HI 2020

Hydraulic Institute White Paper for

Vibration Characteristics
of Stationary Engine
Driven Rotodynamic
Pump Systems
Foreword (Not part of White Paper)

Purpose and aims of the Hydraulic Institute

The purpose and aims of the Hydraulic Institute are to promote the advancement of the
pump manufacturing industry and further the interests of the public and to this end,
among other things:
a) Develop and publish standards.
b) Address pump systems.
c) Expand knowledge and resources.
d) Educate the marketplace.
e) Advocate for the industry.

Definition of a Hydraulic Institute White Paper

An HI White Paper defines a product, material, process or procedure with reference to
one or more of the following: nomenclature, composition, construction, tolerances,
operating characteristics, applications, performance, quality, rating, acceptability criteria,
testing and service for which designed

Comments from users

Comments from users of this white paper will be appreciated to help the Hydraulic
Institute prepare even more useful future editions. Questions arising from the content of
this white paper may be directed to the Technical Director (HITechnical@Pumps.org) of
the Hydraulic Institute. If appropriate, the inquiry will then be directed to the appropriate
technical committee for provision of a suitable answer.

Revisions
White Papers of the Hydraulic Institute are subject to constant review, and revisions are
undertaken whenever it is found necessary because of new developments and progress
in the art.

Disclaimer
This document was prepared by a committee of the Hydraulic Institute and approved by
following Hydraulic Institute procedures. Neither the Hydraulic Institute, Hydraulic
Institute committees, nor any person acting on behalf of the Hydraulic Institute: 1)
makes any warranty, expressed or implied, with respect to the use of any information,
apparatus, method, or process disclosed in this document or guarantees that such may
not infringe privately owned rights; 2) assumes any liabilities with respect to the use of,
or for damages resulting from the use of, any information, apparatus, method, or
process disclosed in this guideline. The Hydraulic Institute is in no way responsible for
any consequences to an owner, operator, user, or anyone else resulting from reference
to the content of this document, its application, or use.

300 Interpace Parkway, Building A, 3rd Floor, Parsippany, NJ 07054

Phone: 973.267.9700 l Fax: 973.267.9055
www.Pumps.org
This document does not contain a complete statement of all requirements, analyses,
and procedures necessary to ensure safe or appropriate selection, installation, testing,
inspection, and operation of any pump
or associated products. Each application, service, and selection is unique with process
requirements that shall be determined by the owner, operator, or its designated
representative.

Units of measurement
Metric units of measurement are used, and corresponding US customary units appear in
parentheses. Charts, graphs, and sample calculations are also shown in both metric and
US customary units. Since values given in metric units are not exact equivalents to
values given in US customary units, it is important that the selected units of measure to
be applied be stated in reference to this guideline. If no such statement is provided,
metric units shall govern.

Committee list
This Hydraulic Institute White Paper was produced and approved by a working
committee that met many times to facilitate its development. At the time the White Paper
was approved, the committee had the following members:

Chair - Michael Coussens, Peerless Pump Company

Vice-chair – C. Kerby Pope, Patterson Pump Company
Committee Member Company
Paul Boyadjis (Alternate) Mechanical Solutions, Inc.
William Marscher Mechanical Solutions, Inc.
Jennifer McGrath Pentair – Berkeley
Maki Onari (Alternate) Mechanical Solutions, Inc.
James Roberts Xylem Inc. – applied Water Systems
Robert Slattery Kop-Flex Inc., Regal Beloit America
Ken Wauligman Clarke Fire Protection Products, Inc.
Clint Zentic SULZER
Matthew Witter (Invited Expert) Structural Dynalysis, Ltd.

Special Acknowledgement
Special thanks to Hydraulic Institute member Clarke Fire Protection Products Inc., which
provided an initial draft for this paper, that was reviewed, edited and updated by the
committee prior to balloting and approval by the Hydraulic Institute.
 Vibration Characteristics of Stationary Engine Driven
Rotodynamic Pump Systems

Introduction
ANSI/HI 9.6.4-2016 Rotodynamic Pumps for Vibration Measurement and Allowable
Values standard excludes pumps driven by stationary reciprocating engines. The
purpose of this paper is to describe stationary engine vibration characteristics and to
inform readers that the Hydraulic Institute is working on a data collection project that
has an end goal of expanding the scope of ANSI/HI 9.6.4 to include guidance for
acceptable vibration limits of rotodynamic pumps driven by stationary reciprocating
engines.

Certain pump applications, such as emergency fire pumps, can use diesel engines as
drivers. The vibration level of an engine drive is significantly different than that of a
typical electric motor drive. This paper will illustrate the operating differences between
engines and motors, outline some typical vibration signatures for various engine
configurations, help interpret vibration data, and determine the severity based on
existing standards.

Electric motor vs diesel engine drivers

Electric motors are commonplace in industrial pump, fire protection systems, and other
process machinery applications. For motor driven pumps, vibration is often used as an
indicator of machinery health. One tool to aid in determining a pump’s health for motor
driven equipment is shown in Figure 1, from ANSI/HI 9.6.4. Historically, this has been a
useful guide to determine acceptability of a new or repaired pump, but engine driven
equipment are outside of its scope. A goal of the ANSI/HI 9.6.4 committee is to improve
the standard to provide guidance on engine driven equipment such as fire pumps.
 Figure 1 - Allowable motor driven rotodynamic pump vibration, example from
ANSI/HI 9.6.4. Refer to ANSI/HI 9.6.4 for details (pump types, operating region,
etc.) on determining acceptance for rotodynamic pumps.

Engines do not have the same characteristics as that of a motor drive. An electric motor
produces a very smooth torque to drive the load and does so without any reciprocating
parts or significantly unbalanced rotating components. Typically, the rotor is balanced to
a tight tolerance to minimize vibration caused by 1x rpm unbalance forces. There may
be some slight torque ripple at line frequency or higher harmonics of line frequency, but
these are typically in a range of one to two percent of mean torque, if they exist at all.
Excessive motor vibration due to residual imbalance or slight shaft bow will be exhibited
at 1x rpm, just like in a pump. Misalignment can result in excessive vibration at 1x and
2x rpm.

An engine generates torque in a fundamentally different manner than an electric motor.

Fuel is burned in a cylinder, and the expanding gas generates torque on a crankshaft
through a slider crank mechanism containing a piston and connecting rod, as seen in
Figure 2. The generated torque is not smooth. A single cylinder, four cycle engine
produces a torque pulse as seen by example in Figure 3 (Nestorides, 1958). The
resulting load on the pump will include both lateral forces at the bearings as well as
torque transmitted through the engine coupling. This load versus time is not a smooth
sine wave but a complex wave shape that leads to generation of strong harmonics at
the firing frequency and it multiples. Furthermore, the cylinder to cylinder combustion
process is typically not uniform and in a four cycle engine there are noticeable, and
possibly dominate, harmonics at half engine running speed. These frequencies can be
transmitted to the pump. In either case, the engine firing process inherently generates
greater vibration levels as compared to electric motor drives.
Engine vibration has a number of unique characteristics:
1. The torque delivered to the load must be reacted against by the engine mounts.
Because the torque is non-uniform and pulsating (Figure 3) it has strong
harmonic content. This causes the vibration forces generated at the mount
locations to also have strong harmonic content. If any of these harmonics
coincide with a natural frequency that involves significant motion or reaction at
the mounts, a strong resonance can occur that may violate vibration
specifications. This can lead to damage such as fretting wear or metal fatigue in
the engine mounts, the engine itself, or the driven pump.
2. In reciprocating engines, forcing is made up by two components: (1) the pressure
forces generated by the various cylinders; and (2) the inertial forces that are due
to parts in alternating motion and to the crankshaft not being axisymmetric
(parametric excitation). The piston and connecting rod are a significant
reciprocating mass, which are accelerating and decelerating in a non-sinusoidal
manner. This can cause the generation of high harmonic-content vibration forces
and moment couples depending on cylinder layout.
3. The crankshaft is fitted with counterweights, not only to balance the offset crank
pins and rotating eyes of the connecting rods, but to also cancel out the shaking
forces and moments generated by the various reciprocating masses, including
the connecting rods, cross-head, and pistons. However, complete cancellation is
not typically possible, leading to vibration levels that are typically higher than
would be expected of a purely rotating machine.

For example, Figure 2 shows a slider crank assembly of a typical in-line four cylinder
engine, which will naturally produce a secondary shaking force in the vertical direction
(Taylor, 1985). This vibration can be minimized through the use of counter rotating
balance shafts rotating at twice crankshaft speed, at the expense of cost and efficiency.
An inline 6-cylinder engine is inherently balanced with no shaking forces in any direction
generated by the reciprocating or rotating parts (Taylor, 1985), although any non-
uniformity of parts or clearances leads to a degree of the shaking forces be present. In
any case, the pulsating torque is reacted laterally by the engine mounts. These
vibration-inducing forces are proportional to load on the engine and cannot be
minimized or eliminated using clever counterweight or balance shaft design.
 Figure 2 - Slider Crank Assembly

Figure 3 -Typical single cylinder torque pulse for a four cycle engine

The torque profile depicted in Figure 3 is inherent to how a single cylinder of an engine
produces power and is completely normal. It is how the engine is applied and installed
that will determine the vibration severity.

Machinery mounting
Pumps must be installed per the manufacturer’s recommendations.

In a typical stationary horizontal split case or end suction pump application, the driver
and pump are both mounted to a common base to maintain alignment. The unit is then
shimmed and fastened to a foundation (housekeeping pad) and is often filled with grout.
In an engine driven vertical turbine pump application, the engine is typically mounted to
a base separate from the right angle gear. Irrespective of this difference, the mounting
of the base to a foundation is typically the same as described above for horizontal split
case and end suction pump systems.

From a vibration standpoint, rigid mounting of the engine is acceptable when the
foundation is of sufficient mass to attenuate the vibration produced by the system. The
rigid mounting arrangement must be designed to avoid a resonant condition being
excited when operating at running speed. If properly executed, engine vibration is
minimized, although vibration transmitted to the facility may pose a problem if the
foundation is not of sufficient mass or isolated from adjacent structures.

To minimize the vibration transmitted to the base and the facility, compliant rather than
rigid mounting can be used. This involves supporting the engine on vibration isolators,
which would be installed between the engine and the base. In addition, isolators can be
installed between the foundation and a common ungrouted base, with no additional
vibration isolation between motor and pump.

Measuring vibration on an engine

Locations to measure horizontal, vertical, and axial vibrations on the pump and engine
are shown in Figures 4, 5, and 6. A typical horizontal split case pump installation is
shown in Figure 4, an end suction pump in Figure 5, and a vertical turbine pump in
Figure 6. However, for a vertical turbine installation, the measurement locations
normally taken on the pump would instead be taken on the right angle gear. The
locations where measurements are to be made and logged are listed in Table 1.

Figure 4 – Typical Measurement Locations – Horizontal Split case

(Applicable to any between bearing pump (BB1 – BB5))
Figure 5 - Typical Measurement Locations – End Suction Pump
(Applicable to OH0 and OH1 types)

Figure 6 - Typical Measurement Locations – Vertical Turbine

(Applicable to VS1 type)
 Table 1 - Measurement Point Definition, Location, and Direction

Name Location Direction

AEO Axial
HEO Engine Outboard Horizontal
VEO Vertical
AEI Axial
HEI Engine Inboard Horizontal
VEI Vertical
API Axial
Between Bearing (BB) Pump
HPI Horizontal
– Inboard Bearing Housing
VPI Vertical
APO Axial
Between Bearing (BB) Pump
HPO Horizontal
- Outboard Bearing Housing
VPO Vertical
AG Axial
HG Gear Input Shaft Housing Horizontal
VG Vertical
A Vertical
Y Vertical Pump (VS1) – Parallel to Discharge Flange
X Discharge Head Perpendicular to Discharge
Flange
AP Axial
End Suction Pump (OH0 &
HP Horizontal
OH1) - Bearing Housing
VP Vertical

Note: Measurement locations can be altered if required for safety. Caution should be
exercised near the coupling and other rotating elements. Do not remove any guarding.

ANSI/HI 9.6.4 specifies recommended vibration instruments, methods of measurement,

and data recording methods. Care must be taken in the mounting of accelerometers to
ensure accuracy and consistency. ANSI/HI 9.6.4, paragraph 9.6.4.2.2.2 provides some
guidelines and precautions that should be employed to eliminate sources of
measurement error.

Vibration Data presentation

Depending on the analyzer type and/or the purpose of collecting the data, the vibration
measurements can be presented as either an Overall RMS level or as a spectrum. An
example of raw vibration data collected in the time domain on an inline 6-cylinder
engine with a rated speed of 1760 rpm (29.3 Hz) is shown in Figure 7.

Overall RMS vibration is a measure of the total RMS vibration magnitude obtained using
instruments that integrate the vibration within a fixed frequency range over a fixed
period of time. It is the simplest measurement to present, but contains the least
information. For example, the overall level would be useful when commissioning a
machine and reporting vibration data back to a governing agency or customer. The
overall level is also used when trending the vibration behavior over time in a predictive
maintenance program. Changes in the overall vibration value can indicate wear or
imminent failure of components.

However, a vibration spectrum shows the discrete (filtered) frequency content of the
signal, as seen in Figure 8. Using frequency domain spectral analysis will give the
engineer much more detail about the nature of the vibration and the specific harmonic
frequencies that are present in the data. In this example it can be seen that the majority
of the energy is contained at 88 Hz (three times running speed). This is the kind of
information necessary to diagnose and mitigate problems.

Figure 7 - Time Domain Vibration Measured on an Engine Installation

 Figure 8 - Frequency Spectrum of Vibration Measured on an Engine

Note: Vertical chart axis is in 0-pk units; which is the RMS value divided by 0.707

Data interpretation
There is an ISO standard referring to the vibration of reciprocating machines (engine
driver), which could be considered when evaluating an engine driven system. That
standard is ISO 10816-6 Mechanical vibration – Evaluation of machine vibration by
measurements on non-rotating parts – Part 6: Reciprocating machines with power
ratings above 100 kW. This section of the standard provides specific guidelines for
interpreting vibration levels on reciprocating machines above 100 kW, and it also
includes a severity chart with additional machine classifications. Classifications 5 and 6
apply to stationary diesel driven sets.

For example, an engine driven pump system having an overall level measured on the
engine at 28 mm/sec RMS would be considered unacceptable for a rotodynamic pump
by ANSI/HI 9.6.4, but may be acceptable for a class 5 reciprocating machine above 100
kW per ISO 10816-6.

Of course, these are only guidelines, not firm fixed rules. Every machine is unique in its
tolerance of vibration and fatigue.
When determining whether or not a pump, engine, or intervening gear set possesses
vibration levels within acceptable limits, it should be kept in mind that the vibration level
may be important for two separate issues.

1. Is the vibration excessive such that damage would be expected, either from
clearances being used up or surfaces sliding, resulting in rapid wear, or from
local stresses becoming too high and leading to component fatigue?
2. Does the vibration level of frequency content indicate the possible or likely
presence of a mechanical or hydraulic problem in the pump/driver system?

For electric motor driven equipment, with the possible exception of cavitation detection,
both of the above issues are typically adequately addressed by determining the RMS
vibration level at the bearing housings and comparing that level to standards such as
ANSI/HI 9.6.4. When troubleshooting, vibration frequency spectra are obtained to
determine which frequencies are dominant, and the physical implications. The same
approach works well for reciprocating engine drivers as well, but for reciprocating
engines paying attention to the time domain (i.e. time-waveform plots of vibration versus
short scale time) is also very important, as illustrated in Figure 9. This is because of the
impulsive nature of the engine firing process, as shown in Figure 3. In addition, this
“nature of the beast” impulsive behavior is typically expected to lead to much higher
RMS vibration levels, as well as specific harmonics (Figure 10), than an electric motor,
even in a well-tuned, properly installed engine. Some of this added vibration and high
harmonic behavior transmits to the pump, resulting in higher overall and filtered RMS
vibration levels at the pump, as is also illustrated in the field data presented in Figure
10. The pump itself therefore also would be expected to have higher RMS vibration
levels. The specific RMS level that the engine can tolerate is recommended by ISO
10816-6, and the pump levels are an active subject of HI investigation. Clearly, the RMS
levels representative of typical, acceptably behaving pumps driven by reciprocating
engines are expected to be higher than those listed for electric motor driven pumps in
ANSI/HI 9.6.4 (i.e. the levels will be increased to account for reciprocating harmonic
transmission). These levels must also satisfy the issue that the vibration levels do not
cause damage to the pump, which is the subject of current HI investigation.

Figure 9 - Axial and Torsional Vibration Transmission from a Diesel Engine

through a Right Angle Gear Drive to a Vertical Pump
Based on the way an engine produces power, half order harmonics are normal and to
be expected. However, unacceptable vibration may occur when one or more of these
harmonics line up with natural frequencies of the structure, even when the forces may
be modest. If at the time of commissioning (i.e. new system), the vibration is shown to
be unacceptable, it may be due to a structural resonance. If the half order vibration level
increases over time it may be related to engine tuning. For example, Figure 10 is a
1760 RPM engine driven vertical pump frequency spectrum that shows half order
harmonics with the 1.5 order harmonic being amplified by a structural resonance.

Figure 10 - Frequency spectrum illustrating harmonic content typical of a

reciprocating engine
Conclusion
The type of driver coupled to a rotodynamic pump can have a significant impact on the
pump vibration characteristics. This paper described some of the fundamental
differences in electric motor versus reciprocating engine operation that leads to different
pump vibration levels. ANSI/HI 9.6.4 currently does not take into account engine
operation factors. Misapplication of the motor vibration levels will often lead one to
believe that a normal, healthy engine driven system has a significant vibration problem.
There are some standards, such as ISO Machinery Standards, that provide some
guidelines regarding acceptable engine vibration levels. HI would like to expand
ANSI/HI 9.6.4 to include pumps driven by reciprocating engines. Hydraulic Institute is
conducting a field vibration data gathering campaign to obtain useful operational
information about engine driven pump applications. To participate in the program or
assist with providing field data, please contact HItechnical@pumps.org.
Appendix A: Bibliography

Nestorides, E. (1958). A Handbook on Torsional Vibration. Cambridge, Great Britan:

Cambridge at the University Press.

Taylor, C. F. (1985). The Internal Combustion Engine in Theory and Practice (Vol. 2).
Cambridge, MA, USA: The MIT Press.

ANSI/HI 9.6.4 Rotodynamic Pumps for Vibration Measurement and Allowable Values,
Hydraulic Institute, www.pumps.org.

DIN ISO 10816-6 Mechanical vibration - Evaluation of machine vibration by

measurements on non-rotating parts - Part 6: Reciprocating machines with power
ratings above 100 kW,

DIN ISO 5348 Mechanical vibration and shock - Mechanical mounting of

accelerometers

CLARKE (2018) ETB 013 - Vibration Characteristics of Diesel Engine Driven

Emergency Fire Pump Systems