HEXAGON - PPM - CAESAR II Dynamics One Lesson Guide

CAESAR II® DYNAMICS ONE
CAESAR II® ONLINE VIDEO TRAINING SERIES

CAESAR II DYNAMICS ONE
CAESAR II® Dynamics One
David Diehl, Course Author

First Edition - December 2018
“© Hexagon AB and/or its subsidiaries and affiliates. All rights reserved
http://www.pipingdesignonline.com
No duplication permitted without express written consent
HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 1

Copyright
Copyright © 2003-2018 Hexagon PPM® Corporation. All Rights Reserved. Hexagon PPM is part of Hexagon.
Including software, file formats, and audiovisual displays; may be used pursuant to applicable software license agreement; contains
confidential and proprietary information of Hexagon PPM and/or third parties which is protected by copyright law, trade secret law,
and international treaty, and may not be provided or otherwise made available without proper authorization from Hexagon PPM
Corporation.
U.S. Government Restricted Rights Legend
Use, duplication, or disclosure by the government is subject to restrictions as set forth below. For civilian agencies: This was
developed at private expense and is "restricted computer software" submitted with restricted rights in accordance with
subparagraphs (a) through (d) of the Commercial Computer Software - Restricted Rights clause at 52.227-19 of the Federal
Acquisition Regulations ("FAR") and its successors, and is unpublished and all rights are reserved under the copyright laws of the
United States. For units of the Department of Defense ("DoD"): This is "commercial computer software" as defined at DFARS
252.227-7014 and the rights of the Government are as specified at DFARS 227.7202-3.
Unpublished - rights reserved under the copyright laws of the United States.
Hexagon PPM Corporation
305 Intergraph Way
Huntsville, AL 35813
Documentation
Documentation shall mean, whether in electronic or printed form, User's Guides, Installation Guides, Reference Guides,
Administrator's Guides, Customization Guides, Programmer's Guides, Configuration Guides and Help Guides delivered with a
particular software product.
Other Documentation
Other Documentation shall mean, whether in electronic or printed form and delivered with software or on Hexagon PPM Smart
Support, SharePoint, or box.net, any documentation related to work processes, workflows, and best practices that is provided by
Hexagon PPM as guidance for using a software product.
Terms of Use
a. Use of a software product and Documentation is subject to the End User License Agreement ("EULA") delivered with the
software product unless the Licensee has a valid signed license for this software product with Hexagon PPM Corporation. If the
Licensee has a valid signed license for this software product with Hexagon PPM Corporation, the valid signed license shall
take precedence and govern the use of this software product and Documentation. Subject to the terms contained within the
applicable license agreement, Hexagon PPM Corporation gives Licensee permission to print a reasonable number of copies of
the Documentation as defined in the applicable license agreement and delivered with the software product for Licensee's
internal, non-commercial use. The Documentation may not be printed for resale or redistribution.
b. For use of Documentation or Other Documentation where end user does not receive a EULA or does not have a valid license
agreement with Hexagon PPM, Hexagon PPM grants the Licensee a non-exclusive license to use the Documentation or Other
Documentation for Licensee’s internal non-commercial use. Hexagon PPM Corporation gives Licensee permission to print a
reasonable number of copies of Other Documentation for Licensee’s internal, non-commercial. The Other Documentation may
not be printed for resale or redistribution. This license contained in this subsection b) may be terminated at any time and for
any reason by Hexagon PPM Corporation by giving written notice to Licensee.
Disclaimer of Warranties
Except for any express warranties as may be stated in the EULA or separate license or separate terms and conditions, Hexagon
PPM Corporation disclaims any and all express or implied warranties including, but not limited to the implied warranties of
merchantability and fitness for a particular purpose and nothing stated in, or implied by, this document or its contents shall be
considered or deemed a modification or amendment of such disclaimer. Hexagon PPM believes the information in this publication is
accurate as of its publication date.
The information and the software discussed in this document are subject to change without notice and are subject to applicable
technical product descriptions. Hexagon PPM Corporation is not responsible for any error that may appear in this document.
The software, Documentation and Other Documentation discussed in this document are furnished under a license and may be used
or copied only in accordance with the terms of this license. THE USER OF THE SOFTWARE IS EXPECTED TO MAKE THE FINAL
EVALUATION AS TO THE USEFULNESS OF THE SOFTWARE IN HIS OWN ENVIRONMENT.
Hexagon PPM is not responsible for the accuracy of delivered data including, but not limited to, catalog, reference and symbol data.
Users should verify for themselves that the data is accurate and suitable for their project work.
Limitation of Damages
IN NO EVENT WILL HEXAGON PPM CORPORATION BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL INCIDENTAL, SPECIAL, OR

PUNITIVE DAMAGES, INCLUDING BUT NOT LIMITED TO, LOSS OF USE OR PRODUCTION, LOSS OF REVENUE OR PROFIT, LOSS OF DATA,
OR CLAIMS OF THIRD PARTIES, EVEN IF HEXAGON PPM CORPORATION HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
UNDER NO CIRCUMSTANCES SHALL HEXAGON PPM CORPORATION’S LIABILITY EXCEED THE AMOUNT THAT HEXAGON PPM
CORPORATION HAS BEEN PAID BY LICENSEE UNDER THIS AGREEMENT AT THE TIME THE CLAIM IS MADE. EXCEPT WHERE PROHIBITED
BY APPLICABLE LAW, NO CLAIM, REGARDLESS OF FORM, ARISING OUT OF OR IN CONNECTION WITH THE SUBJECT MATTER OF THIS
DOCUMENT MAY BE BROUGHT BY LICENSEE MORE THAN TWO (2) YEARS AFTER THE EVENT GIVING RISE TO THE CAUSE OF ACTION
HAS OCCURRED.
IF UNDER THE LAW RULED APPLICABLE ANY PART OF THIS SECTION IS INVALID, THEN HEXAGON PPM LIMITS ITS LIABILITY TO THE
MAXIMUM EXTENT ALLOWED BY SAID LAW.
Export Controls
Hexagon PPM Corporation’s software products and any third-party Software Products obtained from Hexagon PPM Corporation, its
subsidiaries, or distributors (including any Documentation, Other Documentation or technical data related to these products) are
subject to the export control laws and regulations of the United States. Diversion contrary to U.S. law is prohibited. These Software
Products, and the direct product thereof, must not be exported or re-exported, directly or indirectly (including via remote access)
under the following circumstances:
a. To Cuba, Iran, North Korea, Sudan, or Syria, or any national of these countries.
b. To any person or entity listed on any U.S. government denial list, including but not limited to, the U.S. Department of
Commerce Denied Persons, Entities, and Unverified Lists, http://www.bis.doc.gov/complianceandenforcement/liststocheck.htm,
the U.S. Department of Treasury Specially Designated Nationals List, http://www.treas.gov/offices/enforcement/ofac/, and the
U.S. Department of State Debarred List, http://www.pmddtc.state.gov/compliance/debar.html.
c. c. To any entity when Licensee knows, or has reason to know, the end use of the Software Product is related to the design,
development, production, or use of missiles, chemical, biological, or nuclear weapons, or other un-safeguarded or sensitive
nuclear uses.
d. d. To any entity when Licensee knows, or has reason to know, that an illegal reshipment will take place.
Any questions regarding export or re-export of these Software Products should be addressed to Hexagon PPM Corporation’s Export
Compliance Department, Huntsville, Alabama 35894, USA.
Trademarks
Hexagon PPM, the Hexagon PPM logo, CADWorx, and CAESAR II are trademarks or registered trademarks of Hexagon PPM
Corporation or its subsidiaries in the United States, and other countries. Microsoft, and Windows are registered trademarks of
Microsoft Corporation. Oracle, JD Edwards, PeopleSoft, and Retek are registered trademarks of Oracle Corporation, and/or its
affiliates. AutoCAD, ObjectARX, and RealDWG are trademarks of Autodesk. HOOPS is a trademark of Tech Soft 3D, L.L.C. Other
brands, and product names are trademarks of their respective owners.
CADWorx Design Review contains Autodesk® RealDWG by Autodesk, Inc., copyright © Autodesk, Inc. All rights reserved.
Autodesk, Inc., 111 McInnis Parkway, San Rafael, California 94903 ("Autodesk") is a third-party beneficiary to this Agreement to the
extent that the Software Product contains RealDWG, and to the extent that the provisions herein relate to Licensee’s use of the
Software Product containing RealDWG. Such provisions are made expressly for the benefit of Autodesk, and are enforceable by
Autodesk in addition to Hexagon PPM.
Hexagon PPM Corporation

305 Intergraph Way
Madison, AL 35758
Phone: +1.256.730.2000
Toll Free USA: +1.800.345.4856
Fax: +256.730.2048

Welcome to these Special Videos for Learning

CAESAR II® Dynamics!
These CAESAR II Dynamics One training video is the fourth course in our video training
series for CAESAR II software. These lessons were developed by David Diehl, the
Director of Training at the Hexagon Corporation, and are designed to teach you Dynamic
Analysis topics using CAESAR II. We think it's the easiest, fastest way there is to learn
CAESAR II software. As you watch the videos and follow along doing the exercises you'll
learn many of the commands, features, and techniques that will make you more
productive and accurate in your work. We hope you find these topics interesting and
beneficial as you expand your knowledge of CAESAR II and Pipe Stress Analysis.
So, thank you for your commitment to training, and let's get started!

About the Author
Drawing from over 30 years of engineering, technical support, and

training experience, David Diehl has created an outstanding course in
CAESAR II Statics Two. Serving as the Director of Training at the
Intergraph® Corporation, Mr. Diehl is recognized as one of the top
CAESAR II trainers in the world. Now you can benefit from his
experience as he shares with you his knowledge of using CAESAR II
software in a powerful, yet easy to understand course.

TABLE OF CONTENTS
Modal Analysis .............................................................................................................................................. 8
Lesson 1 – Calculating Modes of Vibration ............................................................................................... 9
Video 1 – Introduction to Dynamics ....................................................................................................... 9
Video 2 – Eigensolution, Examples of Modes of Vibration .................................................................. 14
Video 3 – Continuing Examples of Modes of Vibration, Basics of Modal Analysis ............................. 19
Lesson 2 – Dynamic Input Review .......................................................................................................... 24
Video 1 – Lumped Masses and Snubbers ........................................................................................... 24
Video 2 – Control Parameters (Nonlinear Considerations and Stiffness Factor) ................................ 30
Video 3 – Control Parameters (Eigensolution, Mass Model, Sturm Sequence) .................................. 34
Lesson 3 – Results Review ..................................................................................................................... 38
Video 1 – Reviewing Results in Dynamic Output ................................................................................ 38
Lesson 4 – Exercising the Program ........................................................................................................ 49
Video 1 – Up and Down Single DOF Model (from Course Files) ........................................................ 49
Video 2 – Left and Right Single DOF Model and Two DOF Model (from Course Files) ..................... 58
Video 3 – Comparison of Cantilever with One Element vs Ten Elements (from Course Files) .......... 76
Video 4 – Comparison of Cantilever with Hundred Elements (from Course Files) ............................. 89
Video 5 – Complex System with Non-Linear Conditions (from Course Files) ................................... 101
Video 6 – Pipe Lift Off and Friction (from Courses Files) .................................................................. 117
Video 7 – Piping and Structure Interaction (from Course Files) ........................................................ 137
Video 8 – Static Model Adequacy and Determining Acceptable Criteria for Modal Analysis ............ 157
Harmonic Analysis .................................................................................................................................... 164
Lesson 1 – Harmonic Overview ............................................................................................................ 165
Video 1 – Introduction to Harmonics.................................................................................................. 165
Video 2 – Harmonic Load Sources and Effects on Pipings ............................................................... 172
Video 3 – Preparation of INLET PIPING Model for Harmonic Analysis (from Course Files) ............ 181

Video 4 – Reviewing Static Input Model for Harmonic Analysis ........................................................ 191
Lesson 2 – Harmonic Input.................................................................................................................... 204
Video 1 – Dynamic Input Processor, Excitation Frequencies ............................................................ 204
Video 2 – Harmonic Forces, Harmonic Displacement ....................................................................... 210
Video 3 – Control Parameters ........................................................................................................... 214
Video 4 – Static Load Cases and Static Analysis of INLET PIPING Model ...................................... 217
Video 5 – Defining Excitation Frequencies, Harmonic Displacements and Control Parameters ...... 223
Video 6 – Error Checking, Analyzing, and Harmonic Output Processor ........................................... 228
Lesson 3 – Reviewing Harmonic Analysis Results ............................................................................... 236
Video 1 – Effects of Damping, Frequency-Phase Pair, Frequency-Phase Dialog ............................ 236
Video 2 – Harmonic Output Results for Displacements, Loads and Stresses .................................. 244
Video 3 – Fatigue Evaluation of Harmonic Loads ............................................................................. 253
Video 4 – Continuing Fatigue Evaluation........................................................................................... 259
Lesson 4 – Proper Stress Equation ....................................................................................................... 264
Video 1 – Addressing a Fatigue Approach, Inputting Fatigue Curve Data ........................................ 264
Video 2 – Apply Fatigue Data to INLET PIPING Model, Review Harmonic Results ......................... 272
Video 3 – Selecting Frequency Phase Pairs ..................................................................................... 278
Video 4 – User Selects Frequency/Phase Pairs and Reviewing Results .......................................... 289
Video 5 – Conclusion and Closing Comments .................................................................................. 295

MODAL ANALYSIS

LESSON 1 – CALCULATING MODES OF VIBRATION
Video 1 – Introduction to Dynamics
(Reference Video: C2_D_M_L1_V1)
1. We are now beginning a series of videos covering the dynamic analysis capabilities of CAESAR
II. The first step along that way is calculating natural frequencies of the piping system. So, our
first group of sessions will be on calculating modes of vibration. This will be the building block,
which we will use then for harmonic analysis, response-spectrum analysis, and time-history
analysis. So, it's good to get this clear understanding of what we're talking about when we talk
about calculating natural frequencies of a piping system.
2. What we will have in this series of videos is, first of all, a textbook definition of modal extraction.
Won't take too much detail, but it's to get the components that we need to understand later on
well-reviewed and understood at this point.
3. In the next video, we'll talk about developing the input for CAESAR II for modal analysis.
4. Then, another session will be on reviewing the results out of CAESAR II.
5. Finally, the last three sections will all be combined into one or two videos-- what you should do to
make a better model for dynamic analysis, as opposed to a static analysis.
6. We'll look at some of the criteria that are used today for modal analysis-- for example, keeping
your first mode of vibration above, say, 3 to 5 hertz. Then, we'll wrap it up and get ready to start
with actually loading the piping systems.
7. This first session is going to discuss modal extraction out of CAESAR II. It's also called Eigen
Solution-- Eigen being the German word for characteristic. It's a characteristic of the system.
8. Now, when we were working with static analysis in CAESAR II, basically we just worked with the
equation F equals KX, force equals stiffness times displacement or rotation. K was a
big stiffness matrix. For F we had a load vector, and X solved for the position of the piping
system. Once we knew the position of every node in the piping system, we could back-calculate
the forces and moments at each end of every pipe and from there get our stresses.
9. Again, we know our stiffness (K). We know our load (F) -- a constant load, and unchanging load,
and we can determine our position (X).
10. We all know that piping systems do move. They can move from an installed position to an
operating position. Now, that's very slow, and we'll talk about that rate of change later. However,
it could also be a motion caused by a response to what we might call an occasional load--
something that's not always on the system, a changing load. Now, in many cases, these
interludes you get from state A to state B are ignored. So, what if we wanted to know more about
the changing response?

11. Well, if I wanted to get involved in evaluating this change, I have to introduce a new term called
time (T). Now, I could look at individual nodes in the piping system moving with respect to time.
Now, I'm talking about velocity. I could say velocity equals position divided by time.
If I want to work at instantaneous terms or get calculus involved, I could say,

velocity equals dx dt-- change in position with respect to time.
Or V equals X dot-- the first derivative with respect to time, position.
12. Now, velocity too can be changing with respect to time. I could have a constant
acceleration. So acceleration is velocity divided by time.
Or again, instantaneously, it's the change in velocity with respect to

time, or the second derivative of position with respect to time.
Or A equals X double dot.
I'll be using X dot and X double dot just to simply mean velocity and
acceleration.
13. The load in the system can change as well, not just the position changing, trying to get the
equilibrium. So F is no longer constant. I'm just going to say F is also a function of time .
Here we have time involved with velocity, acceleration, and the

changing load.
14. Now, I'm going to still use this equilibrium, but now, equilibrium is changing between different
components. So, we see my equation as shown below:
M X double dot-- mass times acceleration. Acceleration will create a load based on mass. I also
have C X dot, or damping times velocity. So velocity times damping will create another load that's
changing the response. I still have my KX term, and that will all be balanced by the changing
force with respect to time F(t).
15. Now, that was a pretty complicated equation. What we're going to do is simplify it, and not just for
this lesson today, but the program itself will simplify in this fashion.
16. Let's remove damping (C) from our equation. Now, what does damping do? It removes energy
from the system. So we call it a drag, resistance.

17. There are several types of damping that we would have in piping systems. One would be
damping in the material itself. For example, there are special alloys that are used to make bells in
order to improve to sustain. So, that sustain is a function of low damping – the sound lasts longer,
the vibration in the bell lasts longer with lower damping, and that's a good thing.
18. In our systems structures and piping systems, there are documented values for system damping
for welded steel systems. Things like piping systems or welded structures, or bolted steel
structures. The bolted structures have higher damping than welded structures. Think of it as
slipping where those bolts are.
19. In our piping systems, we might have sliding supports. Friction would be a damping term, and that
is like a point damping. Again, we don't handle this directly in CAESAR II; we'll talk about that
more later.
20. You could have hydraulic snubbers in your piping system, and basically, you're pushing a
hydraulic fluid through an orifice. That hydraulic loss is associated with fluid flow, which slows
down the piston motion-- a type of damping.
21. We could also talk about driving down the street in your car. As you drive faster, you will have
higher drag or higher damping associated with your velocity, displacing the surrounding fluid, or
air. So, these are all types of damping that pull energy out of the system.
22. If you have no damping, none at all, there will be no loss to the system, and you'd be able to
support perpetual motion. Perpetual motion machines are basically machines that have no
damping at all. You put some initial displacement on them, then you let it go, and it just keeps on
going forever.
23. Now, we can remove damping from our equation and really not change the concepts that we're
talking about here. So, if I let C equal to 0, I will have M X double dot plus KX is balanced by
some force with respect to time.
24. Well, I'm going to simplify this even more. I'm going to say that for me to evaluate this changing
load, I'm going to say it's sinusoidal, or harmonic.
25. It would be probably easy to conclude that if the applied load is sinusoidal, the response too
would be sinusoidal and at the same frequency. So, I'm going to say that my force is a function of
time is some magnitude F times cosine qt, where q is the angular frequency of the changing load
and t is time.

26. Below we see a plot of it. I say cosine rather than sine. I guess you might expect me to say sine,
but it suits CAESAR II better to work on a cosine wave, where we start from our maximum, and
go through 0 to a minimum, and back up again.
27. Now, just a little more calculus and we'll be ready to get on with the Eigen Solution. If I say my
position of any point in the system is defined as X equals A times cosine qt

again, that's sinusoidal response associated with a sinusoidal load-- then the velocity, or my X
dot, is just the first derivative of this term with negative q A sine qt . This will
be 90 degrees out of phase with the X position.
28. Acceleration as X double dot, second derivative is negative q

squared A cosine qt. So that's 180 degrees out of phase with
position.
Notice that X double dot is negative q squared times X.
If I put that back into my original equation as it stands now, M X

double dot plus K X equals Ft,
I get M negative q squared X plus KX equals F

cosine qt,
Or just factoring it out, K minus M q squared, the

quantity, times X equals F times cosine qt.
29. One more change, let's say F equals 0 , that there's no applied load. You might say, well,
if there is no applied load, how do I get the thing to move in the first place? Well, there's two ways
to get it to move-- that is through applying a load or through initial conditions.
30. I'm going to say that spring with a weight on it, I pull it down a little bit and I let it go. That's my
initial condition. With no damping, it'll just continue on forever.

31. My equation then is as follows:
I'll say no applied load (that 0), and I get the equation K minus M times
q squared, the quantity times X equals 0.
There are two solutions to this equation-- X term equals 0, or term equals 0.
Well, of course, if X equals 0, that's just sitting there not doing anything. That's a solution.
32. This is a nice one-- if I'll let K minus M q squared, that quantity equal 0 , then
I could get any kind of position out of the piping system.
When q squared equals K/M -- and again, remember that q, it was in our equation
-- I can get motion.
When releasing from some initial condition, the system will move forever at an angular frequency-
- that's that q term-- radians per second based on the system stiffness (K) and the system mass
(M).
33. Now, we have the players in place. We are going for a characteristic frequency based on system
stiffness (K) and system mass (M).
34. Our next video, we will take a look at these terms, how they play together, and how we can start
to evaluate piping system model response.

Video 2 – Eigensolution, Examples of Modes of Vibration
1. We have our terms in place, and we're going to be working with frequencies of vibration. These
frequencies of vibration are a function of the stiffness and the mass of our piping system.
Frequencies Stiffness Mass
2. This is all involved with the timing of our response. So, here we're talking the word Eigen solution.
There's an eigenvector and an eigenvalue. Again, Eigen is German for characteristics. So, these
are characteristic solutions, or characteristic frequencies, or characteristic shapes of the piping
system that will be used for evaluating dynamic response.
3. Our equation, again, was K minus M times q squared, that quantity,

times position equals 0. That's just a scalar equation. There is one K,
there's one M, there's one X.
4. We could also turn that into a vector and matrix operation. I can have a large stiffness matrix--
stiffness matrix of the piping system-- large mass matrix for our piping system, and a group of
displacement vectors, and a group of frequencies. So, I could turn this single equation, scalar
equation, into a matrix equation as seen below:
5. I could end up coming with the following:
The eigenvalue equals the square root of the stiffness matrix divided by the mass matrix.
There will be a corresponding displacement vector for each one of the Eigen solutions (|q|).
|X| would be my mode shape. So natural frequency, we're calling |q| the natural frequency, and
|X| is going to be the mode shape in another form.
So that solution, this q vector-- that's not absolute value, that's a vector q-- represents the family
of system shapes that can be achieved with no applied load. he |X| shape and the q value

associated with it. In mathematics, that q vector is the Eigen solution, the characteristic solution,
and the corresponding vector position is the eigenvector.
6. Now, if you would Google Eigen solution, it will show something like this in
Wikipedia. It's a mathematical term-- A minus lambda I, the identity matrix, that quantity times the
eigenvector equals 0. So, the lambda is the eigenvalue and v is the eigenvector. Similar to the
matrix equation that we looked at earlier.
7. Now, in our mechanical systems --this equation is just straight mathematics-- in our
mechanical system, that q is the angular natural frequency in radians per second. In most texts, it
will show it as omega . We'll use omega from now on, but I was using q to just sneak up on
this point.
8. The matching shape, the X vector, for each omega is termed the mode shape. Now, the thing is,
this is a shape. Focus on the word shape, not magnitude. When I calculate natural frequencies,
there is no load involved. That was the whole reason for this thing. We wanted to find a
characteristic shape that could be achieved with no applied load. So, when we talk about modes
of vibration, we are not talking about magnitude at all. It's all the shape. I could talk about the
shape of a sine wave and you'd know what I'm talking about. You'd say, OK, what's the
magnitude? That is added to that shape.
9. Now, we're going to look at several examples, in this case, a very simple
system. This is a single degree of freedom system. I'm watching this mass
bounce up and down with the spring. Now, I called it a single degree of
freedom system. What is that? It's a system for which motion can be fully
monitored by a single node moving along a single vector. So, if I would put a
little LED light on the mass and I turn out all the other lights, all I would see
is the square box bounce up and down. I don't see the spring. I don't see
anything moving, any other point, but that single lighted point moving up and
down. Now, what is the eigenvalue and eigenvector for this system? Or what
is the natural frequency of this system?
10. Well, we talked about it a minute ago. Square root of K/M. So if I said my mass is 100 kg and my
stiffness is 10 N per MM, do the math. Omega equals square root of K/M. Its 10 radians per
second. That's the characteristic frequency.

11. What is the shape? Basically, it's just bouncing up and down. I'd say that's vertical. That's
bouncing up and down in the vertical direction. A single degree of freedom produces one
frequency and one mode shape.
12. Now we see another one. This is a cantilever pipe, let's say, anchored, and I'm watching the
mass move laterally.
My mass is 100 kg. Length of the pipe is 5 m. It's 4” pipe, with a standard wall (wt). Other term I
can use my Young's modulus (E). 2 times 10 to the 8th kPa, 30 million psi. Calculate my I. Now
K, 3xExI over L cubed is a close approximation of the lateral stiffness of the cantilever. That gives
me 15 N per mm. My omega (w), square root of K/M. There's M, there's K. About 12 radians per
second. That's the frequency, the angular frequency, radians per second.

13. The shape would be just wagging back and forth. That's the only thing I can see. It's not moving
in and out of the page. It's not moving up and down, just left and right.
14. Now again, I was talking about the angular frequency in radians per second. That's our omega,
but that's not that common in engineering terms. We usually talk about cycles per second or
Hertz, a natural frequency in cycles per second or hertz. Since there's 2 pi radians per 1 cycle,
we can just divide my radius per second by 2 pi, and that'll give us cycles per second.
15. The first example, 10 radians per second, is 1.59 hertz, and that 12.12 radians per second is
about 1.93 hertz. So, we have radians per second or omega, called angular frequency. We have
cycles per second, called f, cyclic frequency.

16. We could also talk about it in terms of period. It's just the inverse of the frequency, and that's
seconds per cycle, or just called seconds. You have a period of 2 seconds, that'd be 1/2 cycle per
second.

Video 3 – Continuing Examples of Modes of Vibration, Basics of Modal Analysis
1. Now, let’s look at a more complicated system. So, now we see a two
degree of freedom system. I have a LED on M1 mass, an LED on M2
mass. How many shapes and how many frequencies can I calculate?
2. Well, just thinking about the shape now, I could define a shape where
they both move in the same direction.
3. It's rather a low frequency, both the masses moving up, and then both
masses moving down. Both springs breathing together-- opening and
closing, this is extending the whole system up and the whole system
down. That'd be the first mode of vibration.
4. We could also define a second mode of vibration. I'm not going to put a value on it. But this is
where they are working against each other. So, the K1 spring is collapsing, while the K2
spring is extending. It would be at a higher frequency, but it would have that shape, where
they are kind of working against each other.
5. I have two degrees of freedom that I'm watching, and I produced two modes of vibration.
6. Now, let's talk about a cantilever. Now, a simple cantilever can have many, many degrees of
freedom all along the beam. I could go to every molecule on that beam and call it a degree of
freedom which is ridiculous. However, in CAESAR II in your model, I could define a single
cantilever with one node on either end of that cantilever. Since it's anchored at one end,
those degrees of freedom on that end are not free anymore. So, in terms of degrees of
freedom, I will only have the free end moving up and down. It could also move left and right
and in and out, if you wish.

7. So, right now, let's just talk about those three degrees of translational motion. A lot of piping
systems ignore rotational degrees of freedom. They've been considered unnecessary for
inadequate dynamic analysis. So, when I talk about a node in space having really six degrees
of freedom-- three translational and three rotations-- when we get to dynamics, just the three
translation terms are what we'll work with. Now, we could add those rotational terms, but we'll
talk about that later. Now, there's going to be as many modes of vibration that we can
calculate as there are degrees of freedom.
8. We now see an example of a cantilever. In a regular CAESAR II model, I would anchor node
10 and free end at node 20. So, it's the same cantilever that I have down below the 10 to 20--
same length-- but now I have more nodes along my run as seen below.
9. For my first example, how many translational degrees of freedom are in the model shown
below? Well, I've got two nodes, and each of them will give me three degrees of freedom-- X,
Y, Z. That just gives me six.
10. However, node 10 is anchored. I can't move that. It's

not free. So, I'll reduce by those three. That means I
only have three degrees of freedom at node 20. I can
only define three modes of vibration with these three
degrees of freedom. You can see them. 20 could

move in and out. That's a pretty high mode of vibration, because it's so stiff axially. It can
move up and down and also in the Z direction. Actually, 2 and 3 are kind of going to be the
same value, but a different shape.
11. 2 and 3 don't have to be pure Y and Z. I could have one mode going on a diagonal with X and
Y. Then, the other mode of vibration would be in the other diagonal. However, they would be
perpendicular to one another. So, model number one gives me three modes of vibration.
12. Model number two has many more nodes. I've got 11 nodes, and that'll give me 33 degrees
of freedom. Again, node 10 is anchored, so, I'll reduce by those three. That means I have 30
degrees of freedom, and I can calculate 30 modes of vibration.
13. Now, also take a look at the degree of freedom. Again, I was saying earlier that it's like
putting an LED on every node. I can show you a much more complicated shape with all the
LEDs at all the nodes on my beam than just that one out at 20. So this will be something we'll
keep an eye on in CAESAR II. Do we have to add extra nodes along the run?
14. So now, if we look at a cantilever, we see now I'm

showing the first four modes of a cantilever. The
results out of CAESAR II. Now, you can see how
detailed the shapes are.
15. I had a lot more nodes along the Mode 2 run in

order to find that nice, fine shape. It's not all
chopped up. So these are the first four modes of
vibration of a many degree of freedom system.

16. Let's look at those another way. Now, this is the
characteristic shape, but I'm going to just throw
four reference lines along each one of these
cantilevers. Again, I'm not worried about the
magnitude. It's the shape I'm after.
17. So in number one, I can see that all four nodes are going up together.
18. Mode number two, I've got three up and the last one down.
19. Mode number three, we have up, up, down, up.
20. The last one Mode 4, we have up, down, up, down.
21. Given this series, I can define four modes of vibration. Again, don't think about magnitude,
just think about relationship or the shape of the beam itself.

22. Now, we're talking about CAESAR II is a

beam element program. We're following
nodes moving off the center line. The
illustration at the right is from a published
document "Piping Vibration Analysis," and
these images are showing shell modes of
vibration, shell distortion.
23. We do not do this in CAESAR II. We do not

consider the changing shape of the
circumference or the wall of the pipe. We're
watching the pipe move off its center line or
along its center line. So, if you're involved in
this type of evaluation, CAESAR II is not the
tool for you.
24. In conclusion, mechanical systems have basic characteristics based on time. We're going to
call them our modes of vibration, mode shapes, or natural frequencies. We have a family of
characteristic frequencies and their related shapes. They're based on system stiffness and
system mass.
25. We can measure that timing in terms of angular frequency, cyclic frequency, or period. These
modes are all independent of one another. They're all orthogonal. This is going to be quite
important later on when we start evaluating piping systems for load over time.
26. Now, monitoring more points in the system in more directions increases the number of modes
of vibration that can be determined. Talking about the degrees of freedom, more degrees of
freedom, and more frequencies. It increases the accuracy of those calculations-- more nodes,
more degrees of freedom, better calculation. These modes of vibration will be the key
component in calculating system response to changing load.
27. There you have it. We have the basics now. We can go in our next session in CAESAR II and
start calculating natural frequencies of piping systems and their mode shapes.

LESSON 2 – DYNAMIC INPUT REVIEW
Video 1 – Lumped Masses and Snubbers
1. Now that we've completed the overview of modal analysis, let's take a look at the dynamic input
in CAESAR II for calculating natural frequencies of piping systems. What we'll do is, we'll start
up the Dynamic Input Processor and from the main menu of CAESAR II you simply click
on the button Dynamic Analysis. That will get us into the Dynamic Processor.
2. Now normally, you would be running a Static Analysis before you run the Dynamic Analysis.
That's just the natural progression of input. Now, you do not need to have a static analysis on the
machine in every case, but if the model has a set of certain characteristics, you must have a
static analysis available before you can run dynamics. This current PowerPoint screen kind of
shows you what models require a static analysis. If you're going to have hangar selection in an
analysis, those springs must be selected before you go to dynamics. So therefore, yes, you have
to run the static analysis with this hangar selection.
3. If there are any non-linear restraints in your model, that too would require a static analysis to set
those non-linear restraints for the dynamic analysis. All CAESAR II analysis and dynamics are
based on a linear model.
So, if I have a plus y restraint, is that support active or inactive? If it's active, we'll put a
double acting y support. If it's inactive, we remove it completely.
If there's a gap on a restraint, that's another nonlinear characteristic. That would require a
static analysis.
If you have friction.
If you have large rod rotation, these roll restraint characteristics.

Or bi-linear restraints, like soil restraints, with K1 and K2. Those require a static analysis.
4. Now when you go into the dynamic input process, or after you click that button on the main menu,
you will open up the Dynamic Input screen. You can see all these tabs along the top as seen in
the image below. A lot of these tabs have nothing to do with the modal analysis that we wish to
talk about today.
5. What we can do to clear out this large list is to first reset Static Analysis Type.
If you click on this drop arrow next to Analysis Type as seen below, you will see a lot
of different selections. One of them is Modal. Go ahead and select Modal.
Then, that will show you Modal, and it will reduce the number of tabs that are associated with
modal analysis. So now we have a much simpler list of tabs to work with.

6. As you might imagine, once we start looking at response spectrum analysis or time history
analysis, or even harmonic analysis, we're going to be using other tabs in this long list. So now
the rest of these sessions on modal analysis will only work with these tabs you see above.
7. A little background on the toolbar list. We have three symbols -- for Save, Error Check, and Run
the Analysis.
8. I'm in the Lumped Masses tab, and we have three lines of data.
I can modify those lines or Add lines or Delete lines by clicking on these two buttons. These
other buttons that are grayed out, they are not used in modal analysis. We'll talk about them
later in these sessions.
As far as the data that I'm showing in the Dynamic Analysis screen on the Lumped Masses
modal, the first column says Comment (Cmt.). If it's a check mark, that means do not process
the data. So, this is how you could annotate your input if you wish. You could type in any kind
of input on the lines. These examples kind of relate to the different columns that we show.
But you can put the straight text in the columns as well.

9. Now, if I wished to modify the Lumped Mass, I would get the plus sign as shown above, and that
will give me a new line in my data set. It will not check the comment field, which means this will
be live data then for the analysis.
I click the plus sign, and I get a new line, line three. It's open for 1, 2, 3, 4, 5 different
columns of data.
10. Now in the Lumped Mass solver or the Lumped Mass algorithm in CAESAR II, the program will
basically take every run of pipe. In the example, I can call them from left to right node 20, 30, 40
in my model.
The program will automatically break each piece of pipe in half and throw half the mass into the
preceding node, and the other half in the following node. I think of it more as the program creating
a system that looks like a bunch of bowling balls connected by springs rather than pipes
connecting to one another. So, we have these concentrated masses that are separated by the
stiffness between those two points. That's how the entire model will build it up. That happens
automatically.
You would be using this Lumped Masses tab if you wish to modify that mass allocated to each
element in the system.
11. If I say I have zero mass in a certain direction at a certain node, that will zero out the total mass at
that point. If I put a value in number 3 of the Mass column, certainly a positive value, it will add
that mass to the existing node in the mass distribution. For example, if I had a pair of flanges that
I ignored in the static model, I could specify a mass to represent that pair of flanges as an added
mass at a certain node in the piping system. So you can modify that in that manner.

12. You can specify what direction this mass is to be applied-- any
one of the six global directions, or all three translational, or are all
three rotational.
13. One point on rotational mass, if we're going to talk about lumped mass versus consistent mass
modeling; lumped mass does not have any rotational inertia. Consistent mass does.
14. The node where you wish to apply this added mass or to zero out the mass will be specified in
the third column. If I specify a single node in the Start Node column, it will only apply to that
node. But I could actually use the other two optional fields-- optional-- to run through a string of
nodes. Like I said, you can start at node 5 to node 550 by 5's. Then every node between 5 and
550 by 5-- 5, 10, 15, 20, would have whatever mass is specified in line 3. So it's either a single
node or a range of nodes.
15. The next tab that we have here on modal analysis is Snubbers. Now Snubbers are dampers,
they’re like the basic doors when you go into a building that will have a snubber on the door to
keep it from slamming shut, or to control the closure. In CAESAR II, we do not have damping
directly in our analysis. We have stiffness and mass, but not damping directly.
What we do for snubbers, or these dynamic supports, is we specify stiffness rather than a
damping value for these terms. What we're going to do is put a stiff spring-- you specify the
stiffness and the direction and the node number-- and you have two different types of dampers
out there in the field.

They could be mechanical dampers like the one
shown to the right, or a hydraulic damper like the
one shown below. These are devices out in the
system and they allow slow motion, like thermal
growth, but they prevent rapid motion, like a
hammer load or a seismic load coming to the
piping system.
16. The third column is where we're going to do most of our work, is on Control Parameters. I will
not discuss Advanced in this presentation. For the many years that I've worked with CAESAR II,
I've had very little reason to even touch on the Advanced section. We're going to do most of our
work though in the Control Parameters. This is controlling the dynamic analysis. We have a list
of values to set, and a definition of those terms.
The first column says DEF. That stands for default. If I press that button, it converts the
setting back to the default setting. I provide values in the settings, and you can see the
default values listed in the Dynamic Analysis. If I have any question about what they
contain, I can use F1 or the question mark to get help on any one of these cells.

Video 2 – Control Parameters (Nonlinear Considerations and Stiffness Factor)
1. Let's take a closer look at the nonlinear considerations in CAESAR II for dynamic analysis.
Remember that we are basically solving this equation, K minus M omega squared the quantity
times x equals 0. We did that in our first session talking about a simple approach to modal
analysis. This K is a constant stiffness matrix. It doesn't change with position.
2. Our nonlinear models change stiffness with position. So, we have to somehow resolve the lift-off,
or resting supports in a static model into a consistent linear stiffness in our dynamic analysis.
While the program will reset those restraints automatically for us, the nice thing is that if you run
CAESAR II in a static analysis, if you wish to evaluate the dynamic situation for the operating
position, if the pipe lifts-off support in the operating condition, that support would not be included
in the dynamic analysis. If it's resting, it would be included.
3. In the dynamic analysis, you would tell the program what state of the system, static analysis
state, the operating state, the installed state, some upset state, whatever nonlinear configuration
you have for that specific state, that would be the linear state used for the dynamic analysis. Let's
take a look at a few examples.
4. Here we have a simple piping system with a plus Y support. It will be resting on that support in
the cold position.
5. But let's say that there's enough thermal strain here that you could overcome the dead weight, so
that when we look at the operating position, there is a lift-off.

6. So, this is the static operating case. There is a gap between the
support and the pipe. If I use this load case to set my non-linear
restraints, the dynamic model would not have a restraint at all in that
direction. It doesn't matter how much of a gap you get here, if we do
dynamic analysis it could wag tremendously here. But the idea is that
this support is not going to be in the dynamic analysis.
7. If instead the dead weight was still large enough that the thermal
strain did not overcome the dead weight sag, we might have this
operating position where it's actually in contact with support. The
dynamic analysis in this case would be a double acting Y restraint.
It's a linear restraint, so this pipe cannot move at all up or down.
8. Now dynamically, it might want to slap the pipe. CAESAR II does

not do that type of slap analysis. It's either there or it is not. So, we always run a linear model for
dynamic analysis.
9. Here is a situation where we might have friction defined in analysis where the pipe slides when it
goes into operation, and we have a normal load. We can also accommodate friction in a certain
manner in our dynamic analysis. What we do in CAESAR II is that if there is a normal load on a
friction support, we will take that normal load times the coefficient of friction times a value that you
specify. We call it the stiffness factor for friction. And we'll use that to develop the stiffness applied
in CAESAR II. This is not a physical term, stiffness factor for friction, you're not going to find it in
any textbook, but it is an adjustment factor that we have in CAESAR II.

10. Now, what we want to do here is that normal load could be quite large, and that pipe, though it
might be in a dynamic situation, might still slip around at that support. Now, we are not asking for
a zero mean, friction is not used in a one. Typically, users will use a value something like 1,000.
That's what I'd typically use. But it's not that sensitive to the magnitude of that value, the stiffness
factor for friction. 200 will probably give the same results as 100.
11. The idea, though, is that if I have a large normal load at one support and a small normal load at
another support, I'm going to get a larger stiffness term with a larger normal load. And it kind of
makes sense that it kind of distributes this friction stiffness equitably throughout the system. You
will see if you do not have friction in your dynamic analysis that you're going to have wild sliding
over these frictionless surfaces. Friction is not directly used in CAESAR II dynamics. It will be
represented by these points supports, these spring supports.
12. Now some of you might be familiar with ASCE 7, the US structures code. And they state in that
document that friction resulting from gravity load shall not be considered to provide restraints to
seismic forces. So you might have reason to not include friction, but we're not necessarily running
a seismic analysis here. We're just calculating natural frequencies. It's up to you as a user to
determine whether or not-- maybe through specification-- whether or not you're going to have any
accommodation for friction in your dynamic analysis.
13. But this is how you would do it with this stiffness factor for friction. If you're trying to examine the
results in a system out in the field right now, it's not design work but forensic engineering, then
you could use this as a tuning parameter to tune up the results from CAESAR II to match better
what you're seeing out in the field. What is always a question is how right is that number? You're
really guessing at this case. We're trying to match it with the real world.
14. We have some other settings in CAESAR II that works with that stiffness factor for friction. Here's
row 2, by default it says 0. That means there is no friction use. This is that value that I mentioned
earlier, maybe 1,000 or 200 or whatever. As an aside, there are units associated with this. If you
do the dimensional evaluation of that equation I showed you earlier, its units are per inch. So I

have 1,000 per inch. And when you do the math, then it will turn out to a stiffness using the
friction force.
15. Now the first load case, the first row here identifies the load case that will be used to set the
nonlinear restraints. In my example earlier where we had lift off, if I pick a load case where we lift
off a support, if I pick that load case here, that support would not be in the dynamic analysis. If I
pick a different load case where it is resting, that would be a double acting support.
16. Here I'm dropping in the operating case. We are using the operating case to set the non-linear
supports in a model but we will not include friction. In the next video we'll talk about some of the
other items in this control parameters list.

Video 3 – Control Parameters (Eigensolution, Mass Model, Sturm Sequence)
1. There are some other items that are defined on the controlled parameters list. One is how we
stop the Eigensolver. If I have a piping system with n degrees of freedom, I have the ability to
calculate up to n modes of vibration. Are all n modes important? Well, no, not for what we're after
here. What we're after here is the lower frequency modes of vibration because those are the ones
that will develop a lot more bending.
2. A lot of the higher frequencies are axial modes, or shell vibration. We don't do shell vibration. We
work with bending modes of vibration, beam bending. So, in a lot of cases, the interests are those
frequencies below 33 Hertz. That's the default setting in CAESAR II, calculate all the modes up to
and just over 33 Hertz. The reason why we go to 33 Hertz is because that is the typical cutoff
frequency for seismic analysis. With ground motion, the ground can't move fast enough to
activate modes higher than 33 Hertz. The piping system acts like a rigid body above 33 Hertz.
3. So, there are situations, though, where you want to go above 33 Hertz. So, we'd have to change
when we stop the Eigensolver. Do we stop it at 33 Hertz or do we go higher? Well, if you're doing
a time history analysis of a fluid hammer, some high frequency content event, then you might go
well above 33 Hertz, maybe 100 Hertz or so.
4. The reason why we're worried about lower modes too is not only because we get the bending,
because that's where we get a lot of contribution from the environment. There's a lot of
environmental loads, equipment vibration, or flow induced vibration inside the piping system or
pressure pulsation in the line. These types of situations could occur at very low frequencies,
below 3 Hertz or 5 Hertz. So, people will run the CAESAR II just to calculate natural frequencies,
to assure that the first mode of vibration is above, let's say, 3 Hertz or 5 Hertz to avoid these
environmental loads.
5. We have two ways of stopping the Eigensolver for this event. Again, we have the frequency cut
off-- it's the maximum frequency by default, it's 33 Hertz. And we also have another second check
based on the number of calculated modes of vibration. So, if I say I want to go to 33 Hertz, but I
only want the first five modes of vibration, whichever one of these two limits gets hit first will stop

the solution. In many cases, people use a frequency cut off alone. Again, if we zero out the count
for the calculated modes equals zero, then it's not going to do that check.
6. So, if I look in our input screen, we have these two rows, three and four. By default, we do not
use the count of eigenvalues calculated, or the number of modes of vibration. We work mainly
with just the frequency cut off. Again, if I want to go out to 100 Hertz, I change at 33 to 100 and
the program will just slowly crank out the natural frequencies up to 100 Hertz. As soon as a mode
is above 100 Hertz, it'll shut down the solution.
7. The next item on our input screen is the lumped mass method versus the consistent mass
method. For many, many years, CAESAR II only had what we call the lumped masses, and I
mentioned that earlier, that we just cut every piece of pipe in half and throw half the mass to the
From node, the other half to the To node and then we have, like I said earlier, a bunch of bowling
balls moving around in space connected by the springs representing the pipe stiffness between
these bowling balls. That has been in the program from day one.
8. There are some more sophisticated methods. One is called the consistent mass approach. And
that is now in CAESAR II. The machines are much faster and bigger, they can handle that fully
developed mass matrix that we have with the consistent mass. It is a more accurate method for
calculating modes of vibration, but you still may have to add extra mass points, breaking longer
runs into shorter runs. That might still be required for the consistent mass in order to develop the
shape of the mode of vibration. Consistent mass is more accurate, but you still may have to add
more nodes in the system in order to get a more representative mode shape with each frequency
calculated.
9. Here is a description of the size of this mass matrix. If I have lumped masses, we just work with
the main diagonal. This is, let's say, node one, node 2, node 3 in our system, and we're
representing the mass only on the main diagonal of this mass matrix-- only translational terms.
No rotation. Here is the second node, no rotation. Everything off diagonal is 0. Very small matrix
represented by the non-zero diagonal.

10. In a consistent mass matrix, there is other terms that are populated in this mass matrix. So it's a
much more bulky set of data to carry but, like I said earlier, today's machines can handle this
quite easily. We still default to the lumped mass. I would suggest that you change this to the
consistent mass method. Click on that cell, you get a drop arrow, and change it to consistent
mass to use the consistent mass approach.
11. Now, when we look at the modal extraction, there's one more check or one more item to specify
in the control parameters, and that is to whether or not to perform the Sturm sequence check.
This is a back calculation on the calculated frequencies. Think of the Eigensolver as a search
routine, where the program will guess what the next frequency is and then see if that solves the
equation n minus omega squared k equals zero, with Omega squared being the frequency it's
using. So, at times, the search routine may discover natural frequencies out of sequence. It might
do 1, 2, 3, 5, 4-- first mode, second mode, third mode, fifth mode, fourth mode. If that second to
last mode calculated is beyond the cutoff, the program would shut down and it might skip over
one mode that it should have calculated.
12. The Sturm sequence check will identify where this happens, where the solution is out of
sequence. It's a very cheap insurance if you're not missing any modes in the analysis and also
say that, with today's PCs and the higher precision in CAESAR II, this is not a major problem
anymore. But that's the last item, yes or no, do you want to perform the sequence check? And I
would say just leave it alone, leave it at yes.

13. So, there you have it. Basically, we had very little to talk about in Lumped Masses and Snubbers
and we reviewed the Settings and the Control Parameters for the input. First one is for non-linear.
The second is also a non-linear consideration for friction. Three and four work as a pair,
whichever one you hit first will shut down the solution. And then five is whether it's lump or
consistent mass. And then, finally, the system mass or the Sturm sequence check, whether you
wish to run that or not. Just takes a second or two, so I'd leave that alone. In the next video, we'll
talk about results review in CAESAR II.

LESSON 3 – RESULTS REVIEW
Video 1 – Reviewing Results in Dynamic Output
1. Now let's take a look at the type of results you should expect to see when you run a modal
analysis in CAESAR II. If I look at the output menu, this is a different menu than what we've seen
in static analysis. It's a simpler menu, and if we're looking at a modal analysis, there is not even a
load case specified; there are no reports to look at.
2. Modal analysis doesn't calculate response to a load. It is just the system characteristics that we're
calculating. What is the first mode of vibration? What's the shape of that first mode? What's the
third mode of vibration? What's the frequency in the shape? So all of our results are in the
General Results and in the Animation .
3. Taking a look at this window. If I look at my toolbar, we see:

• Open.
• Save the report as text.
• Send to the printer.
• Animation modes.
• Go back to the input (either dynamic input or static input).
• You can edit the titles of the dynamic load cases.

• We can review the dynamic load cases.
• We can view the selections on the screen.
• Or send them to Word.
These are the different tabs we have up on the toolbar in the Dynamic Output.
4. In these presentations, we're going to bring the output right to the screen, not to the printer or to a
text file or to Microsoft Word. We're going to focus our attention on the different columns or the
different values in the right column.
5. Again, no load, no reports, all of our results will be these General Results.
6. So, what we're going to do now is review these different types of general results in the dynamic
output processor.
7. The first one, Natural Frequencies, we see in the model data below. We have six modes of
vibration that were calculated. Apparently, the frequency cutoff was still set to 33 Hz. We get the
first one that's over 33 Hz listed as well. We show frequency in cycles per second, then frequency
in radians per second, and then the period of vibration.
8. These are related to one another. We see our frequency in cycles per second or Hz.

9. Now the radians per second, and there's 2 pi radians per cycle. So we can do that math to get
from 0.4 to 2.9.
10. Then finally, the inverse of the frequency or cycles per second. So that's 0.5. This should be 2.
So we see how those inverse. All right?

11. The next two items in that final column are the mode shapes. These are
just numbers associated with each mode of vibration. Each mode shape
will be listed, first mode, second mode, third mode, and a long list. There
are two different reports. One says Modes Mass Normalized and the
other one says Modes Unity Normalized.
12. Those reports are basically the same, but the Mass Normalized report includes the amount of
mass that is participating in that mode of vibration. The Unity Normalized is the same shape, but
it's all normalized to 1.
13. So, if this is the mass normalized, if node 200 is the maximum response of 1.6 in the y-direction,
in the unity normalized, it would be 1.0. That's the maximum value and everything's ratioed down
below that.

14. We now see that same report for Unity Normalized. We have the same shape. This is a y mode
of vibration, looks like a cantilever, and you can see that same shape in the Unity report, that it's
normalized to 1, not 1.6.
15. Typically, if you're going to look at these reports, you'd be looking at the Unity Normalized to get
an idea of the shape. Frankly, we get more information by looking at the animation, rather than
looking at these numbers, but they're in our output.
16. The fourth item in the General Results is the Dynamic Input echo. This will give us a look at all
the input that was used, the dynamic input for this analysis. So if this is the same model, yeah, we
see my frequency cutoff of 33 Hz. I'm using consistent mass, and I want the Sturm Sequence
Check. It must have been a linear model because I did not have to specify the static load case for
non-linear restraint, and there's no friction in the model.

17. The next item, the Mass Model. We can see the mass model below that has been used in the
analysis. Now, again, I was saying that this example is showing us a cantilever, and it only has
mass in the y-direction. So apparently the mass in the x-direction and the z-direction was zeroed
out by the user in the input for this model. We see a nice consistent mass, 26 to 53. So if that
(26.9469) is 1, the 53.8938 values are 2, 2, 2, and on, and then finally 1 again. So we can see
how we can break down the mass between all the nodes in the model.
18. In the Consistent Mass approach, again, x and z have been zeroed out by the user, but we see
a different list of values for the y-mass or y-inertia and we also see we have some rotational
terms in this mass as well. So, this is the more sophisticated, more, I'll say, correct mass
distribution for the analysis. Although, many people have been using the lumped-mass model for
many years and they've worked just fine as well.

19. The next item is the Active Boundary Conditions in the model. Now again,
if the model is non-linear, this is the report you would go to see which
restraints are active for the dynamic analysis. Below you can see an
example, a piping system that has two non-linear supports at 30 and 40.
They are plus y supports. We also have a restraint at node 10. That's an
anchor.
20. If I look at the operating case, this is static analysis. It actually does lift off of node 30. And it's
resting at 40. So if I use this operating configuration to set the supports for the modal analysis,
then I would want to see 30 being inactive and 40 being active, and I would look at the Active
BCs report to see that.

21. Now again, I'm going to choose the operating state in order to collect the 30 and 40 numbers
from the operating static analysis.
22. If I look at my Boundary Condition Report, active boundary conditions, I'll see that node 30 is 0,
meaning that there is no support at node 30 in the dynamic model.

23. However, at node 40, we see it has a value of 10 to the 12, or 1 times 10 to the 12 and that is our
rigid stiffness. So we show you the stiffness for each restraint. If I had friction in the model, the
friction restraints would be shown as well with their stiffness values.

24. Finally, we have the input echo. This is the input echo of the static
model or the original model that was built up. We have selections in the
Input Listing Options, they say, show me all the elements, show me
the allowable, and show me the restraints. Most people would go back
to their static solution that to look at these values. But again, this is the
basic model that was built that was used for the dynamic analysis.
25. One final item that we have on our output menu for dynamics is the Animation button. I
mentioned earlier that we have our modes of vibration here. Most people find the animation is a
much simpler way to get a good view of the overall response of a system to an event or, in this
case, the actual shape of the system for each mode of vibration.
26. There you have it. Now, what we're going to do next is go through several examples. Some
examples that we built up earlier in this presentation and some additional examples to exercise
the program to better understand modal analysis in CAESAR II.


LESSON 4 – EXERCISING THE PROGRAM
Video 1 – Up and Down Single DOF Model (from Course Files)
1. We reviewed the background on modal analysis and then took a look at the CAESAR II input for
modal analysis and the type of output we get. Let's take a look at some examples now with the
program. So first I want you to prepare your machine.
2. With the set of instructions you received, there are several files from CAESAR II that you should
have installed on your machine. You could be running 2017 or 2018. You can use the files that
are provided directly in the folder 2017.
3. If you are running an earlier version of CAESAR II, go to the folder 2011, and then that will give
you the files that you will need in order to run this analysis. Take those files, and put them in
whatever your working folder is for your video training series.
4. These models should be run in your version, so you may have to run all the static models before
you get to the point where you can look at them in this video series. Once you get that set up,
come back and watch this video.
5. Startup CAESAR II. The first model that I have loaded up is called Up & Down. So if I go back
to CAESAR II now, we see my input of my main menu for CAESAR II, and I'm pointing to
the file called Up & Down, as you see in the image below.
6. I'll go to the input file. Click Piping Input.

7. Then, we see the model as shown in the image below. It's not a big system.
8. Let's go back to our PowerPoint, and I'll show you what Up & Down is all about. So, we see the
first model. It's called Up & Down, and we have a simple mass that can only bounce up and
down. This is a single degree of freedom system. We spoke of that earlier on. Single degree of
freedom, all this mass can do is wag up and down. So we have a stiffness and a mass.
9. What are our terms? We saw this calculation earlier. I went a mass of 100 kg sitting on a spring
with a spring rate of 10 N per mm. I can quickly calculate my natural frequency by hand, square
root of K over M. That's 10 rads per sec, and with 2pi rads in one cycle, I can say 10 rads per sec
is the same as 1.592 Hz.

10. Now we have to go see what's happening with CAESAR II. As far as the mode's shape, it's going
to just bounce up and down. There's no interaction with the horizontal directions, and it's a rather
a simple mode shape.
11. Again, let's go back to CAESAR II now and take a look at our
data. Again, we want 100 kg on a 10 N per mm spring. So it's a very
simple model. 10 to 20 is one element.
12. Click Next Element, and we see there's only a second element, 20 to 30.
13. Just as an aside, I had to break it up into two elements in order to get six restraints to work with.
We're going to have rigid restraints in five degrees of freedom at the anchor, and one in the y
direction will be 10 N per mm. Since I needed six restraints and in earlier versions we were only
allowed to have four of restraints on an element, we had to have two elements in order to get
enough slots in order to load up all six.

14. If you're running 2018, now we know we have the ability to specify six on one screen, and I could
have made it with a simple model.
15. Let's look at the first element, click Previous Element, 10 to 20. It's very short. It is a rigid
element, click Rigid. What's the weight of that element? It is 490 N.
16. There are two elements in this system. Click Next Element and we see the second element, 20
to 30. This one too is 490.33 N. What does 100 kg weigh? It weighs 980.666 N. So I put half of
that weight on each one of these two rigid elements.
Again, in dynamics, we work with mass. In CAESAR II, we're specifying weight. This is in N, and I
now have a total mass of 100 kg.
17. Regarding that stiffness in the y direction, again, the other five restraints are all rigid. Click
Previous Element, and then click Restraints. However, in the y direction, you see 100 N per
cm, again, 100 N per cm. We wanted 10 N per mm. So 10 N per mm is 100 N per cm.

That's the system we want to run. That's all it is. Just has two elements bouncing on top of the
spring in the y direction.
18. I'll do the error check. Click Start Run.
19. We have to at least run the error check here to present the data for the dynamic analysis. There's
the total weight, which is 980 N. The Total dead weight is 100 kg.
20. Now, we do not have to run a static analysis. It’s a linear model. I don't have to get any static
results for the dynamic analysis. I can go right to the Dynamic Analysis. Let's do that. Click on
Dynamic Analysis.

21. That brings up this window shown below. My data is already specified. We're going to do a modal
analysis.
22. I did not modify the mass model.
23. Click Snubbers. I did not modify the additional dynamic restraints model.
24. Click Control Parameters. I've specified, only calculate the first mode vibration. Well, this is a
single degree of freedom system. We only can calculate one mode of vibration. If I would have
left row three (Max. No. of Eigenvalues Calculated (0 – Not Used)) at 0, the program would
have had trouble trying to get out to 33 Hz, because there are not enough degrees of freedom to
get out there.

25. The other change that I made in Control Parameters is at row five; I changed the Setting from
lump to consistent mass, again, the more accurate calculation.
26. Click the running man.
27. The program then goes onto the analysis. We see the Dynamic Eigensolver screen. We also see
the first mode of vibration, 1.59.
28. I'm now in the output processor for dynamics for our model. Once again, we did not load the
system. So, there are no displacements, no forces, no moments, and no stresses to review, but
we do have our general information. These are our frequency calculation.

29. I'll look at, first of all, the Natural Frequencies. Click Natural Frequencies. Click on the second
to last button, View Reports.
30. We then see our data. The first mode of vibration, the only mode of vibration, at 1.592. That's 10
rads per sec or as a period of 0.628. It can finish one cycle in 0.628 seconds.

31. Now, if we go back to our PowerPoint presentation, that's the number we wanted, 10 rad per sec
and 1.592. Very simple situations, so yes, CAESAR II is predicting that natural frequency
.
32. Going back to CAESAR II now, we have other reports available to us. With this simple model,
they would be kind of nonsensical to take a look at. We'll review these results in some of the
more significant models. In the next video, we'll start taking a look at some of the other models
that were reviewed earlier in this presentation.

Video 2 – Left and Right Single DOF Model and Two DOF Model (from Course Files)
1. We'll move on to the next example. Now in this example we see we have the mode of vibration
where the cantilever just swings to the left and right. Our mass is 100 kg again. We have a pipe
length of 5m, OD, wall thickness, Young's modulus. We can calculate the stiffness of the
cantilever, and its 15 N per MM. With my mass of 100 kg at the tip, I can then calculate my
natural frequency, squared of K over M, 12.122 radians per second or 1.929 Hz.
2. We go back to CAESAR II and click File, Open.
3. Select Left & Right.C2. That's the model we're running now, Left & Right.C2. Click Open.

4. Again, you probably have to run through at least the error checker in this model in order to get it
for your proper version of CAESAR II. I'll go into the input processor. Click Piping Input.
5. This is one piece of pipe, 10 to 20. We can see an anchor at 10, and also our length, OD, and
wall thickness.
6. Click Restraints. We see I just have an anchor at 10. I'm not worried about the mass because
I'm going to put a 100 kg mass at the tip in the dynamic processor.

7. We'll do the error check on this pipe, and we have no messages. I don't care about the total
weight. Click Start Run.
8. Now we'll go into the dynamic input. I'm in the

static input. I don't have to go back to the
main menu; we have an option for the
Dynamic Analysis on the toolbars. That option
became available to me since I did the error
check. Click Dynamic Analysis.
We now see the dynamic input.
9. We are going to make a modification on the Lumped Masses. Firstly, I'm going to use, (let's go
to Control Parameters), and I'm going to use Consistent Mass.

10. I now have responsibility for the rotational masses as well as the translational masses. Click
Lumped Masses to go back. First I will 0 out all the mass in the model. The All we see at
number 4 means all translation. They all go to 0. All the rotational masses go to 0 as well. This is
from node 10 to 20 by 10. This is node 10 and 20. They're massless, both translational and
rotation, but then I added a 100-pound mass in the x direction only at node 20.
11. Click Snubbers. We see I have no snubbers.
12. Click Control Parameters and once again, I want to calculate just the first mode of vibration and
I'll use consistent mass.
13. I can then click Start Run (the running man), but before I
do that, we could also, if you've got a bigger model with,

let's say, it's not modal analysis. You may want to run just the error checker by itself. I'll do that
this time. Click the red x.
14. We then see that the program says yes for all that data, and it says OK, this job can run. You can
use this as a means for reviewing your model before you actually submit it for analysis.
15. I'm now going to click Start Run (the running man).
16. That does the same thing, but now it continues on and does the analysis.
17. We then see the analysis, 1.928, and in our PowerPoint 1.929. We got the same number.

18. I'll look at the CAESAR II Dynamic Output menu now. I get Natural Frequencies. What numbers
do I want to see? I want to see 12.12 and 1.92.
19. Select Natural Frequencies, and then click View Report.
20. We receive an output of the report with the frequency, 1.928, and you can see the hand
calculation's 1.929 in our PowerPoint.

21. We see radians per second, 12.117, and in the PowerPoint 12.122.
So yeah, again, a good agreement with the between theory and CAESAR II output.
22. Let's take a quick look at our mode shapes. We're expecting this thing to be wagging back and
forth in the x direction.

23. Going back to the CAESAR II output, I'll close out The Dynamic Output window, and now
I'll click Modes Unity Normalize. Now, this is going to be a very simple report. It lists nodes 10
and 20 only. Click View Report.
Looking at that report, yes, it is moving in the x direction. 10 doesn't move at all, that's the
anchored node. Then 20 is that free node at the end, and that's a single degree of freedom
system. That's just all we can show.
24. Now, if I had put more nodes along the column in my model, I could have a nice mode shape
showing in this data. First mode of vibration of a simple candle lever bending. We'll look at that in
some of the other models. That's all we want to show in this model. So, let's move on to a two
degree of freedom system.
25. We looked at two single degree of freedom systems, and we're now moving on to another model,
that looks like this image below. We're going to take a look at two masses. One mass is M. The
other is twice as large (2M) with stiffness (K) on either side.

26. This comes out of a textbook. Basically, we're going to see one mode where they both move
together. I'm expecting by hand calculation for it to move at 15.6 Hz. Then we'll have a second
mode of operation at 30 Hz where we're looking at, instead of moving the same direction, one is
moving positive while the other one is moving negative.
27. Let's go look at some worksheet 2DOF Modal Values.
28. We see simple calculations for the first two that we already looked at, Left & Right.C2 and Up &
Down.C2.

29. Now we are looking at the 2DOF.C2 system. This is from Thompson textbook that I used in
school for dynamics.

We see we have two masses. One is twice as big as the first

with a spring rate on either side.
We also see my pipe size, so I can then

use that to give you my stiffness.
There's my stiffness in US units.
30. I can then plug them into these equations from Thompson, and it'll say that my first frequency is
15.55 Hz. The mode shape is 0.73 for the first one, while the other one is moving 0.73 for the first
degree of freedom, at 1.0 for the second one. So that's my normalized mode shape.

31. Then for the second mode of vibration, at 30 Hz, this one is going minus 1, while this one is going
positive 0.366. So that's why on the graph in the PowerPoint I show the shorter run at Mode 2.
That's 0.7 So that's 1.
.
This is 1. That's 0.366.
32. Let's go back to CAESAR II. I'll click Home > Open.

33. Select 2DOF.C2, and click Open.
34. We'll take a look at the piping input. Click

Piping Input.
35. We have two nodes in this model. Click Node

Numbers.
An anchor at 10 and an anchor at 40. There's 20 and 30, DX is 10 ft.

Click the line between 20 and 30. I want the same stiffness between them all. They're all
going to be 10 ft. long. Again, we're going to put our masses in the dynamic model. So that's
my static model, so just our 4” pipe running 30 ft. with different masses in it.
36. We'll do the error check on this model, so we can run dynamics.
Click Start Run. There is no static analysis required in this
analysis.

37. Now we'll go to dynamic analysis. Click

Dynamic Analysis.
38. We see our dynamic input. Again, we cleared out all the mass at 20 and 30. Those are the two
nodes that can move.
39. Then we added a mass of 200 lbs. at node 20 and 400 lbs. mass at node 30.
40. Click Snubbers. No snubbers.

41. Click Control Parameters. We're going to run the lumped mass model this time. We're going to
ask for only two modes of vibration coming out of this analysis, and we only have those two
nodes 20 and 30 to work with, and we only have mass in the x direction (which you can see in the
Lumped Masses tab). Basically, we only have those two modes of vibration available anyway.
42. Click Start Run (the running man).
43. Then, the program will go on into the dynamic analysis. We see the two modes of vibration.
44. Now I'm going into my output processor. Again, what are we looking for? I want to see a
frequency of 15.6 and 30, and I want to see the two mode shapes.

45. We'll look at our frequencies first. Click Natural Frequencies, and then click View Report.
46. We get our results and we see 15.553 in the software output, 15.6 in my PowerPoint. 30.046 in
the software, 30 in the PowerPoint. That's in radians per second also at the period.
47. I'll now look at the Modes Unity Normalized, and I want to see these numbers now. Click Modes
Unity Normalized, and then click View Reports.
48. Each mode is listed separately, mode one and mode two. Node 10 and 40 are my anchors.
They're not moving at all. We're only moving in the x direction. There it is, 0.732 and 1. I don't
really worry about the signs, but I see 1 in the Mode 2 in the PowerPoint and then the opposite
sign, 0.366. I'm positive 1 and minus 0.366. That's the same as negative 1 and positive 0.366.

49. There we have a two degree of freedom system. Let's now step on into some more complicated
systems, more piping systems in our next video.

Video 3 – Comparison of Cantilever with One Element vs Ten Elements (from Course
Files)
1. A few more examples in dynamics, modal analysis. We'll go back to our presentation, and now,
we're going to take a look at two cantilevers. Same cantilevers, one of them is just a node at the
From end and To end. There's one element, and a second one with 10 elements. So just break it
down into a more detailed model.
If you recall earlier, I mentioned that, when you want to get a more accurate analysis, you will
break down the pipe into shorter segments. That will give us more information to better
calculate the mode shape and the natural frequencies as well. Let's take a look at both of
these cantilevers.
2. First, let's open up the program. This first model is called Cantilevers 1. I'll click Open.

3. Select Cantilevers 1, and then click Open.
4. We'll take a look at the input. Click Piping Input.
5. We see just one piece, that's the entire model-- 50 ft long, 100, 200, same 4” pipe.

Anchor at 100. Click Restraints to view the anchor.
6. Click Start Run to run the error check.
7. Then, click Dynamic Analysis to run

the Dynamic Analysis.
8. We're not eliminating mass.

9. Click Scrubbers. No snubbers.
10. Click Control Parameters. We'll stick with the lump mass this time. We'll let the machine do its
own thing, and let's see what happens.
11. Click Run the Analysis, (the running man). It says OK,
and it keeps on running.
We see the first mode of vibration listed in the Dynamic Eigensolver.

12. Now if we have more nodes, we can have more degrees of freedom. You can see in the
Eigensolver that we only have one frequency calculated, and the Sturm sequence check failed.
We see it says it couldn't find all the modes that it should have calculated up to 0.3279 Hz. Well,
we didn't remove any mass. I have both the vertical and horizontal modes at the same frequency.
It's a symmetric system. I actually have two frequencies at this exact term-- 0.32793. One will be
in the y direction. It will be in the z direction, or perpendicular to the first one. We only calculated
one, and the machine is telling you that you should have had two. We'll just work with that. Click
OK on the warning, and Cancel on the Dynamic Eigensolver.
Click Yes in the Eigen message.
We see that for this lumped mass model with only one frequency calculated, the program is not
able to continue on.

13. So, what other data is available to us? Well, we’re in the output processor, and we only calculated
one natural frequency. Click Natural Frequencies, and then click View Reports.
We see the value from that report is 0.328 Hz. This cantilever will wag up and down at one
cycle for three seconds. That's just with 1 degree of freedom at the end of that cantilever.
14. So, let's now build it with 10 nodes along the run. The Number two model in our PowerPoint--
now we have that model on the machine already. Let's go back to CAESAR II to open up
cantilever 10. Click Open.
15. Select Cantilever(10).C2, and open that.

16. Go look at the input. Click Piping Input.
17. It still starts at 100 and ends at 200, but we broken into five foot segments. Click to get it all
on the screen. There's the entire model.
18. Let me clean up these properties a little bit. Right-click, and then select Properties > Display
Options.

19. The nodes are a little large. Let's go down to 20 point. Click the Node ellipsis, change the
Color to Black, the Size to 20, and then click OK.
Apply that, and now close out of Plot Settings.
20. Now we have the black nodes along the cantilever. So not just one node at the end, but we have
all these nodes as a more continuous model, and we're going have a better mass model. We're
going to have more nodes in order to calculate or to show mode shapes.

21. Let's error check this. Click Start Run.
22. Once again, go right to dynamics. Click Dynamic Analysis.
23. Now again, in these simple models I'm deleting masses in certain directions. You will not normally
do this in an analysis. You'll leave the masses alone, but I'm trying to make it a simpler model.
The only mass that we're looking at is the cantilever moving up and down. We eliminated the
mass in the z and the x direction, across the entire run from 100 to 200 by 1. Any node between
that series from 100 to 200 will not have any mass in the z direction nor in the x direction.

24. Click Snubbers. No snubbers.
25. Click Control parameters. We're going to calculate all the natural frequencies up to 33 Hz,
lumped mass again.
26. Click Run the Analysis, (the running man).
We see the error check.
27. Now the software moves onto the analysis and we see the Eigensolver calculating. We see our
five modes of vibration. Also, in this analysis, we're only working in the vertical direction, so we
did not get enough degrees of freedom in order to finish out the analysis.

28. Let's go look at those results. Click OK, and then in the CAESAR II Dynamic Output window,
select Natural Frequencies, and click View Report.
29. We now see the five modes of vibration starting at 0.4 cycles per second. You could compare to
that first one. This is more accurate than the first analysis with only the end note.
30. We can also take a look quickly at the modes. Close the Natural Frequency Report, select
Modes Unity Normalized in the CAESAR II Dynamic Output window, and click View Report.

31. We see the first mode of vibration, a simple cantilever first mode of vibration. Everything's on the
positive side or negative. This one is all positive. There's no going negative on this. The
maximum values at tip is one.
32. The second mode, negative 1 for the tip, and we go through zero. We see positive 0.3 and
negative 0.05. So that's again, typically that second mode of vibration for a cantilever.
33. The third mode, we go through 0 twice. First at -0.0584 and then at 0.4579.

34. Then the fourth mode more and more onto the fifth mode. We're getting more shape out of this at
these higher modes. That's why we need these extra degrees of freedom in this model to make it
do so.
35. We can also look at those modes of vibration in

animation. Click Animation.
36. I'm going to look right down the z-axis. I'll click the letter z, and we’re looking down the z-axis.
37. I'll put the node numbers on it, 100, 200.
38. The first mode of vibration-- what does that look like?
Click Volume Motion.
It looks like the image below, first mode.

39. What does the second mode of vibration look like?

Select Load Case 2, and Click Volume Motion.
We see our nodal point right around the area indicated below. It doesn't move up or down.
There's no restraint at that point. It's just that's the nodal point of that mode of vibration.
40. Third mode. Select Load Case 3, and Click Volume

Motion.
41. We have two nodal points, one at 150 and one at 190.
42. The fourth mode and so on. Select Load Case 4, and
Click Volume Motion.
With these extra nodes, we can show the shape you see below. If we only had a node at the
end, we could not define the shape. So again, adding nodes may be important in your
analysis. Then we can go on from this.
43. Now, the exercise continues; there's one other model we can take a look at it, and we'll do that in
the next video.
Video 4 – Comparison of Cantilever with Hundred Elements (from Course Files)


1. Here's that model that had 11 nodes along the run. What would happen if we had more nodes
along the run?
2. Well, let's do that. Let's put 100 elements along this run. I'm
going to go back to my original model. I'll open up
CANTILEVER(1), open that, and go in the input. And again,
just a single element. This is called CANTILEVER(1).
3. I'm going to immediately say File, Save As. I'll click on CANTILEVER(10) and I'll make that
CANTILEVER(100). So, I put another zero in there. This will be the new job. Save that.

4. Now I'm going to just break this element. We can click on the
Break command, break it. And I want to insert multiple nodes.
These will all be equally spaced. This is a good way to add extra
nodes to a model for a better dynamic analysis.
5. Now, we're going to overkill it here. I want a total of 100 elements in this run. So I type in 100. Or I
could have said node step of 1. And then it would have gone from 100 to 101, 102, and stop at
200. But this will do it.

6. And it did one step anyway. Every element is six inches long. Now, this is way too defined for a
sensible analysis. But it shows you how continuous we can get with our mass distribution.
7. I'll say OK with that. So, I've got all these nodes along this run now. You can't even read the node
numbers, there's so many in there. So, I got my numbers. Do the error check. Again, we already
changed the name. This is now 100.
8. If I go to my Dynamic Input, it also duplicated the dynamic

input.
Here's my existing dynamic input for job 100. Once again, no change in the mass model.
Snubbers-- nothing in there.

And I wanted to go out to 33 Hertz with a lumped model.
9. Click the running man. We got our error check.
It goes on to the analysis. Now, there it did get over 33 Hertz. The 10 node problem didn't have
enough degrees of freedom to finish out that cut off of 39 Hertz.
10. We still have a Sturm sequence failure here. Again, this system is symmetric. I get the same
response in the y direction as I get z direction. So they usually work as pairs.

11. Note these other ones up here. There's two. 7 and 8 are both at 16 Hertz. 9 and 10 are at 26
Hertz. I'm going to say 11 and 12 are at 39 Hertz. So, the Sturm sequence check says you've
calculated 11 frequencies at or below this number right here, when in fact there are 12 in this
analysis. And that's what triggered this error.
12. The results are fine for what we have. So, we're going to say OK to that and just let it go on to the
output processor. And we'll take a look at the output that we do have.
13. Here I'm in the output processor. Select Natural Frequencies and click on View Results.
Here are my 11 frequencies. And here we see the pairs. 1 and 2 are 0.4. 3 and 4 are equal. 5
and 6. So again, 11 and 12 would be equal as well.

14. Let's go look at our output for that. In terms of other reports, let's look at
our animation and look at mode number 1, click on Volume Motion.
Now, I can't see which way that's going. It looks like it's getting longer and shorter. It might be
doing that. But I think it might be moving also in the x-- the yz direction.

15. I'm going to look down the axis of the pipe, look down the x-axis.
Click on Left View, and I like the way this works. The first mode of
vibration is working along this direction here.
16. Now, the second mode of vibration is another independent solution of this system. All modes are
orthogonal to one another.
17. If the first mode is working in this direction, the second mode must be working in this direction. It
doesn't have to be in pure y or pure z. But they have to be perpendicular.
18. I'm going to look at mode number 2 now. And just try to imagine-- it looks like right on the 45. If I
look at the second mode, there's the perpendicular. Same frequency, perpendicular response.
And all these modes are going to do the same thing.

19. Number 3, at a higher frequency. Now that one is horizontal.
20. The fourth one, its matching pair, will be vertical. As long as they're perpendicular, they are
unique solutions.

21. We can look at those also then and along the original direction. And there's that nodal point. So,
we got all these extra nodes on here, and a nice, smooth, continuous type of a plot.
22. We could also look at these in the table. Modes Unity Normalized. Show me those results. Here's
the first mode of vibration. And the tip has the maximum response. And it is not purely on the 45.
If it was, this 1 and 1. So it's almost 1 and 1.

23. The second mode. Here is the 1 in the y direction and the almost 1 in the z direction. They are
perpendicular.
24. Here's a third mode of vibration. And that is almost purely in the z. There's a little bit y
component.

But again, the perpendicular one (Mode 4) will have the same magnitude switch, but a different
sign on one of them.
25. That's the 100 node model that will give us 300 degrees of freedom with a lumped system. So
those are simple cantilevers. Let's get something more realistic in terms of piping systems. We'll
do this one in the next video.

Video 5 – Complex System with Non-Linear Conditions (from Course Files)
1. Let's now take a look at the modes of vibration, the natural frequencies of a piping system. We're
going from this anchor over here to these two pumps down here up to another anchor on this
side. There is no friction in this model. There are non-linear restraints. These springs are-- maybe
we have to design them. They might already be defined.
2. Let's open up this model and work from there, so we'll start up CAESAR II and open up the
model PEXP4.C2. That's the model we wish to run. Open that. Let's go into the input for that
model.

3. It's a bigger pipe, 8-inch pipe running from the anchor all the way to that point. There are no extra
nodes in here for dynamics. This is a straight, static model.
4. Show the node numbers on the plot, so we might have some long runs here. Well, there's an
extra node at 50. Maybe these longer runs should be made shorter for a better dynamic analysis,
but let's just take our static model and build up the dynamic input for it.

5. Let's first of all see if we have any non-linears in this model. I'm going to go to Edit, List,
Restraints. Here's my restraint list. Yeah, there's a plus Y. Is there any friction in this model? We
do have non-linear conditions here, lots of them.
6. Do we have to size the springs? I'm just going to click on this piece of pipe. And that should show
me my hanger-- yes, we're sizing springs, so there's two reasons we'll have to run a static
analysis. We have to have a static analysis to set the non-linear restraints. And we also have to
run a static analysis to select these two springs in the model.

7. We'll do the error check. Yes, we've got a size spring, so we have to do the static analysis.

Click the running man. This will give us our static solution.
8. Let's take a quick look at the restraints, click on Load Case Analyzed 3 and 4, then click on
Restraint Summary.

Looking at these plus Y supports, at node 125, the support is active in the operating case, but it's
not active in the installed case.
If I'm running a dynamic analysis, I have to decide whether I want that support in the model for
the operating case or not in the model at all using the sustained case with the installed position.
That's the only non-linear action in the model.
9. And like I said too, we have hangers that we sized. They had to be sized before I could run my
dynamic analysis. Let's take a quick look at those.
Unselect the Load Cases and Standard Reports and select Hanger Table Report and then
View Reports.
And yeah, I picked two springs, a lot of vertical motion. Anvil, figure 82 size 9 and size 7.

10. Now we're ready to go dynamics, so let's close out the static output processor, getting us
back to the main menu. And once again, we can dive right into dynamic analysis. In this
model, we already have some information here.
11. It's already been set for Modal. We already selected the operating case, but again, since I have
non-linear, I have to select the Operating Installed, we'll stick with the operating case (Load
Case 3). We're going to calculate all the natural frequencies out to 33 hertz. And we'll use the
lumped mass model.
12. Let's click the running man.
We have the error check. And on to the dynamic analysis.

We have many, many natural frequencies below 33 hertz, look at them again, in the List. Click
on Natural Frequencies report and then View Reports.
Now one note, usually you don't want to see these very low natural frequencies. They should be
explained or eliminated, but that's just an observation. We're just calculating these at this point.

13. Let's take a look at these in animation, because if we look at them from the data set by clicking
on the Modes Unity Normalized report, it's just too complicated to review. We've got a lot of
things going on here. We're going in the x direction and z direction both, and a little bit of y
component for the first mode. This is the first mode of vibration and here is the second mode of
vibration, so we have a lot of data to go through here.
14. We already have the frequency so, let's look at the shapes this way. We're going to look at the
animated mode shapes by clicking on Animation.
15. Here's the first mode of vibration, first mode right there, click on Volume Motion. Now,
remember, when we talk about the first mode of vibration-- when we talk about modes of
vibration, it's the simple equation square root of K, over M. If this is the first mode of vibration, it
must have a small K, and a big M, square root of K over M.

16. Now what components are actually bending in this model? Well, I see there's some twisting of
this run up here bending this run of pipe. That's a very long run. And this one over here seems to
be bending.
17. We have low stiffness, but we have a lot of mass participation in here. It looks like the entire
center of gravity is wagging back and forth, look from the top, kind of, in the x-z direction. Again,
these are the two elements that are actually deflecting or bending, so that's the stiffness term,
very low stiffness, but we get a lot of mass participation. That's the first of vibration.

18. Let's go back to the overall view. Let's look at the second mode of vibration. As we get the higher
frequencies, we have less mass participation and higher stiffness. Select Load Case 2 and click
on Volume Motion.
19. This one seems to be almost perpendicular to the first one. Looking down from the top again, the
first mode was going this direction. The second mode is going in this direction. Now, they don't
have to be perpendicular. There's enough other reasons to call them orthogonal, but we get
about the same amount of participation of mass. We're just bending different legs. And they're
stiffer legs than earlier.

20. Third mode of vibration, now it actually looks like a torsional mode. The mass itself is not really
moving off its axis, but it is kind of rotating about the y-axis. Again, using the longer legs, they're
more flexible.

21. Fourth mode of vibration, this is kind of interesting. If I look at that by the original view, it's actually
subsets. These stiff sections in here or these heavy sections in here are wagging away from each
other.
22. Now, you might say, you cannot get this mode of vibration. If these are pumps down here, they're
not going to be able to slide back and forth like that. But we modeled this as a frictionless surface,
so our model allows us to have it. So, you might want to discount this mode of vibration or
somehow adjust it, maybe put friction in the model or put in restraints in the x direction down here
so that doesn't happen.
23. As we get to higher and higher modes, we see different parts of the system responding, but,
again, think of it as, we have more stiffness acting in these higher modes. And we have less
mass participating in these higher modes. Each mode has its proportion of stiffness and mass, so
the fifth mode kind of breathing open and closed on that loop.

24. If we go to the last one at 34 hertz, volume plot on that. Now it looks like individual elements are
moving. But again, this is a system-wide calculation of a mode shape. The entire systems
working. It's not just a single elbow whacking back and forth. This looks dead over here, but that's
that mode of vibration and 34 hertz.
25. Now one thing you might use this result for is, if I have some driving frequency, let's say, at 10
hertz, just for a number, 10 Hertz, I'm fairly close to the 10th and 11th modes. What do they look
like? The 10th mode, I'll volume plot that.
26. If I run some equipment at 10 hertz, I might be able to see this system participation in the applied
load. I can see these elbows walking away from each other.

In the 11th mode, I might also activate that one. And I see this elbow wagging up and down.
27. This is how you might use this output right now, just as a simple yes or no. Do we want to change
that frequency so we get away from some driving frequency, like equipment vibration or
something like that?
28. Also we see here that we have a Y support right here at this node. And it's not a nice, clean
bending. Perhaps if we put more nodes in the system, we could define better shapes.

29. We have reasons to maybe review that static model and add extra nodes in order to get a better
mass distribution for our dynamic analysis. So, don't just take your static model and push the
button for dynamics. You might do that to get a rough idea, but if you want a good dynamic
analysis, one of the first things you will do is review the model and see if you have to add
additional nodes.

Video 6 – Pipe Lift Off and Friction (from Courses Files)
1. We have two more models we want to look at. The first one is looking at LIFTOFF. We have a
model on our machine called LIFTOFF.C2. Again, we have a plus Y with friction at both nodes 30
and 40.
Let's take a quick look at this Excel spreadsheet.
2. I have a spreadsheet to relate different ways of looking at this model. Once using the sustained
case to set the non-linears, a second time using the operating case to set the non-linears, and a
third case where we throw in friction in the analysis. So we want to look at it three times.
3. We'll do that quickly, and to go back to CAESAR II. So, from the main menu, I’ll click Open
and then select the model LIFTOFF, and go to that.

4. Look at the input file for that.
5. Another simple model-- now, note that the first element is just that little highlighted bit.

6. Click to add node numbers. Apparently, this model has been very well
distributed, so we have a lot of mass points in the model. It's a very fancy model in that sense.
7. I'll look right down the Z-axis. I have all the extra nodes in the model, and the focus is on the
two Y-restraints.
8. I'll look at the one below first.

9. If you closed Classic Piping Input, double click the element 26-30 to display the content of
the selected Y-restraint.
We see the first restraint. It is plus-Y, and we have friction (Mu). So now, we're talking about
friction and liftoff. So non-linear conditions.
10. We'll run this Static Analysis. Click Batch Run. We have to have
the Static Analysis in order to get to the dynamics because we
have non-linear conditions.

11. We see the operating case and the installed case. I want to look at those two load cases, and
let's look at the Restraint Summary Extended, and then view the report.
Select 1 (OPE) W+T1, hold CTRL and click 3 (SUS) W. Click Restraint Summary Extended,
and then click View Reports.
I selected Restraint Summary Extended so I can see how much the pipe might lift off a support.
We have not just the loads and the restraints, but we also have the position of the node. We see
the anchor at 10. That's at the bottom of the run. That's the anchor. Doesn't move at all.
12. We also see the other two supports.
One at 30. We have the plus-Y support. It lifts off in the operating case. It's in contact with the
support install.
One at Node 40 is in contact in both cases.

13. We have a couple different variables in play here. One, do we want to have node 30 active or
inactive? And two, do we want to consider the friction term in the report?
14. So, let's go on to dynamics. Close out the Output Processor.
15. From the main menu, click Dynamic Analysis.
16. Again, it's already been seeded with a modal analysis. Our important point is: “What load case
are we going to use?”
17. First, we're going to use the installed position (3 W (SUS)), where the pipe is in contact at both
supports. So both supports will be active, but no friction will be included.

Click the running man.
18. So now we're calculating the natural frequencies. 3, 10, 14, and 50.
19. We then see the results.
20. Let's look at those Natural Frequencies again. Click Natural Frequencies, and then click View
Report.
21. As we see in our Excel spreadsheet, with the sustained case, I've got four modes of vibration.
3.59, 14, they're out of sequence here. 14, and then 10, and then 50. This table back on the
spreadsheet that I'm referencing has the order change in order to show different modes of
vibration in the proper order. So let's just go on with these numbers in the Dynamic Output.

22. Let's take a look at these four modes of vibration.

Close out the Dynamic Output window, and then
click Animation in the CAESAR II Dynamic Output window.
23. Here is the first mode of vibration. We have two plots, the Center line Plot (Motion)
and volume plot (Volume Motion).
Click Volume Motion .

This is the pipe wagging back and forth in the Z direction. It is not influenced by those two
restraints. Whether the restraints are active or not, it's still going to vibrate back and forth in the Z
direction (First mode).
24. Second mode. Click the Load Case drop-down, and select 2 – 10.8408 Hz.
Click Volume Motion . Now, it's in the axial direction. Again, not playing too much with those
restraints. It's mainly moving only in the X direction.
25. Third mode. Click the Load Case drop-down, and select 3 – 14.4799 Hz.
Now, it's wagging back and forth, kind of like a second mode of simple cantilever bending.

26. Finally, the fourth mode of vibration. Click the Load Case drop-down, and select 4 – 50.9387
Hz.
We're not really getting influenced by the supports, but it's not moving up and down, because we
have the two supports that are active. That's the support configuration for the sustained case.
We're seeing no sensible Y deflection in the vertical-- in the Y direction of the horizontal run.
27. Let's now go back to our model again, and this time, I
want to go back to my Input, but to my Dynamic Input.
Click Back to Input.
You can go back to Input, static or dynamic, from the Dynamic Processor. We'll stick with
Dynamic Input. Click Dynamic Input, and then click OK.

28. This time we will now use the operating case to set our non-linears. Click Control Parameters.
29. Click on the Setting cell in line number 1, click on the drop arrow, and go back to the
operating case 1 W+T1(OPE). Case number 1, weight plus temperature 1.
This will then have the support configuration from the operating case. Node 30 with a plus-Y will
be removed from the analysis, so the pipe might be able to wag up and down a little bit. Again,
there's no friction in this dynamic analysis. Row 2 says stiffness factor equals 0.

Click the running man.
30. Calculate the natural frequencies, and now we get five natural frequencies instead of four.
31. Before we go on, I'm going to explain some of these values in our table spreadsheet. Again, we
already ran A, sustained case support configuration, Y-restraints both at 30 and 40, and those are
the four natural frequencies listed under A. You'll recall I'm putting Row A out of order, just to
simplify the review of it. Now, I should expect to see Row B frequencies, so let's take a look at
them.

32. Select Natural Frequencies, and then click show me that report.
We see 3.59, 10, 14, 28, and 50, and some are exactly the same as before. The fifth mode wasn't
affected, or the first mode, or the second mode. However, we get a new third mode in, and we get
close to the same in what I'm calling fourth mode, just for numbers.
33. Let's take a look at these for animations. I'll go to the Animation.
The first mode should be the same and looks like it is.

34. Second mode should also be the same. Well, close to it. 10 point something. That's the axial, or
the X run, motion.
35. Third mode at 14.4. Again, the elbow is wagging out of plane, and it looks like it's pinned as a
nodal point at the point indicated in the image below.
36. Now select the mode that shows up that was not in our data earlier (Load Case 4). Click
Volume motion.

37. We can see the support on this is active, and it's wagging up and down, but the other support's
not active, so it allows this mode of vibration to come out.
38. If I look right down the Z-axis, I used to have the restraint at the location indicated in the image.
That restraint's gone. This is a mode that was not available when the support at this point was
active.
39. The last mode of vibration-- let's go back to the original view-- and we see that's certainly not
affected by the support missing. You get the same number, 50.930.

40. Now, one more example. We're going to add friction to our analysis. Let’s will go back to the
Dynamic Input.
41. I'm just going to drop in the stiffness factor for friction under Control Parameters. Row 2 says 0.
I'm going to type in 1,000 and run the analysis for this one now. How does this change the
results? Again, we'll run the analysis.
42. So now I only get three modes of vibration.

Where'd the other ones go? Well, we added stiffness and that kind of changed the response
43. Click Natural Frequencies, and then click View Report.
There's the three that we have.
43. Let's take a look at those three. Click Natural Frequencies, and then click Animation.
44. Now, once again, we had the old first mode. The old first mode was the free end out at the end of
the pipe indicated below, wagging out of plane.

45. Well, now we have friction out at that end, or a spring preventing that wild swinging out at that
point. Click Volume Motion . That mode no longer appears, so we don't have that
mode (C) in the results. However, I do see the elbow wagging out of plane at 12 Hz. That used to
be the same mode at 14 Hz. So a similar shape, but a different frequency, because we have
different stiffness in our model.
46. The second mode that we show is at 28 Hz.
Click Volume Motion . That kind of matches the operating position where the friction is
not really playing a role, and it's pretty much the same value, and now we can see why I threw
this one around.

47. The third mode that we calculate is in the axial direction, but it's at a higher frequency because
we now have an axial stiffness out here modeling the friction at the end point.
Click Volume Motion .
So it's that elbow wagging in the X direction.
48. That’s now my third mode, and that's why I listed the values in my table out of order, 14 to 10, 14,
28, back to 10 and 50. They should be normally rising, but if I look at based on the C row model,
it is a higher stiffness, so I list that one as a third one, rather than the order that we see over in
Row A and B.

49. We can see the effect of adding friction to our analysis. We get higher frequencies in certain
directions, and we have different mode shapes associated with them as well. Let's look at one
final model in our next video.

Video 7 – Piping and Structure Interaction (from Course Files)
1. In this final modal analysis model, we're going to take a look at the interaction of piping and
structure. The interesting thing here-- well, a couple of things are interesting, I guess-- is that yes,
we do get the steel participating in this response. Also, we have to look at the connection
between the piping and the steel.
2. Let's open up CAESAR II. I believe you're going to have to modify the structural steel input for the
version of CAESAR II that you're running. So what we're going to do first is open up-- not the
piping job shown (LIFTOFF), I click Open.
3. So, I'm going to ask for the .str file, the structure file. Select the Files of type drop-down, and
then select Structural(*.c2s, *.str).

4. We then see a file called COLUMN.str. Click on that file, and then click Open to open that
up.
5. Now, go into the structural input processor. Click Structural Input. Now, your screen might look
a little different. I'll open up my screen a little more. There Structures sitting by itself on the menu,
Structural Input, and the model Column displayed at the title.
6. That opens up the Structural Input Processor, and all we want to do is quickly save it. It's a very
simple model, just one piece of steel, 8” by 24 lbs per ft, anchored at the bottom.

7. I'll just click Save.
The software prompts me to save changes, I'll click yes.
So, it's been saved, it's been error checked, and the execution files are now ready. We're going to
close that out, click OK.
8. We now have a file called COLUMN that we can use, and now let's go open up the piping file.
Click Open.
9. The file we are looking for is called PIPE&STRUCTURE. Click PIPE&STRUCTURE.C2, and we
can see our pipe and structure in the preview image. Click Open.

10. Go into the input, static input or basic model input. Click Piping Input.
11. Now, we see the interaction with the piping structure. It is supported in the y direction, so the pipe
cannot lift off the structure. It is also supported in the z direction so they are attached. It's
basically a U-bolt over the top I guess.
12. However, what about the axial direction, the x direction? If

I look inside the pipe… I’ll click on transparent objects.
…I see we do not have our support in the x direction relating the steel to the piping.

13. Let's go and run this through the static analysis. Click the error check (Start Run).
14. Now how did that steel get in the model? Just to review that quickly, that is through Includes
structural files. So, we see there's that Column. That's how we get the steel in with the piping.

15. It's been error-checked. We'll go out and run this through static analysis. I'll click the Batch Run.
I'm not sure if I have to, it doesn't look like a non-linear model, but just in case we'll have it
available.
16. I'll quickly look at the Operating, Sustained, and Restraints Summary. Hold CTRL, and then
click 1 (OPE) W+D1+T1+P1 for the operating case, and 2 (SUS) W+P1 for the sustained
case. Click the Restraint Summary, and then click View Reports. This will tell us whether it is
non-linear or not.
17. We see I've got an anchor at 10. I've got a gap on my Z-restraint. So that is a non-linear condition
at node 40. That's on the horizontal run.

We see my connection at node 90. That's the midspan.
Scroll to node 140, and we see the other guide at the other end.
Scroll to the bottom node 1000, and at node 1000 we see the anchor for the column.
18. Just to review the nodes that we're looking at, I'll use this plot, since it's so easy to get to. Click 1
(OPE) W+D1+T1+P1, to select only the operating case, and then click 3D Plot.

19. We see the gaps on the guides indicated below, and we see gaps on the other ones indicated
below too.
We had plus Ys in the model. So, there were nonlinears. Good thing we ran that. We also have
the node C node combination to connect the piping to the steel, as shown:

20. Let's go on to dynamics. I closed out the output processor completely. I'm back at the main menu,
dynamic analysis. Click Dynamic Analysis.
21. We have data out there right now, simple modal analysis. No change to the mass.
22. Click Snubbers. No Snubbers.

23. Click Control Parameters. We declare on row 1 that we want the operating case to set the
nonlinear. So if the gaps were open on those guides, there'll be no restraints in that z direction in
this analysis. Row 5 we see a simple lumped mass, with row 4 indicating a stop at 33.
24. Click the running man.
We get our input error check. That's fine.
CAESAR II goes on with the analysis, and we see the natural frequencies. We end up getting 15
of them up to 33 Hz.

25. Now I'm on my output processor. Now, let's take a look at the natural frequencies again. Click
Natural Frequencies, and then click View Reports.
It's just numbers in the Dynamic Output. It's hard to get any sense out of this data.

26. Again, we could look at the actual numbers for the mode shapes. Click Modes Unity
Normalized, and then click View Reports.
The numbers for this Dynamic Output of the Modes Unity Normalized also has not just the piping
node numbers, but also the structural steel node numbers as well. Again, hard to see what's
going on in this data as well.

27. So, let's go look at the animation. This is a very good way to get a
good feel for what's going on in our data. Click Animation.
28. All right first mode, click Volume Motion, the loop is wagging back and forth in the z direction.
We have all that mass up at the top that's working on this stiffness up from the vertical piping
points-- well, actually, the restraints back on the lower horizontal section of pipe I guess, and the
structure and the steel are working together. They are connected in the z direction.

29. Mode number two. Click the Load Case drop-down,
and then select 2 – 2.5272 Hz.
It's a little stiffer in response, and it's a little less mass participation.
Click Volume Motion. Oh, look at that. That is interesting. The pipe is moving, actually,
but he steel is not moving at all. Why? Because they are not connected in the x direction. They're
only connected in the y direction and z direction.

30. Third mode-- now, I'm going to say there is no
interaction between the pipe and steel. We're just
naturally pivoting about, since the metric system, that's a nodal point anyway on the piping, no
effect of the steel in the model.
31. Number four-- it looks like they are connected in that

direction. The top of the loop is wagging back and forth,
and the steel is going along with it.
32. Five, it looks like it's connected, but minor.

33. Six-- now, here's a mode of 6 Hz. If you just saw a list of
natural frequencies, you say the piping system has a
natural frequency of 6 Hz. Well, no. That's a natural frequency of the column in the weak
direction, not the strong direction.
34. Again, these are all orthogonal modes of vibration for this system of piping and steel.
Looking at load case Seven.

35. Eight-- they're just bouncing up and down.
We've got our supports active over to the left and right on the horizontal piping, but it's got
enough mass to wag up and down like that.
36. Nine, there they're interacting.

37. Ten-- now we're getting some fancier work on these

runs. Maybe we need more node points in here to get
a better shape of that deflection.
38. Eleven-- now they're working again.

39. We'll look down from the top.
40. Let's go back to number 10. See how these are both
going, same direction like that?
41. Back to 11, and we see they are opposite what 10 is.
42. So these shapes should be somewhat sensible, and you could always learn something about
your system by running natural frequency calculations on your piping systems.
43. Finally, the last one (Twelve).
Let's go back to the original view.
Now we're just wagging up and down at the top side.
Click Front.

44. So, we saw certain modes of this steel, it was playing a role in the response, and in other ones
the steel was doing its own thing. It had to deal with the connection between the steel and the
pipe.

Video 8 – Static Model Adequacy and Determining Acceptable Criteria for Modal
Analysis
1. We took a look at quite a wide variety of piping systems. Let's just wrap this up, then, with a few
PowerPoint screens. I mentioned several times that maybe we could make a better model. So,
let's ask that question-- is the static model sufficient?
2. Well, basically, the image below is what we're trying to model-- a continuous beam, mass all
along it.
3. In CAESAR II, we will automatically throw half the mass to the anchor and the other half out to
the end. We just have this one mass walking up and down. That's the entire model.
4. We see also that we can create a more realistic model if we put extra mass points along it.
5. The question is, how many do I have to put in my model? Well, you can see in the image I've got
a simple cantilever, and I have, first of all, two nodes. That's one node at either end of that run of
pipe.

6. We see the data for my pipe, in the PowerPoint—50’ of 4” pipe. If I run a lumped mass approach
or if I run a consistent mass.
7. We see when you run consistent mass, we can calculate more frequencies, and they are better
calculations of those frequencies.
8. When I go to a 10-node lumped mass-- I put 10 points along the run-- I am approaching that
same value I had with consistent mass. So consistent mass is always going to give you better
numbers.

9. It takes a little more time to do so, but you may have to add extra nodes anyway in order to get
the shapes you see below defined. If I don't have all these extra nodes along the run, I can't show
all these node shapes.
10. If I go to 10 nodes and I use consistent mass, the numbers are even more correct, I'll say.
11. If I go to the extreme-- this is that cantilever-- 100 nodes, we see probably the most accurate
number.

12. If I compare it to hand calculations—looking at this worksheet.
We see my pipe I'm running.
I went to Theory of Vibration with Applications, by Thomson, the book that I used in college.
We see the calculation or the formula for modes of vibration of simple cantilever bending. Plug in
these terms.

We see my frequencies-- 0.471, 2.95, and 8.26, and those are the values we see in the
PowerPoint calculations.
13. If I add all the extra 100 nodes, I get very good model. But I'm spending too much time creating
the CAESAR II input. This is a waste of time. If I go back to 10 nodes with consistent mass, I'm
pretty close. I'm going to say this is probably a better way to do it. I would always use consistent
mass over lumped mass. But even with 10 nodes, lumped mass and consistent mass are quite
similar.
14. If I just go with the original lumped mass model, I'm not correct for the first mode. I'm not that
close, and I don't even get the other modes of vibration.
15. This should give you some pause before you just push the button, using your static model. You
may want to add extra mass points in the system. But again, we still have the question, how
many do I have to add? Is 10 enough? Is 100 too many?

16. Well, here's a suggestion. There was a paper for seismic analysis of piping system from several
years ago-- a Pressure Vessels and Piping Conference paper on mass lumping technique for
seismic analysis of piping.
They said that the spacing between mass points in a model should be no greater than this term.
You could run through this equation. But reviewing this equation, I'll say the maximum spacing
between nodes should be no more than 1’ per nominal inch of pipe.
17. If I'm running 4” pipe, 4’. If I'm running 8” pipe, 8’ should be my maximum span. That's a good
general rule of thumb.
18. You should also put a node between restraints. If you go from one restraint to the next, put a mid-
span on those runs. Also if I'm going elbow to elbow, I should have at least one mass point
between a pair of elbows. Again, you can simply break the model in the CAESAR II static model
to quickly add those extra nodes. Elbow, elbows-- put a node between two bends, and also use
half that spacing when you're coming into an anchor. So, there you have it.
19. Now, one final point-- we did not load the system. All we did was calculate natural frequencies.
And that might be an end in itself. There are certain specs that say, the first mode of vibration
shall not be less than, let's say, 3 Hz. Where does that come from?
20. Well, one source is the DNV's document, recommended practice D-101. It says the first mode of
vibration should not be less than 4 to 5 Hz.
21. So almost every model that we've ran these sessions had very low natural frequencies. You
would probably want to remove those modes of vibration. You can do so by increasing the
stiffness of the system that will increase the natural frequency.

22. So basically, you will alter natural frequencies. You usually add a restraint or add stiffness to the
model. You've got to remember, though, that it's going to make the system a little more
expensive, and also, you might start to affect your thermal analysis.
23. Now, this modal evaluation will highlight modes of vibration that might be too low, that might pick
up environmental frequencies-- running speed of equipment or flow-induced vibration.
24. Also, the first bullet in the PowerPoint (and shown below for reference) says, a lot of your piping
systems are designed for vertical direction and for flexibility. You might see your systems in the
very low modes of vibration wagging in the horizontal plane, indicating that you might have to add
guides. This natural frequency check will tell you where you might want to add those guides to
prevent that very low natural response.
So this model evaluation is very useful in its own right, in order to qualify the system.
25. In closing, model evaluation that we've gone through so far will be used in the other types of
dynamic analysis-- the modal methods, the response spectrum method or the time history
method. Those are both modal methods. You can use them as a qualifying tool or use
frequencies as a screening tool, to make sure that you don't have too many low natural
frequencies.
26. We didn't get any real numbers, like stresses, or forces and moments, but these modal analyses
will be useful in the response spectrum analysis, and where we will get deflection, load, and
stress.
27. When we run those dynamic analyses, we'll also see that the program will tell you which mode of
vibration gives the greatest contribution to that output. So that's going to be a very good
diagnostic tool-- which mode of vibration you want to change.
28. One way that we can correct problems with response spectrum analysis or time history analysis
is to alter the modes that are causing the response. So again, we have the ground rules for that.
Adding stiffness or reducing mass will increase natural frequency. If you have a bad mode of
vibration, you want to move it. Usually, you add stiffness. It's not a common tool to change the
mass of a piping system to solve dynamics problems.
29. Modal analysis is a building block for other dynamic analyses. But it is also a good tool in its own
right to learn more about your piping system. Thank you.

HARMONIC ANALYSIS

LESSON 1 – HARMONIC OVERVIEW
Video 1 – Introduction to Harmonics
(Reference Video: C2_D_H_L1_V1)
1. In this group of videos, we will discuss the harmonic analysis capabilities in

CAESAR II. We will review the static model that we're going to run in dynamics.
Then we'll focus on the harmonic input, and then the steps through the
harmonic analysis itself, the interaction there, and then review the results. Of
course, there are always some adjustments you can make to your models for
harmonics, and we will discuss them as well before we close.
2. In our initial overview, cyclic loads, even a very small magnitude in input, can
create large response in the piping system and fail the piping system. So any
shaking of the piping system in the field is a cause for concern.
3. We can use CAESAR II to replicate the piping response that you see in the field
with those applied loads. Once the computer model matches the physical
system, the model can be confidently adjusted to detune the response.
4. Our point is that we have a forcing frequency, omega f, and a natural frequency, omega n. When
they coincide, that's when you get the large magnification of the applied load and get a large
dynamic response. Whether it's by misfortune or accident or lack of concern, these things just
pop up in the field.
5. So, what we have to do is evaluate the magnitude of the stress in the line because of this, and if it
is overly large, then we would have to fix the system. We usually do that by detuning the system
or a change in the natural frequency of the piping system. So, what we're going to do in this
presentation is review the input that's required for a harmonic analysis. Then, we’ll also bring in
the engineering concepts used for harmonic analysis.
6. Now, you might want to try to use harmonic analysis in initial design, but that is in many cases
fruitless. You'd have to guess at what the frequency of this applied load is and the magnitude of
the applied load. Usually what people do for new design is just calculate natural frequencies and
make sure that these natural frequencies of the piping system, or the proposed piping system,
don't match known excitation frequencies in the field, just making sure that they don't match.
7. So, it's more of a qualitative approach rather than a quantitative approach when you calculate just
natural frequencies. We are not calculating stress, we're just calculating frequencies.
8. Now, when you have systems in the field, they do vibrate. Usually, the way we see it with
CAESAR II is that they're caused by mechanical vibration, fluid flow-induced vibration, or
pulsation. We'll talk about these in more detail, but they have separate mechanisms, what causes
the response.

9. In a mechanical vibration, there's some kind of displacement, cyclic displacement of a boundary
condition, like a pump nozzle wagging back and forth due to imbalance in the pump.
10. On flow-induced vibration, there is flow, and the cyclic load on the piping system is perpendicular
to that direction of flow.
11. Pulsation is usually around reciprocating compressors or systems. There is a force in the
direction of flow, as opposed to perpendicular to the flow. Again, we'll talk about these in a minute
or two.
12. So, this is not so much design engineering, because we actually have a problem in the field. So
we're going to call it forensic engineering. We're trying to decipher what's going on out in the field,
and then match that with our computer model. If we make some changes in the computer model
that detunes the system, then we have high confidence that it will also work in the field.
13. Now, you might say forensic is late in the game for doing your engineering work, but the
advantage is that you have the real system in the field, and that will be our benchmark. You will
know you have a good model in CAESAR II when it's performing like the real system out in the
field.
14. Now again, what does that say about harmonics for new piping design? Well, like I said earlier,
you just have to guess at the magnitude and frequency of the applied load. You'd never run out of
guesses.
15. So in new design systems, it's usually more sensible just to calculate natural frequencies. Make
sure those natural frequencies do not occur near an expected forcing frequency, or at least above
some minimum value-- for example, 5 hertz.
16. If we would look at the math behind this, the big thing is the ratio of the forcing
frequency to the natural frequency.

17. If we look at the equation from Theory of

Vibrations by Thompson, we see that ratio,
omega f (ω⨍) to omega n (ωn).
When the highlighted section of the

equation goes to 1, we really take off the
amplification.
Amplification is a function of damping , forcing frequency , and natural frequency ..
18. Now what confidence do we have in these numbers?
Well, we might get the omega f ( ) pretty close. So that's a forcing frequency. If it's a
mechanical vibration, it's the spinning shaft in the compressor or the pump or the reciprocating
compressor frequency. That, you can get pretty close.

However, the natural frequencies ( ) in CAESAR II are only as good as your data set. A lot of
times, the stiffnesses that we assign to piping systems, restraints and the like, are not that
accurate. They don't have to be that accurate for static analysis.
19. When we do not have a very accurate stiffness matrix for our piping
system, we don't have the high confidence in our natural
frequencies.
So, we do the best we can for that omega n ( ), the natural frequency, and try to match what
we're seeing out in the field.
20. The point, though, is that we want to work on the ratio, omega f ( ) to omega
n ( ). Since I have low confidence in the actual stiffness here, my omega n
that the program would calculate may not be exactly right.
If I have a very firm set forcing frequency, I will still sweep through a variety of forcing frequencies
to account for this lack of certainty in the natural frequency. We will talk more about that late; how
we sweep through a range of forcing frequencies.
21. If we look at the amplification factor, if I have 1% critical damping, (damping is my first term in the
equation and piping systems have low damping), I can get a 50 times multiplier on my response.
So, a lot of dynamic load factor there.

22. If it's 3% critical damping, that's about 16 times amplification when I'm running right at the natural
frequency.
23. If I use CAESAR II, and if I can match the observed response in the field with what the program is
giving me, then I have good confidence in the bending stresses in the piping system, and

therefore stress in general. So, that's what we're going to do. We're going to try to match what we
see in the field and use that to our advantage and gain confidence in the stresses that we
calculate.
24. Now, one very important point, and this is a common trap: “Do not enter observed results as the
input load. Find out what the cause of the load is and develop that in your load vector.”
If it is mechanical vibration, the load comes through a restraint.
If it is flow-induced vibration or pulsation, you have to put the loads in the right directions at the
right points.
25. Below is an example of a piping system. We'll run with this later.
26. If you see a very large deflection at that elbow, do not grab that elbow and pull it back and forth.
That elbow is an observed response.

27. In our example, we're going to put some harmonic displacement at the connection point down at
the end indicated in the image below. We're going to try to make the machine develop the back
and forth response. So, keep that in mind when you're developing the loads.
28. In our next video, we'll talk about the sources of these types of loads in our piping system.

Video 2 – Harmonic Load Sources and Effects on Pipings
1. What is the source of these harmonic loads in our piping system? We spoke of three major
sources, and we'll discuss each one now.
2. First of all, mechanical vibration. Load is transmitted into the system as a displacement through a
support. So, it has to come in through a boundary, because you don't just grab some elbow and
start yanking it back and forth, trying to match some observed response out in the field. So you're
going to apply the displacement at a support point in the same direction as the support. If it's in
the y direction, you'd better have a y support there.
3. Typical example of mechanical vibration would be out of balance equipment. Well, really, if it's out
of balance equipment, it is a mass that is rotating out of balance and it's causing the casing of the
equipment to maybe wag back and forth left and right, and that is the load into the piping system,
that deflection.
4. I also mentioned flow induced vibration, but in the PowerPoint I'm saying the flow induced
vibration activates the structure, and the structure then moves the pipe. So again, this is just a
harmonic load cycling through. I'm showing you a cosine wave, but it's showing this deflection as
a function of time. This is what we're trying to model in CAESAR II.
5. In flow induced vibration, basically it is steady flow through or around an obstruction. This could
either be inside the pipe or outside the pipe. If it's flow inside the pipe, let's say a liquid line, the
frequency of this vibration is 0.2 to 0.3 times the velocity, fluid velocity, divided by the diameter of
the constriction through an orifice plate or something like that.
6. If it's outside the pipe, it could be vortex shedding, again about 0.2 velocity divided by the
diameter. The wind (V) blowing on the piping system could start shaking the pipe up and down.

7. Again, the load is going to be a force, not a displacement, and it'll be perpendicular to the
direction of flow. Now, it's rather difficult to guess what the magnitude of that force will be. Again,
you can build a model, and then tune up the applied load to match the field deflection.
8. Some examples of flow induced vibration:
• Here's a flow through an orifice where the vibration is perpendicular to the direction of flow
• Or even through a reducer. If it's a very short reducer, that velocity through the reducer could
cause some vibration, either up and down or in and out of the plane of this view.
• If we're talking about flow around the pipe, we could talk about wind load on the pipe, and
that will then start to-- we'll call it vortex shedding-- start bouncing the pipe up and down.
These are the forces, these red numbers (arrows) would be the applied force to the piping
system.
9. The third case was pressure pulsation, which is caused by reciprocating equipment. It's a
differential pressure along the run of pipe. So, since this differential pressure is along the run of
pipe, it is going to be a load applied in CAESAR II parallel to the direction of the flow. Again, a
typical example would be reciprocating compressor.

10. You see that compressor, were pushing the piston shown below, and if I could take a snapshot of
the pressure in the line, where there are big fat bars as seen below, its high pressure, low
pressure. This is just the changing pressure because of the pulsation at the piston. It is either by
dumb luck or poor design; you can get a situation where as the pressure wave is moving through
the system at the speed of sound, we might have a high pressure on one elbow, and a low
pressure on the other elbow as shown below.
11. It's a function of the speed of the equipment, the speed of sound in the fluid, and the length of the
run between the two elbows. This could happen on any elbow, elbow pair in the system.
12. What we want to do is we want to develop the load that will be in the line of the run. I'm
examining just the run pipe shown below.
13. If I calculate the pressure on the high elbow, looking inside the pipe, and I look downstream,
whatever surface I see, that's the surface that has the pressure on it. We see we have high
pressure times the area, high pressure gauge.

If I look the other way, I have low pressure.
This run of pipe is now out of balance and it's being yanked back and forth by
this delta P, p high, minus p low, times the area of the pipe. That will then pull
the pipe back and forth in the axial direction.
14. Let's go through a few examples:
Here's a piping system we use in our seminars for harmonic analysis. The script is that we have
an elbow at 20, that when the system goes into operation, we have a large deflection in the z
direction on the elbow, very, very large.

While that 20 elbow is wagging back and forth in the z direction, the elbow at 60, indicated below,
is wagging back and forth in the x direction.

While the elbow at 35 indicated below, is not moving at all.

15. Now one thing that we just described, since we see a large response, is that we're probably
describing a mode of vibration of the piping system. So it's important to collect all these little
concepts for a piping system, what's moving, what direction it's moving in and that will be used to
confirm that our model is accurate enough for this evaluation.
16. What is the cause of the load here? Well, I don't know. Maybe it's a mechanical vibration. It could
be that the pump itself, the casing, is wagging back and forth in the z direction, causing the elbow
at 20 to wag back and forth in the z direction. We know if the natural frequency, enforcing
frequency, are on top of each other, we can get a very large amplification of a small deflection
down at the bottom near that anchor. We saw that amplification factor of 50 for a system with 1%
critical damping. So that's a candidate valid.
17. Well, what about flow induced vibration? Well, there's a control valve as shown below. That
control valve might not be fully open, and that constriction could cause some vibration in a
direction perpendicular. Since 60 was wagging back and forth in the x direction, maybe it goes in

the direction shown by the red arrow. Usually, when you're working around control valves, its high
frequency vibration, it's more of a hazard for your hearing, more so than hazard for the safety of
this piping system. So, I would discount that one in this situation.
18. Third, is it pulsation? Well, we do have a reciprocating pump at the bottom as shown. So maybe
we might have a wave that is standing or a displacement that's standing in each leg. I would have
to check each leg. Now, if the leg indicated below is wagging back and forth, I would guess this
might be the right leg for it. However, we are told the 35 elbow doesn't move, and if I push that

leg back and forth in the x direction, I'm not expecting a response in the z direction at the 20
elbow.
19. I'll check each leg. If I loaded in the leg indicated below for imbalance, it will be hopping up and
down. If I applied load to the leg horizontal leg shown below, it should be moving in the x
direction. If I put it on the vertical leg with the control station, it should be hopping up and down. I
really can't find a leg that really allows me to get this pulsation in to this model.

20. I'm going to say for this model, it's probably going to be caused by mechanical vibration of that
connection in node 5 (you can see node 5 in the image above). So perhaps the most difficult part
of the analysis is trying to determine what the cause of this harmonic response and your piping
system is. In the next video, we'll start working with CAESAR II.
Video 3 – Preparation of INLET PIPING Model for Harmonic Analysis (from Course
Files)

1. Let's take a look at the model that we're going to be using for the demonstration of the harmonic
load capability in CAESAR II. You might have seen this model before in one of the previous
sessions. It's a Turbin model, inlet piping and exhaust piping, and we're going to be looking at just
the inlet side. So, if we have an issue with, let's say, vibration of the inlet system out in the field,
we would want to model this in CAESAR II, and then either do natural frequency calculations or
actually apply some sort of harmonic load. So, since we're doing a harmonic exercise, we'll apply
a harmonic load.
2. What we're seeing is that the elbow at 1050 is wagging back and forth in basically the z direction,
maybe the xz direction, sufficiently to cause concern. So, the idea is to somehow model the
applied load, harmonic load in the system, whether it's pulsation or flow induced vibration or
mechanical vibration. Then, try to replicate that deflected shape of the entire system to evaluate
the stress or bending moment and stresses related to those harmonic loads, and then evaluate
them.
3. Now, the elbow at 1050 is wagging whenever the piece of equipment is running. So, there's a
good indication that the forcing frequency will be a function of this, 330 revolutions per minute.
So, what we will do is we will, hypothetically, establish some harmonic load running at the 330

frequency and apply it at the appropriate point in the model. Then, we see if we can reproduce
the large deflection.
4. What we're going to do is we're going to assume that it's a mechanical vibration. In other words,
the nozzle indicated below is attached to the equipment, is the cause of the large displacement
up at 1050.
5. Here's our model. The model is quite simple.

6. We have the header line and then the branch, and it's a branch that's doing the vibration.
7. Now, what we're going to have to do, though, when we start talking about dynamic analysis -- and
we talked about this in the modal analysis-- is we have to perhaps better distribute the mass to
the system in order to capture the dynamic shape of this piping system. So, what we're going to
do is we're going to add these extra mass points, these red points, and you can see them in the
model already. The original model for static analysis had no reason to put these intermediate
nodes in the piping system. They weren't necessary for input or output. Now that we want to get a
better mass distribution, we have criteria on spacing between these mass points. We're going to

quickly add these breakpoints to our static models. So again, we're going to first update the static
model for the dynamic analysis.
8. However, first, let's take care of some housekeeping. Now, we're going to be using that model of
the inlet piping in our harmonic analysis, and it's available for you from the online video training
website. So, before you continue, I'd like you to go to that web site and under the Workbook
and Support Files tab, download the Dynamics Files.
In the Dynamics Files folder are several files that we will use for harmonics. We will move three of
them to our machine:
One of them is either Inlet Original, that's the CAESAR II model we're going to start with, or Inlet
Original 2011. The 2011 version is version 5.31 CAESAR II. The Inlet Original.C2, is the current
version of CAESAR II 2017.
9. If you're using 2017, go ahead and download or capture Inlet Original.C2. If you're using an
older version of CAESAR II, go for Inlet Original.
The two other files I want you to download are:
• The units file. You might have this from some of the other video series files, but I have it
again called Euro file, Euro.fil. That goes into the folder where you're running the models.

• Another file called Appx W.fat. That's a fatigue file. We didn't talk about it yet, but we'll use it
later on.
10. I'm going to ask you to place the CAESAR II file and the .FIL file in the working folder, and we're
going to move the .FAT file to the program System folder. I'll do that all in the next step, we'll take
care of that.
11. Then, we will start working on that Inlet Original file that you download, and that's what we're
going to be using for our harmonic analysis. So, go ahead and start up CAESAR II once you
download the files.
12. Now, I have them downloaded on my

machine. I have a folder called Dynamics
download, it's sitting on the desktop, and
you can see the four files that we are
talking about today.
There might be other files in this folder or

other segments of the dynamics
sessions, but these files are what we're
focusing on for this lesson.
13. I have those files on my machine, now, I'm going to start up CAESAR II. I'm running the current
version, 2017. By default, the program sends you to the Examples folder. What we're going to do
is we're going to create our own folder on a drive of this machine. So, this has nothing to do with
the Downloads folder.
14. I'm going to click File > Open.

15. Again, I'm pointing to the Examples folder. I'm going to just build up a folder to work with. I'll go
back to the PC route, click This PC, and I've got the folder drive H. I'm going to open up
that H:\ drive folder.
16. I'm in H drive now, and I'm going to click Create New Folder.
17. I'm going to call this Dynamics_Training.
18. I can click Open for that new folder, and now we will go to those files that we pulled down from
the web site. I stored my files, just the harmonics files, on a folder called Dynamics Download on
the desktop. So, what I'm going to do is I'm going to open up that folder and I'm going to drag in
the Inlet Original without the year on it. I will be working with this current version of CAESAR II.
Just drag it and drop it in Dynamics_Training. I'll do the same thing with the other file.
Now, if you are not using the current version of CAESAR II, bring in the 2011 version. That's an
older version of CAESAR II and that will run for you.

19. The other item I'm going to drop in, I'm kind of being sneaky on this, I'm going to drag and drop
the Euro.fil in there as well. It won't show up in the Dynamics_Training folder because it is not a
C2 file. However, it is putting that unit's file into the folder.
20. We have our units file and our data file in the folder, and we're going to have to modify this
drawing a little bit. We'll start that up once we open up the CAESAR II file.
21. Now, there's also that other file we have to bring in as well, the Appx W file, the fatigue file. That
does not go into the data file set. It has to go into the System folder. I'm just going to click
Cancel to close this out now. We already got our file.
22. I'm going to go to Utilities from the C2 Main Menu and click Open System Folder.

23. This is an easy way to get to all that data that's in the System folder. Again, from these other
videos you'll recall that this system folder carries a lot of data for the program, amongst other
things, flange limits, and input formats.
24. I already have my Appx W file in the folder.

I have to pull it in again to show you. There
are .FAT files in the folder, and these are
the fatigue files that the program
references. The ones we ship with the
software-- here's one for TD12-- are
associated with the ASME section 8
division 2 fatigue curves. We're bringing in
our own. I will describe how this was
created later.
25. Right now what I want to do is go back to that Dynamics Downloads folder and just click
on Appx W and drag it and drop it in the System folder.
26. Now, you might have seen that I had one already. I'm going to replace the file and destination. So
there, I now have the FAT file in the System folder. So, we've taken care of the data transfer from
the web site down to your local machine. We're done with the Dynamics Download folder, and

we're done with the System folder so close both.
27. I can go back to my home screen, and I'm going to click Open. We are not in the Examples
folder, we have our Dynamics_Training folder. We're going to click This PC, and I had my
folder built on the H drive, click H:\. I'll just scroll down to Dynamics_Training folder, and
there our file the Inlet Original.
28. I'll click Open for that file. Click Piping Input. Open it, and we'll go to the input processor.
29. In the next video we're going to work with this new model and call it Inlet Piping.

Video 4 – Reviewing Static Input Model for Harmonic Analysis
1. So, you’re now in the input processer, and look where I'm pointing:
H:\Dynamics_Training\Inlet Original.
2. Now, before we continue, I want to immediately click File > Save As.
3. I'm going to call the drawing inlet piping. We're going to work with this new model, inlet piping.
I do that, so you leave the original out there.
4. Now we're going to make some modifications to this model. We're going to clean it up a little bit
and add those extra nodes that I mentioned earlier.
5. Let's review what we're going to do with this model:

• We already saved it.

• We're going to delete that rigid element at the very end of the model, from 291 to 10
• Pull out a c node.
• Add our dynamic nodes.
• We’ll error check and analyze the model for static analysis.
Let's go back to our C2 model.
6. The first thing I want to do is delete the element at the end of the thin run of piping, as shown
below:
It's just, in my opinion, in the way. That element is

modeling the body of the equipment, and we're going to
wiggle the nozzle. The nozzles at node 10, 90. So I don't
need that rigid element at the bottom. Click on it, and
there it is at 1091 to 10, that's the very last element.
If you can't find that easily, just click Last Element , and it goes to the same element.

We want to delete the element, so click the trash can, Delete
Element .
Click Yes, I'm deleting it. We deleted that element.
7. Now, the second thing I said is on the anchor, it is golden and that's our color for CNODE
restraints.
8. So, I expand the Classic Piping Input screen, and we see there it is, 1090, that's my restraint. It
says Restraint, and we see the anchor.

9. Remove 1091 from the Node box. That was part of the element that
we deleted.
10. Now, we just have to add those other nodes into the model. Click on the first element as seen
below. That's the whole element and actually the element is 3000 mm long. For a 12” pipe, 3000
mm might be a good spacing, but I just want to bring home the point again, that you may have to
add more nodes to make a better model.
11. So, I've highlighted the first element, right-click, and select Elements > Break Element.
12. We're just going to put a mid-point in this system. The way the processor works now, is I can say
Insert Single Node, and it automatically comes up with a mid-point node. The whole element is
3000 mm long and it's putting node 1015 between 1010 and 1020 at the mid-point. I'll just click
OK.

Now, it's broken.
13. Let's put our node numbers on this plot, so we can see them.
There's 1010, 1015 1020.

14. Click on the next element.
Same thing. Right-click on the pipe, select Element > Break Element.
Select Insert Single Node. Click OK.

There's 1025.
15. Now, if you're running an older version of the program, you would do it a different way. Click on
the piece of pipe.

Then, use the Break command on the left side tool bar.
16. If you use an older version, you can select Insert Multiple Nodes, and type the Total Number
of Break Elements. I'm going to type 2 to say I want two elements out of this, and it'll
automatically break it.
17. Then, it puts 1031 in the model. Its 1020, 1040, you figure 1030, but I'm already using 1030 at
the end of the larger element. So, it puts 1031 on the line.

18. So depending on what version of the program you use, you can either use the more up to date
Element > Break, click OK.

Or you can use the Break toolbar button.
Or even from the menu bar, the break command is under Model. Control-B would do the same
thing.

We're going to go along with our current method.
19. I'm just going to go down to the last pipe element. Now this small pipe is 4” pipe. So in my
estimation, we shouldn't go more than a meter, maybe 1200 mm between node points. We talked
about that in the modal analysis section. Basically, max spacing ballpark, don't go more than 300
mm per nominal inch, or don't go more than one foot per nominal inch of pipe. So, we're kind of
overkill in the model, but again driving home the point that you should review the model for mass
distribution. If you're going to go into dynamics.

20. So, the extra nodes are in the model, and we're OK with the model.

21. One other thing. You might recall when we looked at the original model, the anchor at the bottom
was golden colored, and now it's red. The set up in CAESAR II is that if you have no CNODE on
your anchor, it appears red. If you have a CNODE specified, in other word is a relative anchor
rather than absolute anchor, it'll be that golden color. So now we see the red anchor, so that is
grounded.
22. That's where we're going to apply our harmonic input. So this will be the model that we're going to
use for dynamic analysis for harmonic analysis.
23. One other point, this model does have a non-linear restraint, which is a plus y
restraint in this model. That means the program will have to run the static
analysis in order to find out the nature of this support configuration. It's
probably going to rest in every load case, resting on that support.
24. Dynamics are always linear. Harmonics is part of the dynamics family. So, the program will put a
double acting y restraint there for the dynamics model, if it is resting in the load case that we
choose to use to set the configuration of all of our supports. There are a lot of words in that
sentence, but we'll see how that plays out once we start developing our input.
25. So that's our model. Let's save it, File > Save, and now we're ready to move over to the dynamic
input processor.

LESSON 2 – HARMONIC INPUT
Video 1 – Dynamic Input Processor, Excitation Frequencies
1. Now let's discuss the input for CAESAR II to develop this Dynamic Analysis this Harmonic
Analysis in CAESAR II.
2. We're going to start the Dynamic Input Processor, just click Dynamic Analysis as seen below.
That activates the current job. Now, this drawing was Inlet Piping, ours is Inlet Original. Might
want to change your name if you wish, but we'll work with this one.
3. So, Dynamic Analysis opens up the entire set of tabs that are available in the Dynamics
Processor.
4. So before you start getting concerned about all the tabs in the Dynamic Analysis, go ahead and
select your Analysis Type. If you click on the Analysis type drop-down arrow, you'll get a list
and you can then select Harmonic.

That then reduces the tabs to those that are only related to Harmonic Analysis. So, when we click
on that, Harmonic, we will get the Harmonic list of tabs. Now, some of these tabs you have seen
before in the Modal Analysis videos. They had Lumped Masses, Snubbers, and Control
Parameters. We have them listed as well; the Control Parameters will be a little bit different.
However, we also see three tabs that we didn't see before-- Excitation Frequencies, Harmonic
Forces, and Harmonic Displacements.
5. So again, the forcing frequencies, omega f, would be defined under Excitation Frequencies.
6. Then, the vectors that are loading the system, the point of application, direction, and magnitude
are defined either as a Forced Set, or several forces, or several displacements, one or more
displacements.
7. Now, in our model we're running a mechanical vibration, so we're going to be using Excitation
Frequencies and Harmonic Displacements.
8. If we were doing pulsation, or fluid flow induced vibration, would be specifying Harmonic Forces.

9. So what data do we have to put in those tabs? How is the Dynamic Analysis screen set up?
Let's first talk about what we see on the Dynamic Analysis screen. What we have indicated in the
image below are three buttons to Save, Error Check, and Run the Analysis. We have no data yet,
so we're not going to push those buttons.
10. If you wish to add lines in our data set, underneath these tabs, we hit the plus sign.
11. If you wish to delete a selected line, hit the negative sign.
The other three buttons at the end of the toolbar are not used for harmonics, so we won't discuss
them in this lesson.
12. Now, we can add or delete lines, so what's a line of data?
Well, the default data shown in the image below, shows we have two lines of data under
Excitation Frequencies. Now, they both have a check mark in the first column, so that means
they're comments and they will not be processed.

So, we are going to see these types of examples for several of the tabs to give you an idea of
what type of data we want to collect.
13. Now, we also see that there are several columns in the data sets for the examples that are shown
delimited by the spaces.
So, this one says 1.1, that's the Starting Frequency.
3.1 is the Ending Frequency.
1.0 is the Increment.
We're going to be using these columns directly, so we won't be referencing that format statement.
14. All right, what is our data? Well, what we want to do is first of all, set up our Excitation
Frequencies. We're going to be running at 330 RPM. That is 5.5 cycles per second. Just 330
divided by 60.
Now, what we want to do is, focus on the ratio omega f ( ) to omega n ( ).

Now, we can say exactly that our forcing frequency is 5 ½ Hz.

However, we are not that certain about our natural frequency. Not that we have concerns about
how CAESAR II's calculating it, but instead: How are we modeling the system? Do we have the
right stiffness of the pipe itself or the supports or the boundary conditions in general?
15. Since omega n is kind of questionable between our models in the real world, we're going to let
the program calculate for us, well basically calculate that system frequencies based on stiffness
and mass. We're going to tweak the forcing frequency ( ) to account for uncertainties in that
natural frequency ( ).
16. So, we know omega f ( ), but we're not sure about omega n ( ). So we're going to sweep
through the range of forcing frequencies. I'll say plus or minus 10%, to count for that uncertainty.
Again, our focus is on the ratio omega f ( ) to omega n ( ), not just a natural frequency ( ),
not just the forcing frequency ( ), but the important thing is the relationship between them.
17. So, instead of me running just 5 ½ Hz, I'm going to sweep from 5-6 Hz. Hopefully, we don't want
to see any resonance, but for the analysis to be valid we should get some peak somewhere in
that 5-6 Hz that then falls off because we are getting close to that ratio of one.
18. So we will enter data as we see on line 2. We open up a

new line, click the plus sign, and it opens up this row
number two.
It will not be check marked, and I'm gonna say starting at 5 Hz and stopping at 6 Hz. Also, give
me a frequency at every 0.05 cycles per second. So this will give us 21 different analyzes. Well,
at least 21, depending on what I specify on the force sets and the displacement sets, but this will
allow me to run several analyzes in a range. We will not specify anything under Load Cycles at
this point, we'll come back later.
19. Note to that on the titles of these columns it says Hz, Ending Frequency is Hz too, but it says
optional, O-P-T. The Increment and Load Cycles are optional.
20. If I just specify a Starting Frequency, then that is the only frequency that will be analyzed. For
example, I can just type in 5.5 for the Starting Frequency on line 2 and I get analysis at 5.5 Hz.
Again, I want to sweep through the 5-6 range to tune in the maximum response. So that number
2 row is my one line of data for the Excitation Frequencies and I'll get 21 sets of analyzes of any
vector that I specify under forces and displacements. We will pick up with the remaining sets of
data in the next video.


Video 2 – Harmonic Forces, Harmonic Displacement
1. Continuing on, after we define our Excitation Frequencies, we'll develop either Forces or
Harmonic Displacements. Now, we're going to be analyzing a mechanical vibration through the
anchor in this model, the nozzle connection. So that's going to be a displacement.
2. However, first a brief word on the harmonic forces. Again, we have several columns that we have
to fill out. For Forces, we specify the magnitude of the Force (I'm in metric system, so it's N
(Newton)), the Direction, and the node, Start Node or Stop Node.
3. Let's just talk about Force, Direction, and Node which is putting a vector on the model, a certain
magnitude, direction, and location. The other thing I have to specify, which I've been skipping
over, is a Phase Angle. We'll talk about it, certainly, with the harmonic displacements.
4. If we have more than one force in our model, they have to be related to one another. They are
either in phase or out of phase. So, it is the Phase Angle that we can use to establish the
relationship between a variety of forces that might be on the system at the same time.
5. Again, for the Forces, this is basically for flow-induced vibration or pulsation. One problem with
these forces, these magnitudes are hard to predict. Well, even with harmonic displacements they
are hard to predict. So in many cases you will guess at a load and then tune it up in order to get
the deflected shape that you wish from your output. So again, Force, Direction, Phase Angle,
Start Node.
6. The last two Stop Node and Increment are optional. If I wanted to specify the same vector at
several points in the system, I can use Start Node, Stop Node and Increment.
In the same sense I could use-- back in the Excitation Frequency screen, as seen below, we saw
it, Ending Frequency and Increment. So Starting, End Frequency, Increment, so in Harmonic
Forces, too, we have the same thing as shown above.

7. If I use the proper node numbers, instead of me putting several lines of data, one for each vector,
if I give them the proper node numbers, I can string them all in a group, first node, last node, and
increment to get from the first node to the last node.
8. Now, we're going to be using Harmonic Displacements. What we're going to do is we have this
node 1090 at the end of our system, and that's where we're going to apply our displacement.
9. So, I'm putting this 0.1 mm displacement, rather arbitrary, and we'll see how correct that is later or
if we get our results that we wished. It'll be in the X direction. The X direction is perpendicular to
our lateral with respect to that run of pipe.
That then might get the elbow indicated in the image below wagging back and forth, and that's
what we're after.
So, a 0.1 mm in the X direction at node 1090, and the Phase Angle, I'm going to specify it as 0.

10. Now, you might say, well, are we doing a sine wave? Well, the program actually uses a cosine
pattern anyway. Phase Angle 0 is when we're at the maximum of the applied load as seen. So,
Phase Angle 0 will give me the type of load you see below. So we're evaluating that load at 1090
in the X direction, a 0.1 mm in X. It goes to 0, then minus a 0.1 mm through 0, then a positive. It
does that at the frequency of five to six cycles per second, stopping every 0.05 Hz. That's the
displacements that we specify.
11. Now, what about the Lumped Masses and Snubbers? Now, in regard to that magnitude, again,
I'm putting a 0.1 mm. If you actually have data from the field available, you can go ahead and use
that data, but I would not expect it to give us the results that we want. Why? Well, again, we don't
have the same model stiffness and model mass that will match the field.
12. So, since we're going to be calculating a natural frequency based on

your input model, that ratio Omega F to Omega N might be off, and you
will not see your exact data give you the exact response.
13. Now, there are three components in our analysis that will affect the response that you wish to
generate:
• One is the magnitude that you specify, either a force or, in our case, a displacement.
• The forcing frequency can also affect the magnification or amplification. The closer this
forcing frequency gets to a natural frequency, the greater the response we're going to get
from the system. If we run the analysis and we don't get the magnitude we want, not only
can we change the applied load, we can also change the forcing frequency, get it closer
to a natural frequency. That might get us to the response that we want.
• Thirdly, another wildcard is the system damping. Piping damping depends on how you
construct the system. Welded systems might have 1% to 3% critical damping. If you're
sliding on supports, that will add more damping. If you get any yielding in the piping
system-- and I hope you don't have that with harmonic loads, but if you have any yielding,
that will also increase damping.
14. We can modify the input value for system damping to modify or adjust the magnitude of the
results from the program.
15. All three of these terms play a role. Again, since our goal is to replicate the shape of the harmonic
response, it doesn't really matter how I go about doing that unless I get too crazy with some of

these values that I use. ry to keep them realistic but consider them as another way that you can
modify them. The next video will discuss the other tabs in the Harmonic Input screen.

Video 3 – Control Parameters
1. In our harmonic input so far, we've defined our Excitation Frequencies, we have not specified
Forces, we have our Harmonic Displacements. Then, the other three tabs. Lumped Masses and
Snubbers, we discussed in the previous group of videos on modal analysis. I have nothing to add
about these now. If you wanted to add more mass or add snubbers, this is how you would do
that. Snubbers were those restraints that only work in a dynamic analysis. Finally, Control
Parameters which we'll talk about and that will finish our input for the Harmonic Analysis.
2. Again, we already discussed some of the data with the modal analysis.
3. We don't have any non-linear restraints to set for the dynamic analysis. That's the first line, the
Static Load Case for Non-linear Restraints Status.
Basically, CAESAR II will only run linear models in dynamics. So, if we have non-linear restraints
in the static model, what setting 1 W=T1+P1=H(OPE) does is it says-- well, however that restraint
is acting in the load case, that's how it's going to be acting in the dynamic analysis. So, if we have
a resting support that is in contact with the support, it is a double acting Y restraint or double
acting vertical restraint in your model. If it is lifting off of that support in the operating case, it will
be not restrained at all in the vertical direction.
4. Stiffness Factor for Friction. Well, we don't have any friction in our model and it doesn't get used
anyway. We put in a stiffness to mimic that effect of friction in your dynamic model.
5. Damping is new with this setting. We didn't see damping in the modal analysis, and this is a 1%
critical damping. Put it in as a decimal value. 1% is what we're going to do.

6. This is kind of a guess, if you will. 1% to 3% is typical for piping, which is in the elastic range
(Weld and steel piping). If you have other types of connection methods or a lot of friction sliding
supports, you might want to increase damping up. Again, this is another tuning value that we
have available to us.
7. The last item, number four, is the Mass Model. When we ran this, I believe in the modal analysis,
we kept with a lumped mass model. The consistent mass is a more accurate calculation of
natural frequencies. It's a little heavier, in terms of analysis. It may take a little bit longer, but I
would always try to use consistent mass as my default. The default is lumped and I have to
change it to consistent.
8. Then what we're going to do is do the error check on this, and then we'll run it.
9. Now, I'm showing a graph about our friction. A 1% critical damping has a very large dynamic load
factor of about 50, when you're running it right at the natural frequency. So even though we're
putting a very small deflection on our connection point, we might get a large deflection throughout
the system because with that amplification as indicated by the red line in the image below. If we
would put 3%, we'd be down to 16 times as our multiplier as indicated in the image below by the
blue line. So, we got a great variation in our response, based on that damping term.
10. With that, we'll go back to our CAESAR II model and finish up the input for the dynamic model.


Video 4 – Static Load Cases and Static Analysis of INLET PIPING Model
1. Well, now, let's go back to CAESAR II. I'm showing you the main menu for CAESAR II. We're
going to go under our model Inlet Piping as seen on the title bar, and it's been a while since we
spent time discussing the input for dynamics. I want to make sure it's been run. I'll click Piping
Input, and we see the model.
2. I will do the error check just to make sure there are no

problems with the dataset. Click Start Run.
3. We see we’re fine. Now, this model would not have to be

analyzed in a static analysis, if the system was linear. I could go right to dynamics without a static
solution, if the system was linear and there was no hangar sizing.

4. We do not have to size any springs in this model. They're already defined.
5. I do have a nonlinear restraint in the model. I do have to push it through the static analysis.
6. I'm going to look at our load cases. I'll click Edit Stack Load Cases, and that'll bring up the load
case menu.

7. We see that the current load cases for this model include hangar sizing. Now, you might recall
that, once you establish a set of load cases for a model, that set of load cases gets used in any
subsequent submission for analysis. So, the first two load cases, 1 and 2, were associated with
sizing springs.
8. We don't have any springs in this model now, so I'm going to click on Recommended Load
Cases, and we're going to see that those first two load cases will be removed.
9. We have an operating case. If you're familiar with the alternate sustained case, that's our second
load case listed, which is to address a more recent change to B31.3 and calculating all of the
support configurations for sustained stress calculation. This load case 2 will have that plus Y in
the model active if it's active in the operating case. It will not be active if it is not active in the
operating case. That's what we call the alternate sustained load case. This alternate sustained
case is a recent addition to CAESAR II. If you're running an older version, you will not have this.
So, as far as the dynamics are concerned, as long as you have load case 1, we’ll be able to run
the analysis. However, version 2016, I think, was the first version that had this all in the sustained
case.
10. Then, we have a regular sustained case, where the nature of that plus Y restraint is determined
by the L3 load case explicitly.

11. Finally, 1 minus 3 is the standard expansion stress range calculation. Yes, those are the load
cases I wish to run. Now, we may not even reference these, but this is what I would run for this
analysis. I'll click Accept.
12. So, there are our load cases. The check mark under Alternate SUS/OCC means: yes, use the
support configuration from load case 1. Whatever supports are active in load case 1 will be active
for this sustained stress calculation.
13. The L3 calculation will reset nonlinears, so it satisfies this L3 load vector alone.
14. Click (the running man) Analyze load cases to run the analysis.
Then, we see our output screen.
15. So, we'll take a quick look at the operating case. I'll look at the plot, 3D plot.
Click 1 (OPE) W+T1+P1+H, and then click 3D Plot .

16. I'll grow the plot of the model. Not much going on in it. Again, since we have a piece of equipment
at the end of the small piping, it has to be flexible enough not to put too much strain load on the
nozzle connection at that equipment. That means added flexibility. As you add flexibility, you
reduce the natural frequencies, and that might be the cause of any kind of vibration problem,
where we're kind of picking up some vibration off of the pump nozzle.
17. That's our static analysis, and we're ready to go and develop our dynamic input for this model.
We'll close out the Static Output Processor. I closed the outside x, the red x, and that gets me
back to the main menu.
18. Now, we can go down into Dynamic Analysis for the job Inlet Piping. Click Dynamic Analysis.

19. So, we see our Dynamic Input Screen. We see all the tabs for all the different dynamic analyzes
we have in CAESAR II.
20. The way to clean up the amount of tab shown is to select your Analysis Type. I'll click on the
Analysis drop-down arrow, then and select Harmonic.
21. Now, we have a manageable set of tabs to work with. We discussed these through the
PowerPoint earlier. First, we'll define our Excitation Frequencies. We'll define Harmonic
Displacements for the mechanical vibration input. Again, if we're doing flow-induced vibration or
pulsation, we would define a Harmonic Force instead. We can then modify our mass model or
even our stiffness model for this dynamic model by adding Snubbers or adding Lumped Masses
and, finally, Control Parameters.

Video 5 – Defining Excitation Frequencies, Harmonic Displacements and Control
Parameters
1. We are in excitation frequencies. We have two lines; they're both comments, so I'm going to
click the plus sign to add a new line.
2. If I wanted to add just one frequency, I’d just type in one number on that new line, and that's all I
need to set that frequency. Again, we're going to sweep through a range of frequencies.
3. We're going to start at 5 Hz. Type 5 in the Starting Frequency of row 2.
4. Now we're in optional data, we're going to end at 6. Type 6 in the Ending Frequency of row 2.
5. Now at this point, I'm going to do two analyzes, one at 5 and one at 6. If I specify the Increment,
it will then give us more analyzes between the two limits we defined.
If I type 0.05, I'm going to get an analysis at 5, 5.05, 5.1, 5.015, 5.2, and so on, all the way to 6.
So that's why I'll say we'll get 21 analyzes out of this range.
6. Load Cycles is also an optional entry, and we're not going to specify a value. I'm going to show
you some more about how this program works, and we'll leave Load Cycles blank for now.

7. We are not specifying Harmonic Forces, we are specifying Harmonic Displacements. Now,
what I'm going to do is, again, open up a new line under Displacements in mm. Click the plus
sign to add a second row. Now, the units that we're using are going to be the same units that
were used for the input file. I don't care what your system units are for your machine. It is the
units file for our units associated with our input, and that was in mm.
8. I'm going to type in a 0.1 mm deflection. Now that's, again, rather a large deflection, but I'm
going to be using that to get things started. I want to see what happens when I put that kind of
deflection on the anchor note.
Now, we are doing harmonic analysis, and you can get very large amplification. So even though
we're only putting a 0.1 mm deflection, we can get very, very large deflections out of the system if
we run it right at the natural frequency and if there's no damping. We have three terms that can
be used to adjust the overall displacement response. That is the input magnitude, the frequency
that we're forcing the system at, and also then the damping value. We have our 0.1 mm.
9. I'm going to just say it's in the x direction. I can use the drop arrow and say X or just type in
the letter x for the Direction.
10. Phase angle. Now this is where I'm going to just specify zero. It is required. This gives us that
cosine wave. You will recall that from the PowerPoint. Type 0 in Phase. Now, why would I be
interested in a phase angle? Well, if I had several different displacements specified maybe at
several different anchors. You can see how I could have phase angle zero for one, phase angle

90 for the second, phase angle 180 for the third, and 270 for the fourth. And it's like a four-
cylinder engine basically. They're all firing 90 degrees out of phase with one another, but they're
all being analyzed together.
11. Now, we have to specify the node where we want to apply this load. I'm going to put it at that
anchor. That's at node 1090. Type 1090 in the Start Node.
12. Now, if I wanted to put on a string of nodes, I could use that same Start, Stop, and Increment
that we used over in the Excitation Frequencies. I don't see that too common, so we're just
putting at this one node 1090 alone. That's our displacements.
13. We are not changing the mass model. Again, we're going to eventually say, we want consistent
mass as a very sophisticated approach now. We already added the extra nodes in the model for
a better mass distribution, and a better tracking of the mode shapes, so section nodes help with
that. We don't have to do anything in Lumped Masses.
14. We are not putting any dynamic restraints on our model. There is no data for Snubbers.

15. The last field, we're going to fill out. The last tab we're going to fill is Control Parameters.
16. We do have a non-linear restraint in this model, so this setting on the first row must be completed.
Now what I'm going to do just to show you what happens if you don't specify, I'm going to leave it
blank for now, and we'll see what happens.
17. We have no friction on our model, so I don't care about the multiplier at row 2. It won't be used
anyway.
18. I said I want to go back to 1% critical damping rather than 3%. Change 0.03 to 0.01 in row 3.
Piping well seal systems that are hanging run about 1% to 3%. This is, again, that one wildcard
we have. If I make the row 3 Setting field smaller, I'm going to get larger amplification of the
applied load.
19. Finally, on row four, Settings, I'm going to click on Settings, and I'll click Consistent mass.
It's a better model.


Video 6 – Error Checking, Analyzing, and Harmonic Output Processor
1. Now again, I did not specify row number one in the last video. I'll click the red X to error check
it.
2. I receive the following message: Error is found. Well, what type of errors do I have shown? It
shows all the different tabs we have available-- Control Parameters, Lumped Masses, Snubbers--
roll my mouse wheel down, and I'm looking for red ink. Again, this is input echo-- things that
we didn't even specify for harmonics.

3. After scrolling down, we see the error. "The static load case number, the second item on the
control parameter spreadsheet, must point to a valid, already run static case from which CAESAR
II can determine the non-linear restraint settings for the dynamic run."
4. So that's it for the messages. It says-- the first line is that we're running a harmonic analysis.
5. So we need an item in the first line. So, I'll close out the CAESAR II Dynamic Syntax Check.
I'll click on the Setting drop arrow. If there are nonlinear restraints in the model, make them
linear in accordance with the operating load set. Select 1 W+T1+P1+H(OPE).
6. That is set. I'll click the error check again.
Now, it says error check OK. There are no messages in that file. Again, rolling my mouse wheel
to check for messages. So I'll click Close in the CAESAR II Dynamic Syntax Check.
7. Now, that just did the error check again. If I click the
running man, it will do the error check and the analysis. So
I'm going to click Run the analysis.
CAESAR II says, yes, it's OK, and the analysis keeps on going.

8. Now, just so you know, earlier I was using two monitors to run CAESAR II. A lot of the screens
that we're seeing were on the second monitor, and you saw me dragging them to the other
monitor. We reset the format so everything appears on one screen. So yes, you should see the
Harmonic Solver, and Frequency Options type of information. Now we're still in the analysis
portion. We're not ready to look at output yet, but the programs are going to ask several
questions regarding the type of output that I want to review. Right now, we have this long text in
the Harmonics Solver, and in Frequency Options I could either let the program select frequency
phase pairs, or I select frequency phase pairs, or I just go back to the main menu.
Well, we'll talk about the three options in a minute or two, but let's stick with the default-- let the
program select the frequency phase pairs. We'll talk about the long text in Harmonic Solver later.
I'll click OK.
9. Now, CAESAR II shows me that it has run 21 different analyses. Well, there they are. 5, 5.05,
5.10, all the way up to my 6 Hz. The program is going to develop for me these 21 reports-- the
separate analyses. They are all but the same magnitude of deflection at the anchor node, but all
run at a different frequency. I can use these 21 that were selected automatically, or I could reset
them, or I could delete any one of these individually. We'll talk more of that later. I just want to get
the output menu. Just click OK.

10. So, now we're going to the output menu, and we see the static output processor for harmonics.
It's basically a static analysis, the way we work it. We see the 21 analyses requested. They all
say occasional (OCC). They go from five cycles per second up to six cycles per second. That's
the frequency. We also see our phase angle. For some reason there's a different phase angle for
each one of the analyses. We'll talk about that, too. We have, then, all these Standard Reports
that are available.

11. Now, I notice one thing here, that the units on my machine are called English. You might recall
that we dropped in a Euro file, euro.fil, when we were loading files up. Let's go activate that one.
After we get that set up, we'll talk about these different types of analyses.
12. We close out the output processor with the big red X. Back to the main menu, and from the
Home screen, we go right to Configure. Click Configure.
13. It brings up the configuration file. These are those general settings for this folder.
14. I want to go to Database Definitions.

15. We see in Database Definitions, the units file name is English. I'm going to click on the Units File
Name cell, and it gives me a drop list. Then we see a bunch of files that are in the system folder.
All the ones that begin with a star are on the system folder. There is one that does not have a
star, the Euro file. That's in the local folder, and that's the one we wish to use. I’ll click
EURO.FIL, and there's my European units file. We'll see MPa, and mm, and the like.
16. I want to make one other quick check. I want to check SIFs and Stresses. I just want to look
at the General Settings for Yields Stress Criterion. It's called Maximum 3D shear.
17. We have a choice in Yield Stress Criterion-- Maximum 3D shear or VonMises. We want it to be
3D shear. I'm not going to tell you anything more about it now, but we'll come back to that later.
18. We have to save this. Save and exit. Now we can go back to our output processor.

19. Now, in our Home ribbon we see our Static Output button, but where's our dynamic output? I
don't see it listed on the ribbon. If we go to the ribbon bar tabs, or the tabs for the ribbon and
click Output, then we see the Harmonic report. So, it's Static Reports, Harmonic Reports. We
also have Harmonic Animation. Let's click on Harmonic Report.
20. We are back in our menu for a Harmonic analysis, and we see our European units.

21. Before we go to our output for this model, let's go review those screens that we had during the
harmonic analysis to see what was actually going on.

LESSON 3 – REVIEWING HARMONIC ANALYSIS RESULTS
Video 1 – Effects of Damping, Frequency-Phase Pair, Frequency-Phase Dialog
1. When we ran the harmonic analysis, there were several screens that came up before we actually
got to the output processor. This section now in our presentation will discuss those intermediate
screens. They're basically based on damping, and how damping plays a role in our dynamic
analysis.
2. Now, if I have no damping in a piping system or any system, we see a system below, this
cantilever. As the foundation wags back and forth, with no damping, the maximum of my input will
be the same instant as the maximum for my output. If I follow the base and the tip through time,
they are in phase. There is no shift at all between the input and the output.
3. However, if I do have damping in my system, there will be a phase shift. It takes a little more time;
we'll say it that way, for the tip to follow the base. We see the phase shift between the maximum
value at the base and the maximum value at the tip. So, this lag is induced in this instance.

4. So, if I say, what is the displacement of this piping system during this dynamic event, this cyclic
event? You then have to ask: well,
• What node are you looking at?

• What direction are you looking at?
• At what phase angle to the applied load?
Because it's like taking a snapshot of the piping system, you're freezing it in time, and saying,
show me the position of the tip with respect to the applied load. There is a phase shift, and that
phase shift changes based on the frequency and the damping of this piping system.
5. We will have to investigate. Do we want to know what the maximum displacement of the input
point is or the maximum displacement of any point in the piping system? It’s going to have this
lag.
6. CAESAR II will identify that phase angle that produces the greatest result in displacement of each
frequency analyzed. So this will tie the frequency to a phase angle, and we call it the frequency-
phase pair. Certain frequency, forcing frequency, and a phase angle for that frequency that
produces the largest response somewhere in the system.
7. So, these are snapshots of the system. It's like taking a photograph of the system at an instant
when you get the maximum displacement of a node, and the maximum bending moment for that
state, or the maximum stress for that state. It is not always tied to the maximum displacement, but
the maximum results are usually associated with that maximum response.
8. If you are interested in other snapshots, say the program selects a maximum response for a node
that you're not interested in, or if you're interested in a specific node, you can search the results
for the maximum response for that specific node. So, you get a specific frequency-phase pair for
that response.
9. Now, when we ran the analysis-- this is one of those intermediate screens that we saw-- we ran
an analysis from 5 Hz to 6 Hz, stopping every 0.05 cycles per second. So we have 21 analyzes.

10. These are the 21 frequencies that we analyzed, and the program then shows us which node for
that frequency gave us the maximum response. Now in our model, in every case, node 1,049
gave us the maximum response. That's the elbow indicated in by 1049 in the image. So we're
putting an X displacement on the nozzle, and the maximum response no matter what frequency
that displacement was driven at will displace 1,049, the most in this piping system.
11. 0.1777 is that value. When running at 5 Hz, we put a 0.1 mm deflection in the X at the anchor,
and our elbow is wagging 0.17 mm. There is some amplification. If it was a static load applied on
the elbow, you wouldn't see that much deflection. We notice that the deflection is increasing as

we increase the frequency. Now, here is the phase angle that gives us that maximum response.
12. When the piping system is wagging back and forth at 5 Hz, 5 Hz is provided at the anchor. Five
degrees later, we get to our maximum at node 1,049. As we go through higher frequencies, the
phase shift becomes more pronounced. Now, one interesting point is that when you get to a
natural frequency of the system, your phase angle goes to about 90 degrees. So, I can say that
using this result, there must be a natural frequency in our piping system at about 5.65 Hz,
because that's when we get to this 90-degree phase shift.
13. There are two other columns shown. This is another way of reviewing the response. Instead of
looking at a cosine wave, we can watch the vector spin about the axis. We have a real
component and an imaginary component. The magnitude is based on the magnitude in MAX
TRANS, but the real and imaginary components work with that phase angle. It’s just another way
of looking at it. We'll work with the other numbers we’ve looked at.

14. Now, if I would plot these maximum responses-- they're all at 1,049-- and as I increase
frequency, we see it go from 5 to 6; we notice that the response is amplified at about 5.65 Hz. So
again, we must be getting close to a natural frequency, because when the forcing frequency
divided by the natural frequency equals one, we're getting to that resonance. Once again, we
have more indication that that's our maximum response.
15. If I look at the phase angle-- I mentioned this a few minutes ago-- it's just an observation that
when our phase angle gets to 90 degrees, we get to our maximum response. Even though I'm
only pushing the anchor back and forth a 0.1 mm, I'm getting over 2 mm deflection. This is cyclic,
but that's the maximum deflection. It goes from 0 to 2, to 0 to minus 2, to 0 to 2, and so on. We're
just taking a snapshot of that time. We're trying to capture this maximum response of the entire
system.

16. Keep this in mind that if you go through these phase angles when they equal 90, you're running
near a natural frequency. If I wanted to get a larger number here, I could run an analysis maybe
at 5.64 Hz, and I might even get a higher response, because I'm getting closer to that 90-degree
point.
17. Now, when we're looking at our output or our interaction during the harmonic analysis, we see the
Harmonics Solver screen with a lot of text, and I just skipped over it, but basically what we told
the machine is to let the machine choose the frequency phase pairs to show which snapshots we
wish to see in the output. You have control over this. You can say that you wish to select the
frequency-phase pairs. You would use that if you were interested in a node number that was not
the maximum, or you can just bail out and go back to the main menu. However, the Harmonics
Solver dialog in this instance is telling us about the different ways that we can collect frequency-
phase pairs. Again, a frequency driving frequency, and then the phase angle for the maximum

response. We'll talk more of this later. We just stuck with the users of the program selection, and
we'll just go with that for now. We'll come back and talk about this later.
18. Well, that gets us to our last screen during that error check and analysis. You see the 21 different
analyzes that were made for the harmonic analysis. We swept from 5 to 6 Hz, and we just
happened to catch this one, report number 14, at 5.65 Hz where we're at that 90-degree phase
shift. We're pretty confident then, that that will give us the maximum response in this analysis.
19. Now we can go back and review the results for the analysis. What we've done so far is just
explain those intermediate screens. Let's go back to CAESAR II now and review those results.


Video 2 – Harmonic Output Results for Displacements, Loads and Stresses
1. Let's get back to our output for this model, and let's review some of those results.
2. On the Harmonic Output Processor screen, I'm showing you that we want to go into load case
number 14. That's the one that was close to 90 degrees. I'm quite confident that will give us the
maximum response of any of the reports you see in the list. Each one of the reports is a snapshot
of that piping system running at a different forcing frequency on the anchor. Select case 14, and
again, we're using that one because of the 90 degrees.
3. We're going to look at the displacements for case 14.

4. When I click on Show me the results, we see the following image. Now, this is the displacement
report at 90 degrees-- 90.45 degrees-- at 5.65 Hz.
5. The results from the quick look from the analysis processor said that the maximum response was
2.0150 mm. That's that intermediate report that we saw earlier.

6. Yeah, well, if I take my X, Y, and Z deflection of 1049 and combine them by squared sum of the
squares, we can see that we get the same value we show in the report. So yeah, for every
analysis, there is going to be a node that will give us a maximum response, and we're only
monitoring that maximum node. You might be interested in other phase angles, but then you have
to search for the node that you wish to use.
7. We could also take a look at some of the other reports that are available. Again, we're going to
stay focused on that maximum response. Now, we're looking at the restraints report.

8. If I look at the restraints report, yes, once again, it's showing the frequency and the phase angle,
and it shows us the loads on the different restraints in the system. Now, at the bottom we see an
interesting number-- the number for the inlet node number. That's a very, very large number.
That's where we applied the load. Now, if I have an anchor at 1090 and I pull it even a 0.1 mm, I
will develop a large load. But that is not the load in the system.
9. If I go into CAESAR II, and look at the results-- I'm now going to go in my Output on the
machine-- Harmonic Results.

10. I'll go to case 14; again, I was looking at the Restraint report.
Click View Reports.
11. We now see that very, very large number. That's the imposed deflection on the anchor.
12. If I look at my internal forces and moments report-- I'll go to the Global Element Forces
report.

13. I’ll look at that node 1090. We'll see that the load is only 92 N on that anchor. That's the load from
the pipe.
14. Don't work with that anchor load. That was used to develop the magnitude of the deflection on the
rigid anchor. But we see that the load inside the piping system is much smaller than that.
15. If I look at the other results, yes, the anchor load is too large. Apparently, we have a very high
stiffness, and we put the 0.1 mm deflection on it. We get a very large number. But we should also
take a look at the deflection of that point.
16. Now, remember that we are not looking at it at the maximum deflection at the anchor. The
maximum deflection is at phase angle 0. We are looking at it 90 degrees off of that, so we're
pretty much 0 deflection. Let's go look at the amplitude at node 1090.

17. I'll go back to CAESAR II-- look at my Displacements. Now I'm not going to look at node 1049.
Instead, I'm going to look at node 1090.
18. We see the 1090 listing at the bottom of the page. If we were running it at phase angle 0, that DX
would be 0.1. Since we are 90 degrees out of phase, we're close to 0.
19. It's kind of interesting. The maximum response of our node 1049 is when the displacement of the
anchor connection is 0. Again, we're 90 degrees out of phase.
20. We can take a look at other numbers, too. I'll go back to the PowerPoint and take a look at
stresses. It says occasional 5.65 Hz, 90.45 degrees, Stresses. Hit Enter.
21. Again, I'm looking at that maximum report. This is a snapshot of the stress at every node at that
point in time. We have values shown. What is this Allowable Stress? Well, remember that we are
running an occasional load case in 31.3, and the piping code for occasional stresses has an
allowable stress equation for sustained plus occasional at 1.33 times Sh.

22. Is that a proper allowable stress for a harmonic load? Well, no, it's not. We aren't looking at a
force-based load, where we're worried about collapse in the occasional plus sustained case. We
are, instead, worried about fatigue-type failure. This approach right now is not correct. We should
not be using this approach in CAESAR II. We want to change the type of stress that we show for
case 14.
23. Now, again, for the 31.3, if we're talking about occasional stresses for force-based loads, we're
working against collapse. We will sum those occasional stresses with those longitudinal sustained
stresses, and compare them to 1.33 times Sh. That's our 183 MPa.
24. These are harmonic loads, and we have to look at them in terms of fatigue. We have a quote
from API 618, and if anybody knows anything about cyclic loads on equipment and piping system,
it's these people-- Reciprocating Compressors for Petroleum Service. It goes on to say that, well,
if you have vibration, you should limit the stress due to that vibration to less than 180 MPa-- peak
to peak stress. Also include stress concentration factors.

25. Now, we do not use stress concentration factors in CAESAR II. We base it on the piping codes
stress intensification factors, not stress concentration factors, like notches or bad welds. Sharp
changes in geometry would be a stress concentration factor, or a point where to apply a stress
concentration factor.
26. Again, in CAESAR II, we're calculating a stress state, and it's not from peak to peak that API 618
suggests. We are calculating the 0 to peak stress at that point, and we don't have a stress
concentration factor.
27. What type of formulas should we use for evaluating these stresses? I've given you several
reasons why what we're doing right now in CAESAR II is not correct.
28. Well, then we'll have to get involved in developing our fatigue limits for this analysis, and we'll pick
that up with our next session.

Video 3 – Fatigue Evaluation of Harmonic Loads
1. The question is: How do we get the program to represent fatigue stresses and fatigue stress
evaluation in a harmonic analysis rather than the occasional stress that the program works with
by default? Well, remember that first of all, when we look at our fatigue evaluation in the program,
we have our expansion stress limit-- the allowable stress for expansion stress range-- is a
function of f. That's the cyclic factor. As you get more cycles, f drops lower and lower, and
reduces the allowable stress.
2. In addition to this sort of fatigue evaluation, the program also allows us to establish a fatigue
curve rather than using the equation to develop a relationship between number of cycles and
stress amplitude, or stress range. We will use that type of fatigue curve approach with our
harmonic load evaluation.
3. Now, I could go into Section VIII Division 2, Part 3 and find fatigue curves that look like the image
below. These are the typical S/N curves. S is the stress amplitude, and N is a number of cycles. If
I expect 100 cycles-- using my dotted line -- my stress amplitude is about 200,000 psi. Stress
amplitude that means a stress range of 400,000 psi. Now that sounds awfully large, but again,
we're talking about elastic response.

4. Now, I want to look at response up in the upper end indicated below. There is a million cycles,
and when you look at this type of fatigue evaluation, it doesn't take that long to get to a million
cycles. So, we're talking about ranges down in the area to the right of the area indicated.
5. We could use the curve, I guess. Unfortunately, this is for polished steel test specimens-- which is
A106 grade B material. That doesn't really reflect our piping in the field. We have to worry about
weld and stuff like that, the stress concentrations that I mentioned earlier.
6. Now, we should also concern ourselves with stress range versus stress amplitude. Now, in our
curve that we have been looking at, the stress amplitude is zero to peak. So it goes from zero to a
peak, and zero to a negative peak. The range that we have for stress evaluation is going from
peak to peak. We have to worry about, do we have the right allowable or are we off by a factor of

2?
7. Now, if I wish to establish the curve in CAESAR II I'll have to go back to the Input Processor. In
the Input Processor where I have my Allowable Stress field, down at the bottom I'll get the button
for Fatigue Curves. I'll click that button, and then I can put my fatigue curve in the Material
Fatigue dialog box, number of cycles, and the stress range. In this case, when we're looking at
static analysis, we're with range values. When we work with harmonic analysis, we'll work with
stress amplitude.
So, I can click on Fatigue Curves and I could type in my data, or I'll click on 'Read from file' (as
seen above) in order to get the data.
8. Take a look at the help for the Material Fatigue curve table. There's some good information in our
help on showing you how to build up this type of data for a static analysis.

9. Now, whenever you specify the fatigue curve, the program will use the data in two different
methods. When you're in static analysis, we're talking about a number of cycles that will be
defined in the Load Case Editor. In a dynamic analysis, you define the number of cycles in the
Excitation Frequencies tab. We're going to work with the Excitation Frequencies tab.
10. When you have a fatigue curve set, rather than using OCC as a stress type, the stress type will
change to fatigue or FAT. Then, you know that a homemade fatigue curve was used rather than
that allowable expansion stress range calculation that we see in the code.
11. What we want to see in CAESAR II results-- and confirm that we have it done properly-- is
instead of saying OCC, it will say FAT. Occasional versus fatigue. Now, when we ran the job
earlier, it said OCC and the allowable stress was 1.33 Sh. When we see fatigue the allowable
stress will be pulled off that fatigue curve for the number of cycles you specify. This is how we
can get back to a true fatigue analysis in CAESAR II.
12. Now, what we're going to do rather than typing in our N/S relationships is we're going to click on
'Read from file'. The program will go out to the system folder and show us all the files that we
have there that have this data. They are .FAT files. They have extension *.FAT.

13. If I click on 'Read from file', it'll take us to that System folder, and we see all the FAT files that I
have in that folder. Many of them are shipped with the program.
14. We also see the Division 2 values.

15. I made some special materials uniquely on my machine and they're showing as A53B.FAT, and
API5LX65.FAT.
16. We are going to use the Appx W.FAT. This was supplied with the data file that we're using today.
We moved this earlier in the sessions into the System folder. So, you should have that file
available now. So, we're going to say, yes, click on that Appx W.FAT file, and click Open.
CAESAR II brings in the data.
17. Now, what do I have in that Appendix W file, the Appx W.PDF?
There is a new proposed Appendix for B31.3, Appendix W, titled
High Cycle Fatigue Assessment of Piping Systems. I expect this
to be in the 2018 edition of the B31.3.
18. So, it builds upon the existing code fatigue rules and is used for high cycle-- we're talking 10 to
the eighth cycles, 100 million cycles, on offshore structures. The document gives us this fatigue
curve equation that you see below. We see my expansion stress range that we would do in 31.3
anyway, just like we do in our static analysis. We also see the number of cycles available to
prevent failure at that stress range. So, if I know SE expansion stress range, I can tell you what N
the number of cycles that are allowed for that. We're not going to go into the definition of all this
equation, but we want to get the N/S relationship.

Video 4 – Continuing Fatigue Evaluation
1. We're not going to go into the definition of all this equation, but to get the N S relationship.
2. Let me click the Appx W.pdf file in the Power Point to illustrate.
3. What you see now is a math CAD worksheet that I put together to exercise that equation we're
going to see in Appendix W. We see the terms that we're using in that equation:

4. We also see our equation-- number of cycles based on the expansion stress range calculated.
We are, basically, working with a type of expansion stress that we use in CAESAR II all the time.
5. Then, we see the curve-- S N curve.
6. Now, I want to know what the stress level is for a certain number of cycles.
7. What I can do is I can start to work with different stress ranges until it equals 100 cycles. We see
a stress range for 1,000 cycles. So we see my table at 100 cycles. The allowable stress range is

254,000 PSI. 1,000 it is 122,000; 10,000 it is 58,000; and so on. So that's the curve that we're
showing above. This gives me the data that I will use in CAESAR II-- N S.
8. Also, we see the table that I built. This table is what we're going to access with our data-- Appx
W.fat We see those numbers that I just quoted. We see the 10 to the 8th, that's the last one in the
list-- 3,080 PSI. Now I could use this with any model in CAESAR II on this machine once this is in
my system folder. So, that's our fatigue curve.
9. What we're going to do is we're going to bring it in to our model and use it in CAESAR II. You see
the data, and the fat file. A few points on this-- anything that begins with a star is a comment. In
one of these comments you see I'm talking about-- we're using our Appendix W data to set our
fatigue curve.

10. One thing, we have a multiplier (1.000000). We are putting our data in US units, so there's no
conversion value necessary. This multiplier is a multiplier to the numbers indicated below. Usually
this is 0.5 to indicate that we are putting in stress amplitude, and it'll be doubled for stress range.
Well, the indicated numbers are stress range values, so I'm putting my multiplier of 1.
11. Whenever you work with these types of tables, I would suggest you first run a test case to make
sure that the data that you’re supplying is used properly. If there's any misunderstanding in what
you assumed, the program would prove to you that it is right or not.

12. Now the stress that we wish to calculate-- if I was calculating a stress in CAESAR II for 31.3, it's a
basic equation you see below with axial term, bending term, and torsion term. Now this would be
used for the occasional stress, but again, I'm changing this into a fatigue stress to get the fatigue
allowable.
13. Do I have to worry about the equation that gets used? Well, the program does modify the
equation for stress for fatigue away from occasional. The occasional stress is always using the
sort of formula above-- the code defined calculated stress, but in fatigue—we’ll use a user defined
stress formula with the user defined fatigue curve.
14. Now, where do I define my stress formula? Why would it not be the equation? Well, in CAESAR
II, we will use the separate stress calculation available in the configuration editor to set our fatigue
stress formula. It could be VonMises-- which is the octahedral shearing stress calculation--
distortion energy calculation, or the Max3DShear stress calculation. Since we are using the
Appendix W equation which references the expansion stress range, we're going to use
Max3DShear.
15. What we'll do before we build up the model is that we'll first take a look at the configuration file to
make sure that we have the proper stress equation. Then after that, bring in the FAT curve from
Appendix W for our analysis.

16. After this, we're going to go back now to CAESAR II. I'm in the output processor now. I want to go
back to the main menu, close the output process with the X. On the Main Menu, click on
Home, and we'll take a look at the Configuration File.
17. We now see my Configuration Editor.
18. I want to go click SIFs and Stresses, then click Yield Stress Criterion, and if I click on
Max3DShear, I get a drop arrow. Then, we see my two choices-- Max3DShear and VonMises.
We're going to stick with Max3DShear. That's the default setting. Didn't have to change
anything in the Config. If you had to change something, certainly click the Save button, but I'll
click Save anyway. Save it. Our stress equation has been changed for fatigue stresses. Let's go
back into the harmonic input and see how that impacts our work so far.
LESSON 4 – PROPER STRESS EQUATION

Video 1 – Addressing a Fatigue Approach, Inputting Fatigue Curve Data

1. Let's go back to the harmonic input and make some changes to the analysis to get the stress
equation in properly. Now, how do we get the program to show fatigue stresses in the harmonic
analysis rather than those occasional stresses? Well, we have to specify the number of cycles we
wish to use.
2. Then, if I specify the number of cycles, the program will use the equation (Max3DShear or
VonMises) that we just specified in our configuration file to set the stress, and the Load Cycles
will be used to set the allowable stress.
3. I'm going to specify 1E8 as the number of cycles, 100 million cycles. There are several
references in the code to this number. Basically, if you get to this point, it's below the endurance
limit and there should be no failure. It's not an uncommon point.
4. Now, once I put that 1E8 in my input and I run the analysis, now all of my reports are labeled as
fatigue stresses. That's my indication that I've changed from an occasional stress to a fatigue
stress calculation.

5. If I look at my results—we see my stress results for this load case-- I now have a new code
stress. I'm not going to compare it with the other one. I'm just going to tell you that this is the 3D
maximum shear stress equation, not the B31.3 occasional stress.
6. But look-- allowable stress is zero. There is no allowable stress associated with this analysis yet.
We have yet to associate our fatigue curve with this model. We'll have to go back and introduce it
to our model.

7. Indicated below you can see the fatigue curve that we wish to use. This is the one from Appendix
W, and we have to put in the static model, not the dynamic model.
8. We'll go back to the input file. We'll go to the first element of the model, and from the Allowable
Stresses auxiliary field, we'll click on Fatigue Curves, and we'll click on Read from file.

9. Then, that's when we get that Appendix W file from the System folder. Again, that should have
been put into the System folder early on in this harmonic section.
10. When I click on that Appx W.FAT file, it'll bring in the data and show it on the screen. Now, you
might not recognize the numbers because we are running in metric. These are the stress ranges

for the system in MPa, and our data was in US units. This gets it into the model, and I can then
use this with my harmonic analysis.
11. One thing on that file. Below you can see the actual content of the file. We put a value of 1 for the
Stress Multiplier. If you look at all the other FAT files in the system folder, they'll say 0.5. That 0.5
is useful in converting from amplitude-- which comes from a lot of those fatigue curves-- into the
range calculation that we wish to use in the expansion stress calculations. Clearly, I specified 1.0
because the indicated numbers below are range numbers, not amplitude numbers. You can see
the data that I've provided the program.
12. In the other window you can see how it is reflected in terms of MPa versus psi.

13. Yes, the first number, 254,000 psi, is 1,754 MPa.
14. Now, when I put in that Fatigue Curve-- and again I'm asking for 1E8 cycles-- so actually, I want
to see 21 MPa (number 7 in the Material Fatigue image above) as my range calculation, and
when I go to my harmonic analysis, I will divide that by 2, because what we see is a stress
amplitude. I want to use an amplitude for my limit. So, I'm not going to use 21, but 10.62 is going
to be my allowable stress amplitude for 100 million cycles.

15. This is how we want to set this up and we'll go back into CAESAR II now and do that. We will
introduce the fatigue curve and get our outputs set up for this analysis, and we're done.

Video 2 – Apply Fatigue Data to INLET PIPING Model, Review Harmonic Results
1. Let's update our CAESAR II model to reflect the changes we made. There were two changes we
made. One is we introduced the fatigue curve in the static model, and then the Dynamic Input
we specify the number of cycles.
2. So, let's first go to the Piping Input for the Inlet Piping model.
3. We now see the first pipe in this model 1010 to 1015.
I'm looking at the auxiliary data area for Allowable Stresses, level stresses, and at the bottom
Fatigue Curves. Click Fatigue Curves.

4. Then, I can type in my NS pairs going from the smallest number of cycles to largest, or in our
case I'm just going to click Read from file. So, I'm going to click on Read From File and again,
this is assuming that you already placed appendix W.FAT in the System folder that we did early
on in the session when we first got started.
5. I'm going to click Appx W.FAT (appendix W.FAT), and then click Open.
We now see the data.

We see my 100 cycles. I can get to 1754 MPa.
At 100 million cycles, my stress range is 21 MPa. Click OK.
6. Now, we're not going to be using that fatigue curve in our static analysis. We'll only use it for our
dynamic analysis. For me to get this data into the dynamic dataset, I have to do the error check
and run the analysis. So, let's just click on Batch Run.
7. That fatigue curve will not be used in the analysis because my load case is still called expansion.
If I would have gone in and changed it to fatigue style or fatigue type of analysis, then that fatigue
curve would have been used.
8. Yes, we now see my expansion case. It might be instructive to go point that out how we could
change the expansion case into a fatigue. This is not part of our harmonic session but thought it'd
be worthwhile.

9. I'll click Static Analysis.
10. In Static Analysis on row 4, if I wanted to use the fatigue curve, I could change Stress Type
to FAT.
Then, in Load Cycles specify the number of cycles; say 1 million or something like that.
Then, we'll be using the fatigue curve for the limit on that. I'm not going to fool with the Stress
Type. We'll just change it back and leave it alone.
11. Now, that gets our static model in good shape. Now, we have to go update our harmonic input to
specify the number of cycles. Close the Static Analysis out and now back to Inlet Piping, Dynamic
Analysis. Click Dynamic Analysis.

12. For me to invoke the fatigue calculation rather than the

occasional evaluation, I’ll have to specify the number of
Load Cycles. Again, I'm going to type 1E8 because
100 million cycles is a point at which the allowable
stress at that point will not cause any failure over time.
It's at the endurance limit of the material.
13. With 1E8 for Load Cycles, I'll click the Analyze (running man).
14. That goes on into the dynamic analysis, and once again I have screens that ask for my response
on when I want to capture my frequency phase pairs. I'm just using the default settings. We'll talk
about those other choices later. That gets me to my output screen. We see it says FAT indicating
that we are using the fatigue equation. In our case, it is 3D maximum shear stress, and we have a
fatigue curve associated with it.
15. Now, if I click on case number 14 and look at Stresses.

16. Look at that report, click View Reports. We will now have an allowable stress.
The allowable stress is 10.62. You might recall that the range for 100 million cycles was 21 MPa.
Well, then the amplitude associated with that is half of that number so 10.62.
17. Now, we do not have a high stress in the system. Node 1020 has the highest stress ratio-- 53% of
the allowable stress.
1020 to 1031, this is the branch connection and our stress is 5 MPa. Our allowable is 10. We're
using half the allowable. It's a very low allowable. For fatigue, that makes some sense. So, the
system is not overstressed even though we get a lot of deflection of that elbow, this model
doesn't really show us a big problem. It is a good model for demonstrating and developing our
harmonic input.
18. Now, we have more things to look at in the Stresses. We see the stress report (which you can
view above in step 17) -- very calm system. So, those other screens that we came across when
we're going through the analysis to capture different frequency phase pairs, in next video, we'll
start talking about working with those screens.

Video 3 – Selecting Frequency Phase Pairs
1. Up till now, we've been using the default frequency-phase pairs selected by the program. Again,
the program will go through one cycle for every analysis and find the one node that moves more
than any other, resultant deflection, and use that phase angle where that maximum was achieved
to report the entire model.
2. Now, this is a dynamic analysis, the piping system is always moving. So, how can we list one
stress? Stresses are moving as well.
Now, what we do is, basically-- I would call it like it states in the Power Point:
• Take a snapshot of the piping system at the instant when that maximum is achieved.
When I say, when that maximum is achieved, that's at a certain phase angle, because it's going
to be cyclic, it's going to happen again and again.
3. You have some control over when that snapshot is taken. Again, the program, by default, finds
the largest deflection and uses that to set the snapshot, but that might not be the node that you're
interested in. Maybe you're interested in a Tee somewhere else in the system. So, you might
want to select that T and get the maximum response of that Tee.
4. What are we going to do? Well, you remember the Harmonic Solver screen when we did the
harmonic analysis? The program lists all the different frequency-phase angle pairs, frequency-
phase angle, and the largest deflection. We can see the maximum of two. When we were shown
the screen below, we were given the choice of how we wanted to select the frequency-phase
pairs.

5. What if the node that the program lists is not a node that you're interested in? You can select any
node to sort through this check. Now, there is no guarantee that a point of maximum nodal
displacement will also be the point of greatest stress amplitude, so you might have to look around
for these values. You have the ability to choose your own frequency-phase pairs based on the
deflection of any node.
6. When we normally ran this, so far, we've always let

CAESER II select the frequency-phase pairs as
highlighted by the red rectangle below.
7. The screen you see behind the Frequency Options screen is the screen right after the Harmonic
Solver one with the phase angle data. So, you would click OK on the smaller CAESAR II window
as seen in the image on page 1, and then it presents the Harmonics Solver screen with a bunch
of text. I skipped over the text, so I'm going it read it now.
"A phase relationship exists for this solution, because either a nonzero phase angle was specified
for an input load, or damping has caused a phased response."
That's what we have, we have a phased response.
"In a “phased” solution, the maximum response can occur at any time during the system cycle,
ranging from 0 to 360 degrees (corresponding to the range of 0.0 to T seconds, where T is the
forcing period, (1/f)."
So, we're not going to talk about time of this cycle, but instead the angle associated with that
time.
"There are 21 frequency solutions for this job."
We asked for 21, 21 solutions, and we have the ability to review which frequency-phase pairs we
wish to use.
"Each frequency solution can contain up to 18 phase solutions."
360 degrees divided by 18 is, basically, a report every 20 degrees.

“Output (99 max) cases will be built for the frequency-phase pair selected.”
"Users may select frequency phase pairs or CAESER II can select them based on the maximum
displacement for each frequency."
We've always used the CAESER II selection, now we're going to let the user select. If you click
the User selects frequency/phase pairs, and then click OK, the program allows you to, first of
all, select a frequency to work with. I'm going to use 5.6 just like seen below.
8. Then we'll see that at 5.6, Node 1049 still has a maximum deflection, but the deflection is 1.508
mm, at a phase angle of 48.98.

9. We're seeing that the phase angle could change when you pick different frequencies for the
maximum response.
10. Now, once you select the frequency, then you're given the ability to select a node. So, we're
going to say OK, 5.6 looking at Node 1049, that's our maximum response node that we're familiar
with from earlier.
11. Then in the same Reviewing Harmonic Solution window you see the 18 snapshots that can be
taken of that cyclic load. You know 0, 20, 40, 60, 80, 100, 120, 140, 160, 180, and so on, yeah,
every 20 degrees. You'll note, however, that some of these are off a little bit. Instead of being 40
it's 48, instead of being 140 it's 138.98. We'll talk about that in a second.
I can then select any one of these snapshots by double-clicking on them and it'll put a little
asterisk by it meaning it was selected. So, I can select any one of these reports. Now, let's look at
the Z deflection in the DZ column, we see 0.7 0.91 0.08 1.06. We get through a peak at number
3. That's one that's off. It's not 40 as you can see in the Phase Angle column, but it's off a little bit
to get to the peak.

The Z deflection at 8 is almost 0, that's also at the crazy angle of the 138.98. Then we see our
next peak at 12, also not evenly divided by 20. Then, the fourth one at 17, closer to 0, also not
evenly divided by 20. So basically, it's every 20 degrees, but four of these are kind of shifted in
order to get to a maximum or a 0 deflection point.
12. Now I can show this in another method. Below we see I'm showing the cosine wave (the purple
line) of the applied load. This is the load at 1090, going through one cycle. I'm also showing node
1049 (the green line). What I'm plotting are the values through the cycle shown in the Reviewing
Harmonic Solution window seen below. We can see the phase shift. The phase shift is about 49
degrees. So, from the beginning of the line to the max of the green line is about 50 degrees. You
can see that. So that's a damped response. There's a phase shift in the response, as opposed to
the applied load.

13. Now, the reports that I can get are at these times shown below, 0, 20, 40, 60, 80, 100, 120, and
so on. Again, four of these have been shifted.
14. If I do it as seen below, I can show that I want to get the maximum, 0, minimum, and 0.

Maximum
Zero
Zero
Minimum
15. What the program will do is, it will shift to get to those peaks. Maximum one shifts over, first zero
one shifts down, Minimum one shifts over. While they are normally every 20 degrees, there is a
shift point in order to get the max and min values and the 0 values.
16. If I look at the Reviewing Harmonic Solution report, the ones that I'm really interested in are,
maybe, three. I'm just looking for the increment that is not divisible by 20. I can select all the

reports that we want. I'm going to suggest that we select these four, max, 0, min, 0. We'll take a
look at those.
17. If I want to do that, I can just double-click on each index, 3, 8, 12, and 17, and then I'm going to
get those four reports of my output. Click Apply, and I can take it from there. So those are the
reports that I want to look at just to show you how it works.

18. When I select those four and I continue processing, now we see in the Harmonic Solver that I
have four identical frequencies, but four different phase angles, so again, frequency phase angle
pair. We're working with the same frequency with different phase angles.
19. Then, when I go to the output processor, I can come in here and take a look at all four together.
And if I say, let's look at displacement summary, for example.

20. I can see the values. Looking at the first one, we get a very big number for x deflection at the first
phase angle. Then almost 0 for the second, negative the maximum, and then back close to 0. We
see this cosine wave traveling through the system, depending when you snapped your camera,
basically, is how I think of this.
If you're looking for a specific node or a specific direction of response, you can select your own
frequency-phase angles to dig deeper into our default output.
21. If I look at code compliance too, this might be interesting.

22. If I look at the reports, I can see that my stress, I'm not saying this is high, but a node 1020. The
first report I get a stress of 4 MPa. The second one 0 MPa, then back to four again this would be
a negative term. Then, back to 0.
It's when you snap your camera, basically, is when you get the report. These stresses are always
changing. There's no guarantee that you're looking at the maximum stress for the node you're
interested in. Unless you take a look at these other reports to find where there might be more
significant results other than based on the maximum nodal displacement through that one cycle.
These numbers are always changing through the cycle.
23. So, one more trip back to CAESAR II, and we'll do the same thing with our analysis.

Video 4 – User Selects Frequency/Phase Pairs and Reviewing Results
1. Let's go back to CAESAR II. I'm in CAESAR II, and I'm not going to fool with the static model.
We're just going to go right back toward Dynamic Analysis. Click Dynamic Analysis.
2. Again, we want to exercise the frequency phase angle selection. So same data in the Dynamic
Analysis, no change.
3. I'll click the running man. It keeps on going.
4. Then, we see the Frequency Options question: Do we want the program to select our frequency
phase pairs? Or do we want user select?

5. We're going to click on User Select, and then click OK.
6. Now, in the Solutions exist for the following frequencies we can select the report that we
wish. Select number 13. We know number 14 was bigger response because that's closer to the
natural frequency. I'll click Review Current Frequency.
7. CAESAR II asks what node do we wish to look at? Well, I'm going to stick with that 1049. That
was the one that was the largest response anyway from earlier analysis. So, it's going to then

pick out the 18 phase angles that go through one cycle for the response of 1049. I'll type 1049 in
the Display Values For Node, and then click Apply. Again this is at 5.6 Hertz.
8. Then, I double-click the ones that I'm going to select. Again, I double-click number 3.
Double-click number 8.
These are the ones that are the odd numbers out here.
Now, frankly, I'd work off the one that was a maximum, probably just number 3. These systems
are linear for dynamics. When I go to 3 or number 12, it doesn't really matter whether I go positive
Z or negative Z. I'm getting the same stress results.
9. So that's it. I'll just click Return To Frequency Selection.

10. You can see we already have 4 selected. I've already done 5.6. I could select others as well, but
basically, I'm done. Click Done.
11. Now, we see the four phase angles for a single frequency. Click OK.
12. That will get me to the Output Processor, and I'll take a look at all four Load Cases.

13. You might not have the report shown in the Custom Reports called Displacement Summary. If
you don't and you want to look at it, you can look at all four of the Displacement reports; these are
all blended together. Click Displacement Summary and click to view the report.
14. I'll just scroll to our node 1049, and we see that pattern again. Biggest positive, very close to 0.
Biggest negative, very close to 0. Positive 0, negative 0. Positive 0, negative 0, and so on. This is
not looking at four different frequencies. It's a single frequency, but at four different phase angles.
Numbers are always changing
15. Now we can do this one together. I can go to Code Compliance. Code Compliance is a stress
report, but it kind of summarizes all these together. So I expect to see a large stress (1 (FAT)
5.6000 Hz 48.98), 0 stress (2(FAT) 5.6000 Hz 138.98). Large stress (3 (FAT) 5.6000 Hz 228.98),
0 stress (4 (FAT) 5.6000 Hz 318.98). Select all the Load Cases. Click Code Compliance. Click
Report.

16. I look for a significant number. I'm looking at 1031.
1 (FAT) is 2 MPa, 2 (FAT) is 0, 3 (FAT) is 2 MPa, 4 (FAT) is 0 as seen below:
17. Again, the frequency phase pair is an important consideration when you're reviewing results for
the harmonic analysis.
18. Let's get back to our PowerPoint now. We did that demo, we looked at that. We'll finish up then
with a few more closing points to end the harmonic analysis.

Video 5 – Conclusion and Closing Comments
1. Let's wrap up this discussion on harmonic analysis and CAESAR II. So, what we were doing, so
far, is developing that data for CAESAR II analysis. We put up displacement, mechanical
vibration on our anchor. Again, we could also put force sets on for pulsation or flow-induced
vibration. However, there are things that could get in the way of getting these results.
2. One thing that we're doing is basically we're doing our forensic engineering. We have a system
out in the field right now that is vibrating. So, we dug up the old CAESAR II model for statics and
ran it through the harmonic solution in CAESAR II to see what the response is.
3. If we have a real system out in the field and we have a model on the computer, we should be able
to correlate the two of them. If they match, then we should have greater confidence in our work. If
they don't match, then maybe they have to be tuned up. Remember that you control the mass of
your model (M) and the stiffness of your model (K). If you don't match what's out in the field, try to
take a look at the mass and stiffness of your model and see where that might not match what you
have in the field. A lot of times, it might be your boundary conditions.
4. Now, in our model, we did not have high stresses or high deflections-- both reasons for redesign.
There is no reason to modify our system. However, if there was a problem, if we were
overstressed and fatigue, perhaps redesign would be necessary. What would you do to fix the
model?
5. Well, remember that we're working off the ratio of the forcing frequency of the natural frequency (
). If I'm enabled to change the forcing frequency ( ), it might be a fixed speed
equipment. Then, the only way to change the response is to change the natural frequency ( )
of the system. That's what we were talking about earlier, de-tuning the system to not get the
amplification-- large amplification from harmonic loads.
6. If I wanted to de-tune the system, well the natural frequency is a function of the stiffness and the
mass of the system ( ). Well it's kind of hard to change the mass of the system,
but it's rather simple to change the stiffness. When you add an additional restraint in the model,
you're going to increase the stiffness. Usually you want to increase the natural frequency rather
than decrease the natural frequency for more dependable operation.
7. What you will do is you will find a point that is moving a great amount in your results, either in a
harmonic result, or in a moment we'll show you natural frequency calculations. Find a point of
high deflection and put the restraint at that point. Put a point of high deflection in the same
direction. Then, you can test that fix in CAESAR II, make sure it works there before you go install
it in the field.

8. We're getting back to calculating system natural frequencies. Remember, we're doing harmonic
analysis. We were just running certain frequencies through the processor. If we are actually
approaching a natural frequency, it might be valuable information just to calculate natural
frequencies of the piping system to see if it does correlate with what you see in the field.
9. If I go back to the harmonic results for all these different frequencies, then I would plot this
maximum translation which kind of looks like the image below.
10. I'm expecting to see a natural frequency occurring at that peak. It might be quite informative to
calculate those natural frequencies to see what you're starting with. Let's do that, and then we'll
come back and wrap it up.

11. So, back to CAESAR II, and I'm not changing anything. I'm using the last model we ran, and I
click Dynamic Analysis.
12. The Analysis Type is currently set to Harmonic.
13. All I'm going to do is change it from Harmonic to Modal.
14. Now, that all that other input is still with the model as seen below:
I can always change it back to Harmonic and the information will be there.
Click Control Parameters. By default, the program would calculate all the natural frequencies up
to 33 Hz. It's keeping all the other settings. That's fine; I'll click Analyze (the running man).

15. We're running the natural frequency calculations for this model. We're using the consistent mass
approach. It's a little extra overhead to get it going, and look, the very first mode of vibration,
5.649. That's pretty much 5.65, and that's what the program showed as the maximum response.
16. Now, we're looking at the Output for modal analysis, and we can see those frequencies again. I'll
show that report. Select Natural Frequencies, and then click View Reports.

Then, we see the number that the program was showing in the Dynamic Eigensolver. So yes, we
were hitting the first mode of vibration.
17. Now, the nice thing here is that if it didn't match, you can always look at the animations. Click
Animation.
18. Now, it's quite exaggerated, but there you go. You see that's the elbow that's wagging too much
in the x-z direction. Click to start the animation.

19. If I look at the second mode, 9Hz, it's certainly another direction. Select the Load Case drop
down and select 2 - 9.4893 Hz. Click to start the animation. It does participate this one a
little bit too, but it's de-tuned. We're nowhere near that forcing frequency to excite that mode.

20. Third mode is mainly the elbow indicated below.
Select the Load Case drop down, and select 3 - 10.6432 Hz. Click to start the
animation.

21. Let's go back to the first mode. Select the Load Case drop down and select 1 - 5.6498 Hz.
Click to start the animation. If I had to adjust the system to kill this or modify this natural
frequency, I could put a restraint at 1049, but I also could put a guide on the z run of pipe
indicated below, and it would also affect that natural frequency. You don't have to grab the
highest deflection point, but any point that moves in that mode could de-tune the system.

22. Now we have more hints on how to fix a system. Natural frequencies are good to confirm that you
have a good model, in terms of mass and stiffness, and it is also a good way to predict where a
single restraint might be most effective in reducing the response of the system. With that, let's
close it up.
23. Now I'm using the Engineering Dynamics company name only because we worked with them in
the past in other presentations. They do a lot of harmonic analysis. They work with a lot of
reciprocating equipment. Their web page, Engineering Dynamics. They have a lot of good tutorial
information, technical publications that work with the types of stresses that we're working with in
the presentation. I encourage you to look at their site. You can see their website for their
publications, EngineeringDynamics.com shown below.

24. Also, this type of analysis is better suited for forensic work. You wouldn't take a model that you've
just designed, or a system you just designed, and say well:
• What if this elbow wags too far?

• Or what if I've got this pressure imbalance?
• Or what if this deflection is great?
25. You'd go crazy trying to do all those what-if calculations. That'd be poor design work, but I call it
forensic work because it's a great tool. When you see something out in the field you can
reproduce it. If it's a harmonic type of load, reproduce it in CAESAR II, calculate the fatigue stress
of the system, and then use that to model, to de-tune the system. So, we do not get great
amplification. So, if you can't attack the forcing frequency, you can use CAESAR II to attack the
natural frequency of the piping system.
26. I’ve said it before, in many cases, a good design tool, dynamics would be just to calculate the
natural frequencies. If I knew I had a forcing frequency, the equipment speed, at 5 ½ Hz, I would
not want to have a natural frequency at 6 Hz. Usually, when you get to reciprocating equipment,
the lowest mode of vibration should be 20% above the second harmonic of the equipment. We're
getting pretty high frequencies. Natural frequencies are a good screening tool. Harmonic stresses

are mainly for forensic work.
27. There you have it. That is our second set of dynamics. First was modal analysis and we just
finished harmonic analysis. Next, we'll start taking a look at response spectrum methods of
analysis mainly for seismic analysis of piping systems. Thank You.

HEXAGON - PPM - CAESAR II Dynamics One Lesson Guide

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

HEXAGON - PPM - CAESAR II Dynamics One Lesson Guide

Uploaded by

Copyright:

Available Formats

CAESAR II® DYNAMICS ONE

CAESAR II® ONLINE VIDEO TRAINING SERIES

CAESAR II® Dynamics One

David Diehl, Course Author

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 1

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 2

Hexagon PPM Corporation

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 3

Welcome to these Special Videos for Learning

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 4

About the Author

Drawing from over 30 years of engineering, technical support, and

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 5

Lesson 1 – Calculating Modes of Vibration ............................................................................................... 9

Video 1 – Introduction to Dynamics ....................................................................................................... 9

Video 2 – Eigensolution, Examples of Modes of Vibration .................................................................. 14

Video 3 – Continuing Examples of Modes of Vibration, Basics of Modal Analysis ............................. 19

Lesson 2 – Dynamic Input Review .......................................................................................................... 24

Video 1 – Lumped Masses and Snubbers ........................................................................................... 24

Video 2 – Control Parameters (Nonlinear Considerations and Stiffness Factor) ................................ 30

Video 3 – Control Parameters (Eigensolution, Mass Model, Sturm Sequence) .................................. 34

Lesson 3 – Results Review ..................................................................................................................... 38

Video 1 – Reviewing Results in Dynamic Output ................................................................................ 38

Lesson 4 – Exercising the Program ........................................................................................................ 49

Harmonic Analysis .................................................................................................................................... 164

Lesson 1 – Harmonic Overview ............................................................................................................ 165

Video 1 – Introduction to Harmonics.................................................................................................. 165

Video 2 – Harmonic Load Sources and Effects on Pipings ............................................................... 172

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 6

Lesson 2 – Harmonic Input.................................................................................................................... 204

Video 1 – Dynamic Input Processor, Excitation Frequencies ............................................................ 204

Video 2 – Harmonic Forces, Harmonic Displacement ....................................................................... 210

Video 3 – Control Parameters ........................................................................................................... 214

Lesson 3 – Reviewing Harmonic Analysis Results ............................................................................... 236

Video 1 – Effects of Damping, Frequency-Phase Pair, Frequency-Phase Dialog ............................ 236

Video 3 – Fatigue Evaluation of Harmonic Loads ............................................................................. 253

Video 4 – Continuing Fatigue Evaluation........................................................................................... 259

Lesson 4 – Proper Stress Equation ....................................................................................................... 264

Video 3 – Selecting Frequency Phase Pairs ..................................................................................... 278

Video 5 – Conclusion and Closing Comments .................................................................................. 295

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 7

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 8

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 9

If I want to work at instantaneous terms or get calculus involved, I could say,

Or V equals X dot-- the first derivative with respect to time, position.

Or again, instantaneously, it's the change in velocity with respect to

Or A equals X double dot.

Here we have time involved with velocity, acceleration, and the

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 10

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 11

position of any point in the system is defined as X equals A times cosine qt

28. Acceleration as X double dot, second derivative is negative q

If I put that back into my original equation as it stands now, M X

I get M negative q squared X plus KX equals F

Or just factoring it out, K minus M q squared, the

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 12

-- I can get motion.

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 13

Frequencies Stiffness Mass

3. Our equation, again, was K minus M times q squared, that quantity,

5. I could end up coming with the following:

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 14

HEXAGON PPM | CADWORX FREQUENTLY ASKED QUESTIONS 15