AA V7 I1 Full Version

ADVANTAGE
Excellence in Engineering Simulation VOLUME VII | ISSUE 1 | 2013
SIMULATION FOR
AEROSPACE
10
Safe Landing
18
Ready for Liftoff
32
Shaping Up
PUT YOUR BEST ON THE LINE.
EVERY TIME.
Tackle your toughest ANSYS simulation tasks with

the pure performance of Intel® Xeon® processors.
Discover the difference of Intel Xeon processors

at intel.com/pureperformance.
Copyright © 2013 Intel Corporation. All rights reserved. Intel, the Intel logo, Xeon, and
Xeon inside are trademarks of Intel Corporation in the U.S. and other countries.
*Other names and brands may be claimed as the property of others.
W
elcome to ANSYS Advantage! We hope you Realize Your Product Promise®
enjoy this issue containing articles by ANSYS is dedicated exclusively to developing engineering
ANSYS customers, staff and partners. simulation software that fosters rapid and innovative
Want to be part of a future issue? The product design. ANSYS technology enables you to predict
editorial team is interested in your ideas with confidence that your product will thrive in the real
for an article. Contact us. world. For more than 40 years, customers in the most
demanding markets have trusted our solutions to help
The Editorial Staff, ANSYS Advantage ensure the integrity of their products and drive business
ansys-advantage@ansys.com success through innovation.
ANSYS, Inc. For ANSYS, Inc. sales information,

Executive Editor Editorial Contributor Southpointe call 1.866.267.9724.
Fran Hensler ANSYS North America 275 Technology Drive Email the editorial staff at
Support and Services Canonsburg, PA 15317 ansys-advantage@ansys.com.
U.S.A. For address changes, contact
Managing Editor AdvantageAddressChange@ansys.com.
Chris Reeves Art Directors
Ron Santillo
Neither ANSYS, Inc., nor Wall-to-Wall Studios, Inc., guarantees or warrants accuracy or
Editors Dan Hart completeness of the material contained in this publication.
Erik Ferguson ANSYS, ALinks, Ansoft Designer, Aqwa, Asas, Autodyn, BladeModeler, CFD, CFX, Chip Power
Kara Gremillion Graphics Contributor Module (CPM), Chip Thermal Model (CTM), DesignerRF, DesignerSI, DesignModeler,
DesignSpace, DesignXplorer, Engineering Knowledge Manager (EKM), Explicit STR, Fatigue,
Mark Ravenstahl Robin Steed
Fluent, Full-Wave SPICE, HFSS, ICEM CFD, Icepak, Icepro, Maxwell, Mechanical, Mesh Morpher,
Brenda Westcott Multiphysics, Nexxim, Optimetrics, ParICs, PathFinder, PExprt, Polyflow, PowerArtist,
Design PowerArtist Calibrator and Estimator (PACE), Professional, Q3D Extractor, QuickEye, Realize
Editorial Advisor Wall-to-Wall Studios, Inc. Your Product Promise, RedHawk, Rigid Dynamics, RMxprt, RTL Power Model (RPM), SCADE
Display, SCADE Lifecycle, SCADE Suite, SCADE System, Sentinel, SIwave, Simplorer,
Tom Smithyman
Simulation-Driven Product Development, Solver on Demand, Structural, Super-Compact, TGrid,
Ad Sales Manager Totem, TPA, TurboGrid, Vista TF, VerifEye, WinIQSIM, Workbench, and any and all ANSYS, Inc.
Helen Renshaw brand, product, service, and feature names, logos and slogans are registered trademarks or
trademarks of ANSYS, Inc., or its subsidiaries located in the United States or other countries.
ICEM CFD is a trademark licensed by ANSYS, Inc. LS-DYNA is a registered trademark of
Livermore Software Technology Corporation. nCode DesignLife is a trademark of HBM nCode.
All other brand, product, service, and feature names or trademarks are the property of their
© 2013 ANSYS, Inc. respective owners.
EDITORIAL
TAKING FLIGHT
Aerospace and defense leaders rely on
engineering simulation to get their innovative
ideas off the ground.
By Sin Min Yap, Vice President
Industry Strategy and Marketing
ANSYS, Inc.
A
s you will learn in the form the industry — without investing in modeling and simulation industry, to the
overview by Aerospace extensive physical prototyping or wind entire hardware and software system.
and Defense (A&D) tunnel tests. Whether organizations are exploring
Industry Director Rob For more than four decades, ANSYS the use of advanced composites materi-
Harwood, this sector has been helping the industry’s public- als, developing innovative phased array
currently faces many and private-sector leaders to accomplish antennas, or creating radical engine rede-
diverse challenges. The A&D industry is their most ambitious engineering goals signs that incorporate new fuels and com-
under intense fiscal pressure amid sharp via simulation. Today, ANSYS customers bustion processes, today their challenge
defense spending cuts. Consequently, include the top five aircraft manufac is really to re-invent an entire industry.
product innovation has become a key turers, the top 10 defense contractors, the On the beach at Kitty Hawk in 1900,
driver of growth and profit — whether top 10 space agencies, and the top eight Wilbur and Orville Wright could not have
applied to reducing fuel-burn costs, meet- electronics manufacturers in aerospace anticipated that, someday, an unmanned
ing defense department affordability and defense worldwide. aircraft would touch down on a distant
goals, or dramatically lowering the cost of In working with these industry pace- planet. Similarly, none of us can truly
spaceflight, for example. setters, ANSYS has gained deep and forecast what the future of the A&D
industry will look like. At ANSYS, we are
excited to be part of that future, and we
How can aerospace and defense will continue to invest in the solutions
leaders create mind-boggling

that engineers need to accomplish the
next great industry and product develop-
innovations in a deflationary ment transformation.
profit-margin environment?
A question we often ask ourselves at unique insights that guide the develop-
ANSYS is, “How can aerospace and defense ment of our modeling and simulation
leaders create mind-boggling innova- solutions, ensuring that they deliver capa-
tions in a deflationary profit-margin bilities that support industry best prac-
environment?” Engineering simulation tices for product development process
provides the solution. By developing and improvements that directly impact A&D
testing designs in a risk-free, cost-effective companies' business initiatives. These
virtual world early in the product devel- capabilities span structural mechanics,
opment cycle, A&D engineers can explore fluid dynamics, electromagnetics, and,
the effects of the most complex physical with the recent acquisitions of Apache
phenomena on unique product designs. and Esterel, power/thermal management
They can imagine, create, optimize and and embedded software code validation
evaluate multiple design alternatives for (respectively). The depth and breadth of
engines, wings, radar systems and other our technology provides solutions not just
components that have the power to trans- to discrete components but, unique in the
© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 1

TABLE OF CONTENTS
10 18 24
FEATURES
6 18 32
BEST PRACTICES ANALYSIS TOOLS WING DESIGN
All Systems Go Ready for Liftoff Shaping Up

Fuel efficiency, environmental compliance, Recent technology developments from ANSYS Mesh morphing reduces the time
safety, affordability, sustainment and help aerospace engineers address pressing required to optimize an aircraft wing.
innovation are priorities of aerospace and engineering challenges.
35
defense companies. Simulation is helping
24
industry leaders to meet these and other
strategic goals. ANALYSIS TOOLS
AIRCRAFT ICING
Flight Simulator
10 Ice Shape Aeroelastic Prediction Workshop
BRAKING SYSTEMS Fluid flow simulation that predicts ice assesses computational methods for
formation on aircraft can help maintain safety, predicting unsteady flow fields.
Safe Landing reduce test costs and decrease weight.
Esterel solutions help Crane Aerospace &
28
Electronics to design braking systems that
are certified for safety.
ROCKET ENGINES
14 Testing the Next Generation

AIRPORT EQUIPMENT
of Rockets
Structural analysis provides additional thrust
On Board with Simulation for analyzing rocket engine test equipment.
ThyssenKrupp uses virtual analysis to
increase load limits while maintaining safety
and costs for aircraft passenger boarding
bridges.

35 38 51
SIMULATION@WORK DEPARTMENTS
38 48
THOUGHT LEADER ACADEMIC
A Systematic Approach Gathering Dust

Oil and gas leader FMC Technologies is A new simulation approach helps to
making a science out of systems-level improve particle removal efficiency of
simulation. wet scrubbers.
42 51
INDUSTRIAL EQUIPMENT TECH TIP
Magnetic Attraction Designing Solid

Simulation aids a revolutionary
magnetizing machine to produce magnets
Composites
Employing ANSYS Workbench workflow
with precision-tailored magnetic fields,
streamlines the simulation of solid
forces and behaviors.
composites.
ABOUT THE COVER
The aerospace industry has been
45 a pioneer in deploying modeling
and simulation to design
ENERGY
complex engineering systems.
Engineers can imagine, create,
Hot Stuff optimize and evaluate multiple
NEM reduces cost and improves efficiency for
design alternatives for engines,
concentrated solar power generation.
wings, radar systems and other
components.

NEWS
Simulation in the News

WHEN TWO PLUS TWO many new and critical solutions, enhance- than a tenth of overall vehicle energy
MAKES MORE THAN FOUR ments to pre-processing and meshing losses. The consortium’s Chris Thorne,
Desktop Engineering capabilities, and a new parametric high- program manager for heavy-duty vehi-
deskeng.com, September 2012 performance computing licensing model cles, said, “We want to get the consult-
to make design exploration more scalable. ing parties — specialists in engineering
Traditional simulation packages are being ANSYS 14.5 extends the advancements (Romax), modeling (ANSYS) and lubrica-
extended — not only to help different dis- delivered in 14.0 and provides a great tion (Castrol) — around the same table
ciplines work with the same models, but number of new and advanced features to co-engineer a general solution in an
to share the results across higher-level that deliver solutions for customers to environment where the normal business-
collaborations. This roundup of inter- amplify their engineering effectiveness, related constraints have been removed.”
views — which includes Barry Christenson simulate their most complex engineered
from ANSYS — underscores the concept products and accelerate the time to mar-
that “better, faster, cheaper” is today’s ket for their products.
norm. Christenson notes, “Working with
the multiphysics aspects of a single elec-
tronic chip, though complicated, is not
a systems or multidisciplinary point of
view. However, making sure that struc-
tural, hydraulic, electromagnetic inter-
ference and airflow design requirements
are met for an aircraft design definitely
qualifies as the latter.”

DOORS OPEN TO INNOVATION
ANTENNA AIMS AT DUAL BANDS Business & Gentlemen
Microwaves & RF businessgentlemen.it, January 2013
mwrf.com, October 2012
Electrolux discusses the importance of
Antennas are critical components in simulation in creating new business
WLAN and other wireless systems. opportunities. The company works with
With the integration of ANSYS subsidiary
Researchers at Xidian University lev- Esterel Technologies’ SCADE Suite with ANSYS suppliers to incorporate virtual analysis
eraged ANSYS software to realize effec- Simplorer in version 14.5, users can virtually into business routines while conveying
validate power electronic and mechatronic
tive designs for such applications. The the importance of simulation as a design
systems by simulating embedded software with
team used a modified ground structure the hardware, including electrical, mechanical tool. Collaborating with ANSYS, Electrolux
(MGS) to achieve dual-band operation. and fluidic subsystems. developed the initiative Innovate to
The antenna design exhibits a monopole- Compete, which is part of a plan to cre-
like radiation pattern and acceptable peak ate a network of people with the desire to
gains across the operation bands. DRIVETRAIN EFFICIENCY build innovation by leveraging the syner-
OF BIG TRUCKS TARGETED gies with new partners.
Automotive Engineering International
ANSYS 14.5 ENABLES sae.org, October 2012
DEEPER DESIGN INSIGHT FLORENCE TAV RAILWAY STATION:
ansys.com, November 2012 A UK Energy Technologies Institute (ETI) EXTRAORDINARY DESIGN
project expects to boost fuel efficiency of Railway Strategies
Technology enhancements in ANSYS 14.5 large trucks by cutting in half the amount railwaystrategies.co.uk, September 2012
deliver engineering productivity and of parasitic losses that occur in their lower
innovation through unprecedented multi drivetrain systems. These losses, related Architectural firm NInsight studies and
physics analysis and HPC capabilities. to lubricating oil agitation and component modifies architectural designs and build-
The recently released version delivers friction, together can account for more ing materials in a virtual environment.

One of its latest projects is the Florence A FUTURE WITH DRIVERLESS ALL HANDS ON DECK
TAV railway station, an innovative struc- VEHICLES REQUIRES SENSORY Oil & Gas
ture that is already being hailed for unique ADJUSTMENTS europeanoilandgas.co.uk
features, such as a dramatic arching roof. Desktop Engineering January 2013
The roof is problematic from an engineer- deskeng.com, October 2012
ing perspective, so the firm applied ANSYS For open-sea environments, aluminum
software to address its performance under In the future, cars may become self- decking is ideal because it is weather resis-
different rainfall and drainage situations, driven. “A driverless car must have a good tant, has a high degree of functionality,
the erosion patterns of its many glass understanding of, and must keep track of, has a long life, and has low maintenance
surfaces, and the effects of falling rain- the positions of nearby objects. One of the costs. MB Hydraulikk is a leading designer
water on pedestrian safety. The structure technologies that enable a car to do that and producer of lightweight deck equip-
is expected to be completed in 2016. is radar,” said Sandeep Sovani of ANSYS. ment. To adhere to strict quality stan-
“That’s all about antenna design.” The dards, the company uses the latest design
greater challenge is to go beyond simu- software throughout the design process,
lating simply radars and onboard devices including ANSYS tools for strength analy-
and simulate the entire car as an intercon- sis of complex 3-D models.
nected system.
THERMAL STREAM
Go beyond KEM
simulating simply kem.de, January 2013
radars and onboard SGB designs power transformers so that
devices and simulate they are highly reliable and durable —

and so they minimize load losses. The
the entire car as company uses ANSYS solutions to closely
FINITE ELEMENT ANALYSIS
SAILS INTO THE MAINSTREAM
an interconnected examine thermal distribution. Engineers
simulated the magnetic field with ANSYS
Manufacturing Engineering system. Maxwell to calculate ohmic resistance;
sme.org, December 2012 they transferred the losses to ANSYS
– Sandeep Sovani, Manager of Global Workbench for calculating temperature
Improvements in finite element analysis Automotive Strategy, ANSYS distribution.
(FEA) software are speeding up engineer-
ing analysis. In biomedical applications,
combined FEA–CFD solutions provide sys- THE CURRENT STATE OF MODEL-DRIVEN ENGINEERING
tems answers that are critical to projects Chip Design Magazine
like modeling how pacemakers interact chipdesignmag.com, December 2012
with the heart and the overall cardio-
vascular system. “When you manufacture ANSYS participated in an open forum on model-driven engineering, approaching the
a part, you don’t just run a simulation challenge from an electronics point of view. Many component companies are moving
once,” said Thierry Marchal of ANSYS. up the supply chain to create subsystems, including embedded hardware and software.
“Each time, you need to rerun the simu- Today, most analyses are set up and performed by a few experts with Ph.D.s — which
lation. With a full 3-D model, you want to can become a bottleneck, since a systems-level model must be simple enough for all
run that very effectively. With a systems engineers to use. “There needs to be a democratization of simulation to the engineering
approach, this type of modeling can be masses,” said Andy Byers of ANSYS.
done very quickly.”
When you
Concept of Operation and
operations Verification and maintenance
maintenance
manufacture
Pro
Requirements and System verification

jec
architecture and validation
a part, you
tion
t de
gra
fini
Integration, test
don’t just run a

inte
tion
Detailed design and verification

nd
st a
simulation once.
t te
Implementation
jec
Pro
– Thierry Marchal, Director, Industry Project definition

Marketing for Healthcare, ANSYS

BEST PRACTICES
ALL
SYSTEMS
GO Fuel efficiency, environmental compliance, safety,
affordability, sustainment and innovation are priorities of
aerospace and defense companies. Simulation is helping
industry leaders to meet these and other strategic goals.
By Robert Harwood, Aerospace and Defense Industry Director, ANSYS, Inc.

T
he aerospace and defense (A&D)
industry is incredibly diverse,
and its engineering teams face a
wide range of challenges.
Commercial airlines are fighting to
stay in business as rising fuel prices
erode profitability. To meet the needs of
this sector, aircraft manufacturers and
their supply chain are racing to develop
more fuel-efficient aircraft with improved
engines and aerodynamics, alternate fuel
sources, and lighter-weight airframes
and systems. All of this is taking place in
the context of tightening environmental,
emissions and noise regulations as well
as the uncompromising need to main-
tain a safe and comfortable passenger
environment.
The major defense-spending nations
are entering a period of fiscal scrutiny,
with inevitable program scope reduc-
tions and possible cancellations. This
drives demand for sustainment engi-
neering and the ability to extend product
life well beyond that intended. Growth
Astrobotic Technology is pioneering the lunar frontier, planning to provide commercial payload
areas of investment, such as intelli- delivery and robotic services through missions of increasing scope and capacity. The structural
gence, surveillance and reconnaissance components of the Tranquility Trek spacecraft comprise aluminum and lightweight composites to form a
technology (including unmanned sys- high-strength yet lightweight sandwich material. Using ANSYS tools, the company quickly designed and
refined the spacecraft to withstand static acceleration and dynamic random-vibration loads of launch
tems), are driven by the mantra of while maintaining an acceptable level of safety.
design for affordability.
In the space sector, the transition has
started from government-dominated pro-
grams to the commercial space-faring
era. The success of companies like Space
Exploration Technologies proves the schedule and quality has been demon- optimization capabilities. Whether the
power of commercially driven engineer- strated: The U.S. Department of Defense application is complex thermal and fluid
ing innovation to dramatically reduce (DOD) recently estimated the ROI of apply- interaction during the combustion proc-
costs. In addition, emerging space- ing modeling and simulation to military ess in an engine, the impact of a bird hit-
faring nations have started a new race programs at between 700 percent and ting a wing, electromagnetic interference
to the moon and beyond, with a focus on 1,300 percent. for antenna co-site assessments, power
delivering complex engineering solutions and thermal management, or validat-
in the shortest possible time. ANSYS: ON DECK AND READY ing software embedded in a mission-crit-
The A&D industry has long been a pio- FOR DUTY ical system, the ANSYS vision is highly
neer in developing and applying model- Because A&D applications are so com- aligned with A&D needs. These customers
ing and simulation tools, designing what plex, the basic requirements for accurate have trusted ANSYS to provide the neces-
are arguably the most complex engineer- simulation are deep and high-quality sary technology for over 40 years.
ing systems in the world. Time and again, physics along with systems modeling However, companies tell us that to
the return on modeling and simulation and simulation capabilities supported meet the challenges of tomorrow, these
investment in terms of program cost, by high-performance computing and deep capabilities alone are not enough.
A number of key modeling and simula-
tion best practices are required to enable
The A&D industry has long been a pioneer improvements to the product develop-
in developing and applying modeling and ment process.
simulation tools in designing the most EFFICIENT MULTIPHYSICS

complex engineering systems. The way that organizations exchange
modeling and simulation data today is
often siloed and highly inefficient; this
can be a source of inaccuracy and uncer-

BEST PRACTICES
Complex A&D applications must be simulated with deep

and high-quality physics along with systems modeling and
simulation capabilities supported by high-performance
computing and optimization capabilities.
Unmanned vehicles, prominent in military operations worldwide, have highly complex systems that require integration of varied electrical components
within a very limited space. This creates challenges for power and thermal management not seen in previous generations of electronic systems. To be
successful, designs must incorporate advanced power and thermal management strategies from the earliest stages of the design process. They also must
assess power and thermal issues across scales, from component to system.
tainty, as data is lost in the translation is a high-profile concern to A&D compa- constrained environment, which means
from one group to another. For example, nies. Of course, it is critical to retain data; that training budgets are more vulner-
an aerodynamics team may perform very more important is that organizations capable than ever before. ANSYS boasts
high-fidelity thermofluid calculations and ture the rationale and intelligence behind world-class physics-based modeling and
provide these loads to the aerostructural engineering data for re-use long after the simulation expertise; our support and
team. However, incompatibility between originator of the simulation has moved services organization is set up specifi-
tools used in each group may mean that on. In addition, due to the industry’s cally to become an extension of your team
the high-fidelity data is inaccurately inter- program-focused nature, duplication and as well as a project partner. In addition,
polated or, even worse, averaged across a redundancy of problem-solving occur. A&D companies can customize the tools
surface to a single bulk number. These Different engineering teams in the same to integrate with their entire organiza-
inefficiencies have long been accepted organization may be solving the same tion and democratize the use of modeling
as a cost of business, but in today’s fiscal problem or even one that has already been and simulation beyond the expert analy-
climate, and with ever increasing prod- solved, but they have no effective way to sis community — without sacrificing qual-
uct complexity, these issues cannot be leverage the solution. ANSYS has embed- ity of results.
ignored. ANSYS focuses on delivering the ded the Engineering Knowledge Manager
deepest yet broadest suite of simulation (EKM) tool into the simulation workflow TAKING A SYSTEMS VIEW
tools available in an integrated working to address these types of issues. Today’s — and more so, tomorrow’s —
environment — ANSYS Workbench — so technology is not just a multiphysics
that organizations can effectively execute WORKFORCE DEVELOPMENT hardware solution. The DOD reported
an efficient multiphysics simulation. One of the biggest challenges for effec- that the F-35 Joint Strike Fighter has an
tive organizational deployment of mod- estimated 18 million software lines of
KNOWLEDGE CAPTURE eling and simulation tools is finding code (SLOC). In the future, it is highly
Particularly in the western hemisphere, engineers who have the right skill sets. probable that the intellectual property
the graying of the engineering workforce Compounding this is today’s fiscally of aerospace and defense systems will

be contained more in software than in
the hardware solution. Therefore, for
modeling and simulation to be truly rep-
resentative of the final real-world prod-
uct, it must capture both hardware and
software performance and deliver a val-
idated solution. ANSYS’ recent acquisi-
tion of Esterel is a demonstration of our
commitment to staying in tune with
rapid developments in the A&D sector.
THE SKY’S THE LIMIT

Aerospace and defense companies face a
ANSYS is active in research projects for advanced technology required in the aerospace industry such
as participation in the AIAA CFD High Lift Prediction Workshop and the Aeroelastic Prediction Workshop. daunting set of engineering challenges
today, and more are on the horizon. But
with great challenges come incredible
opportunities. This issue of ANSYS
Advantage is packed with stories of how
diverse customers leverage engineering
Concept
Requirements & System Production simulation to emerge as leaders in a diffi-
validation
specifications
cult business landscape.
Requirements
From developing complex software
System
analysis
functional &
Subsystem
integration & that controls brakes to understanding ice
Design architectural
verification
Physical
Verifications
design prototype formation on wings, ANSYS customers are
Component
balancing cost and time concerns with
Subsystem
design
integration &
verification
the most important engineering priority
Requirements Detailed of all: protecting human passengers. By
analysis design
Design Detailed component creating and verifying their innovations
design & optimization
Verifications
in a safe virtual world, these A&D leaders
can have the utmost confidence that their
designs will deliver the expected benefits,
Yes
Concept Design Physical
Testing Analysis Production
while also supporting the secure, comfort-
prototype
No able transportation of passengers.
CAD CAE CAM
Other ANSYS customers are explor-
ing unmanned vehicles designed to travel
to far-off planets, creating electronically
invisible surveillance and reconnaissance
A pioneer of early engineering simulation, the A&D industry has the opportunity to re-invent the
processes that lead to innovative design. Applying simulation at the concept stage — to explore possible technologies, and otherwise moving the
alternatives well before ideas are set in stone — encourages true innovation. industry in new and unexpected direc-
tions. For these pioneers, simulation
offers a no-limits environment in which
any idea is possible.
Whatever the specific engineering
Modeling and simulation continue to challenge, one thing is certain: Modeling
support development of innovative

and simulation will continue to support
development of innovative aerospace and
aerospace and defense technologies defense technologies and deliver solu-

tions in a rapid and cost-effective manner.
and deliver solutions in a rapid and ANSYS is committed to helping the A&D
industry realize its product promise.
cost-effective manner.
Reference
High-Performance Computing Modernization
Program, Determining the Value to the Warfighter —
A Three-Year Return on Investment Study; U.S.
Department of Defense, 2010. www.hpcmo.hpc.mil/
cms2/index.php/roi-report

BRAKING SYSTEMS
Safe
Landing
Esterel solutions help Crane Aerospace
& Electronics to design braking
systems that are certified for safety.
By Gregory Mooney, Systems Software Engineering Lead — Landing Systems

Crane Aerospace & Electronics, Burbank, U.S.A.
C
rane Aerospace & Electronics’ principle is to exceed the needs of its cus- most complex component in Crane’s elec-
mission-critical products help to tomers, a responsibility Crane’s engineer- tronic controllers is an invisible one: the
ensure the safety of millions of ing team takes very seriously. thousands of lines of embedded software
airplane passengers. The company must Crane is the industry leader in aircraft code that ensure efficient, reliable brake
meet stringent government safety regula- brake control systems, with 65 percent of control when required during landing.
tions aimed at ensuring that these prod- the commercial market and 80 percent of To support reliable software perfor-
ucts — in this case, braking systems — will the Western military market. The com- mance, the United States Federal Aviation
perform as expected under a broad range pany has more than 25,000 systems in Administration (FAA) has drafted a set
of real-world conditions. service worldwide today. These braking of guidelines under its Federal Aviation
Crane pioneered the antiskid braking systems are critical in ensuring passenger Regulations requiring proof that soft-
industry in 1947 by developing the Hydro- safety during routine landings, as well as ware “performs intended functions under
Aire Hytrol Mark I antiskid system for the during challenging rejected takeoffs. (See any foreseeable condition.” A means to
B-47. Since then, Crane has provided both sidebar.) Brake control systems designed comply with these regulations — that is,
private- and public-sector aircraft cus- and manufactured by Crane include a a way to show that the system meets its
tomers with a wide range of brake control number of mechanical and hydraulic safety objectives — is a standard called
systems and other products — includ- parts, such as control, shutoff and park- RTCA/DO-178B, Software Considerations
ing power, cabin, fluid management and ing brake valves as well as wheel speed in Airborne Systems and Equipment
sensing systems. The company’s guiding and pedal position sensors. However, the Certification. This standard aims to assure

testing facility or customer lab repli-
cating actual aircraft configuration —
which meant that Crane’s engineers
had to go back to the drawing board and
rewrite the code. This was an expen-
sive and time-consuming proposition
at such a late stage of software develop-
ment, especially after hundreds of devel-
opment hours had been invested.
The resulting high costs, large amount
of rework and scheduling issues nega-
tively impacted the company’s customer
Braking control systems manufactured by Crane Aerospace include a number of mechanical and
hydraulic components — all controlled by thousands of lines of mission-critical software code. satisfaction levels. In addition, Crane’s
engineering team had to assemble a
wealth of documentation at every stage
to satisfy FAA requirements. Crane real-
ized that to accelerate and streamline the
software development process — with-
the safety of software by defining a set of Today, while the basic mechanics of out sacrificing ultimate product integrity
development, verification, requirements braking have not changed, brake control or regulatory compliance — it needed to
management and quality assurance tasks systems continue to evolve as an incred- identify an advanced technology solution
aimed at instilling rigor into the software ibly complex blend of control software, that would model and predict real-world
development process. electronically controlled actuators, and performance of these smart systems at a
Essentially, DO-178B requires devel- high-speed digital communication inter- much earlier stage.
opers to test and verify performance of faces with other onboard systems (as
control software under a broad range of well as humans). For instance, in case of SCADE PUTS THE BRAKES ON
operating conditions, a challenging engi- an electrical short or other unexpected MANUAL WORK
neering task. Since 2010, Crane Aerospace event, the software not only ensures con- Crane evaluated a number of model-based
has relied on SCADE Suite software from tinuing brake performance, it alerts the development environments before choos-
Esterel — now a part of ANSYS — to help flight crew about the issue and sends an ing SCADE Suite. The company selected
develop software that satisfies the critical automatic alert to maintenance staff to SCADE because it is a purpose-built soft-
FAA certification process for braking con- address the problem once the plane lands. ware development tool qualified to meet
trol systems. Obviously, the underlying code is mis- the standards of DO-178B up to Level A,
sion critical, because the consequences the highest level of safety for the aero-
BRAKING CONTROLS: NEW for brake failure can be truly catastrophic. space industry. In addition, engineers
COMPLEXITY, NEW CHALLENGES This new complexity has created were impressed by the support that they
Since Crane revolutionized the aero- exceptional challenges for software devel- would receive from Esterel while install-
space industry with its first Hydro-Aire opers, who must test for each input to the ing and running SCADE solutions.
antiskid braking system, brake control brake control software — as well as for a
systems have become increasingly com- broad range of operating events. This dic-
plex. Wheel speed transducers measure tates extremely rigorous, broad-scope test- Software
high-resolution wheel speeds, enabling
modulation and control of brake pres-
ing and verification tasks, even as Crane’s
customers are working against aggressive
developers
sure during operation. Today, brake con- development schedules. must test for
trols have to dynamically adapt to all
runway surfaces, whether wet, dry or
Prior to 2010, the company’s software
engineers managed the requirements of
each possible
icy. For example, most large commercial DO-178B via a time- and labor-intensive input to the
planes feature brake-by-wire systems,
which apply brakes electronically, first
process, which had some evident draw-
backs. Because of the manual nature of
brake control
introduced by Crane for the U.S. Space the process, the finest details of func- software —
Shuttle program in 1973. Continuing
this development, Crane introduced the
tionality were hidden in the underlying
code. The impacts of customer require-
as well as for
first microprocessor antiskid systems ments and system updates were not fully a broad range
in the early 1980s, which apply sophis-
ticated control algorithms to automate
visible — and thus could not be com-
pletely verified by Crane’s engineering
of operating
braking and achieve antiskid perfor- team — until the software was fully devel- events.
mance in excess of 90 percent. These oped and implemented.
systems, like many Crane innovations, There were often surprises when
have become industry standard. software was run in a Crane in-house

BRAKING SYSTEMS
Since implementing SCADE Suite in

2010, Crane Aerospace has realized signif-
icant cost, speed and efficiency benefits in
its safety-critical software development,
verification and validation process. From
the earliest stages of code development,
the SCADE tool enables software engi-
neers to simulate and confidently predict
real-world results, eliminating surprises
and rework at later stages. SCADE Suite
automates the code-generation process
and enables the testing of embedded soft-
ware code against thousands of inputs,
without the need for a target. This sig-
Software req Systems defect Customer defect
nificantly reduces the need for manual
work involved in code generation as well Software code Customer change Interface
as related software development, verifica-
Integration
tion and validation tasks.
During software simulations, SCADE
Suite allows the engineering team to sub-
A significant number of late-stage defects in braking control systems arise from first-stage defects
ject the design to thousands of inputs in the customer requirement. SCADE Suite allows Crane engineers to identify and address these issues
and events to ensure that the software during initial code validation.
will function exactly as expected when
installed in an actual plane.
SCADE allows customers like Crane
to build customized libraries that include
general-utility operators, primary and
ancillary brake functions, and system
interfaces. These reusable libraries make
software development even faster. In
addition, SCADE Suite generates much of
the process documentation automatically,
eliminating hours of work that once were
invested in meeting stringent government
requirements for record keeping.
MOVING BEYOND CODE

VERIFICATION
While there are many ways to verify that
software performs accurately against cus-
tomer requirements, Crane engineers
leverage SCADE to take logic and control a
In-progress simulation of fault detection subsystem
step further — by validating that software
requirements make sense and protect pas-
senger safety. Because simulation makes
system behaviors visible, SCADE allows
engineers to flag exceptions and identify
problems with initial requirements. If not in corresponding software requirements. engineers to quickly see the ultimate
detected and addressed, oversights in ini- These errors, some major and some minor, impact of any design changes on sys-
tial software requirements can result in were caught and eliminated in early val- tem reliability and braking performance.
late-stage issues that delay projects and idation exercises — before the systems They can quickly identify the problems in
add to cost. ever existed in the real world. Prior to the underlying requirements and adapt the
SCADE implementation, these flaws might software model accordingly.
IN-PROGRESS SIMULATION OF not have been discovered until very late in
FAULT DETECTION SUBSYSTEM the manual software development proc VISUAL MODELS OFFER
Since Crane Aerospace first implemented ess — resulting in significant rework for OBVIOUS BENEFITS
SCADE, the solution has identified more Crane’s software engineers. A number of features in SCADE Suite make
than 150 errors in preliminary system Today when errors are detected, what- it easy to communicate defects and other
requirements and more than 180 flaws if scenarios in SCADE Suite allow software issues to aerospace customers. SCADE

software. Using SCADE Suite positions
Crane to easily incorporate any new
brake control innovations as it simu-
lates how software code functions as
part of a larger aircraft system.
In some cases, Crane’s customers
have actually used SCADE’s systems-
level models and what-if simulations to
;
BRAKESB: JBS STATUS2,1,BRAKESK ; Check status consider enhancements to overall brak-
JBS STATUS2,7,BRAKESK ; ing system prototypes during iron-bird
LD LX,PEDALC[WHEEL] ; Select the greater value testing, a dimensionally accurate struc-
CMP LX,PEDALCCP[WHEEL] ;
JH BRAKESC ;
ture specially fabricated to replicate
LD LX,PEDALCCP[WHEEL] ; the aircraft. Instead of identifying “sur-
SJMP BRAKESC ; Continue prise” software performance variations
;
while in the iron-bird environment, cus-
BRAKESK: LD LX,PEDAL[WHEEL] ; Select the greater value
AND LX,#0FFFH ; tomers are uncovering new systems-
LD AX,PEDALCP[WHEEL] ; level insights that inform the entire
AND AX,#0FFFH ; brake control environment. Supported
CMP LX,AX ;
by SCADE software, Crane can offer its
JH BRAKESC ;
LD LX,AX ; customers industry-leading, forward-
looking braking system technologies,
on-time delivery and the high product
SCADE Suite improves internal and external communication by transforming dense lines of software
code (bottom) into intuitive graphics (top). These reveal the software’s underlying logic in a highly quality they rely on every day.
visual manner that all stakeholders can quickly understand.
Aborted Takeoffs:
enables Crane software engineers to gen- used the software’s intuitive libraries A Special Engineering
erate easy-to-understand graphic models
and reports that reveal the inner workings
and graphical models to meet a two-week
turnaround for a scaled-down customer
Challenge
of software code in a way that all process demo of a new control system.
Rejected (or aborted) takeoffs pose
stakeholders — including Crane custom- Today, Crane is able to deliver on cus-
special challenges for brake con-
ers — can quickly comprehend. During tomer requirements, as well as meet the
trol systems. During a rejected take-
simulations, users can view real-time stringent demands of the FAA’s DO-178B
off (RTO), an airplane has much more
software performance values and eas- guidelines, via a compressed develop-
energy than during a normal land-
ily understand the immediate impacts of ment schedule. Day-to-day engineering
ing due to higher speed and a signifi-
model changes. work is much more quick and efficient,
cantly heavier weight. Moreover, an
SCADE Suite translates lines of unin- and errors are detected at a much ear-
aircraft has typically used up much of
telligible software code into visual, intu- lier stage.
the available runway when the pilot
itive graphics that make the software In addition, SCADE software has
has to make the difficult decision to
logic — and any disconnects — apparent, increased Crane’s agility and speed in
abort. While most landings may use
even to non-SCADE users. This has helped adapting to new developments in brak-
only 20 to 30 percent of the plane’s
to create a higher level of customer sat- ing control systems, such as improved
available braking capability, during
isfaction, especially when compared microprocessors, offering more robust
an aborted takeoff up to 100 percent
with the original time- and labor-inten- and sophisticated fault- detection
can be needed to stop safely, and, thus,
sive process, as Crane’s engineering team
the control software must ensure near-
interacts with customers during software
development, verification and validation. SCADE Suite has ideal performance. Since pilots are usu-
ally focused on the conditions that led
CUSTOMER SATISFACTION
enabled Crane to the rejected takeoff, Crane brake con-
TAKES OFF Aerospace to help trol systems offer an automatic braking
With the incredible pressures that air- its customers meet function that applies full brakes during
an RTO. Because rejected takeoffs rep-
craft manufacturers face today, SCADE
Suite supports Crane’s efforts to help its
tight budgets and resent such a great challenge, Crane’s
customers meet tight budgets and sched- schedules. software engineers focus particular
attention on this demanding event dur-
ules — by providing faster, highly accu-
ing their SCADE simulations.
rate software development. Shortly after
implementing SCADE, Crane successfully

AIRPORT EQUIPMENT
On Board
with
Simulation
ThyssenKrupp uses virtual analysis to increase
load limits while maintaining safety and costs
for aircraft passenger boarding bridges.
By Antonio Murias, Research and Development, ThyssenKrupp Airport Systems, Mieres, Spain
F
or most airports, a passenger design by evaluating the performance of tion facilities are located in Spain, China
boarding bridge is not just a many design iterations while taking non- and the United States. The company’s
system to enhance passenger linear material properties and contacts bridges consist of glass and steel side-
comfort on the way from the terminal into account. Parametric analysis and walls with one, two or three tunnels to
to the plane, but a strategic safety ele- shape optimization delivered the required service planes with multiple entrances.
ment. ThyssenKrupp Airport Systems safety margin with the least material A column supports a substantial por-
has installed more than 3,000 passen- possible. The end result is a part with 33 tion of the complex structure’s weight
ger boarding bridges, also called fingers, percent higher allowable load limits that and connects with the moving frame
at airports around the world. The drive- keeps manufacturing costs under control using a roller just below the rotunda.
system frame of the passenger boarding by using 10 percent less material than the The lift-and-drive system incorporates
bridge (known as the bogie) supports original part. a telescopic frame joined to the tunnel
50 metric tons (55 tons) of structure structure near the entrance to the air-
while providing the forward, backward STRATEGIC SECURITY ELEMENT craft. The bogie supports the lift system
and lateral movements needed to dock ThyssenKrupp Airport Systems is a busi- and uses motor-driven wheels to provide
to the aircraft. ness of over 400 people focused on forward, backward and lateral movement
ThyssenKrupp recently used ANSYS building, installing and maintaining pas- for the bridge. The bogie includes a lift
simulation tools to develop a new bogie senger boarding bridges. Its three produc- jack opening that is used when replacing
the wheels. A cabin on the end of the pas-
senger boarding bridge connects the tun-
The end result is a part with 33 percent higher nel to the aircraft.
allowable load limits that keeps manufacturing
MOVING TO
costs under control by using 10 percent less SIMULATION-BASED DESIGN
material than the original part. The bogie originally used in ThyssenKrupp
passenger boarding bridges was designed
in Germany 20 years ago using hand
calculations. The limitations of manual

ThyssenKrupp triple-stand passenger
boarding bridge (PBB) at the Dubai Airport
during the first docking in the world between
a PBB and an A380-800 aircraft
design methods meant that prototypes

had to be constructed and tested in evalu-
Simulations address complex physics
ating each design iteration. Testing identi-
fied problems with early design iterations
that are far beyond the capabilities
but provided limited diagnostic informa- of hand calculation.
tion, so engineers had to rely on intuition
to modify the design. The time required
for physical testing limited the number of
load-case scenarios that the engineering
team could evaluate. Historically, it took
18 months to iterate to a bogie design that plastic deformation. The technology gives design geometry so that design engineers
met specifications. the design team the ability to accurately can evaluate many different design varia-
Recently, ThyssenKrupp Airport predict real-world performance, making tions in a short time. ANSYS Mechanical
Systems integrated simulation into its it possible to postpone prototype build- offers the ability to analyze the wide
design process. The team still regularly ing and testing until the latter stages of range of physics — stresses, deformations,
uses hand calculations, but finite element the design process, when the build-and- vibration characteristics, reaction forces
analysis is now the primary tool to evalu- test method is used to validate the simu- and residual strains — required to deter-
ate proposed concept designs. This simulation results. mine performance of a passenger board-
lation takes the actual geometry of the ThyssenKrupp Airport Systems uses ing bridge.
part into account so engineers can more ANSYS Workbench, ANSYS DesignModeler
accurately determine the response of the and ANSYS Mechanical. Workbench pro- REDESIGNING THE BOGIE
structure to realistic loads while exam- vides an integrated environment to sub- Recently, the decision was made to rede-
ining detailed diagnostic information. stantially reduce the time required to sign the bogie using simulation as an
Simulations address complex physics that prepare models for analysis — including integral part of the design method. As a
are far beyond the capabilities of hand cal- defining loads, constraints and contacts. first step, the original CAD geometry was
culation, such as nonlinear material prop- DesignModeler delivers a fully paramet- transferred into DesignModeler. At the
erties, nonlinear contact conditions and ric environment for defining and editing same time, physical tests were carried

AIRPORT EQUIPMENT
Rotunda
Column
Rotunda and column Lift-and-drive system
out to validate material properties and determine loads based on to determine the location of the fixed points where the bogie
the original design. Engineers used ANSYS Mechanical to eval- attaches to the wheels and lift. Loads and fixed supports were
uate the stress behavior of the original part and the effects of applied. The software approximated the initial structure as a
design changes, such as increasing the radii in high-stress areas. large volume. Material was then reduced at locations with low
The team decided to change the material from cast iron to steel, stresses, resulting in a more optimal structure.
because steel provides better control over material properties, The resulting optimized design weighs 10 percent less than
and then began to redesign the part from scratch. the original design while providing an allowable load of 80 met-
ThyssenKrupp engineers then created a new design in ric tons (88 tons), 33 percent more than the previous-generation
DesignModeler. The model of the previous design used all solid design. Engineers ordered a prototype of the new design and
elements because of the complex geometry of the cast part. It had performed component-level physical testing that verified the
a total of 1.3 million nodes. The model of the new design with simulation results.
a combination of shell and solid elements used only 300,000 Confident in the design results, rather than prototyping just
nodes. Running this analysis with all solid elements would have the drive system, engineers went ahead and built a complete
taken about 12 hours per load case. The use of shell elements bridge with the new bogie design. Testing showed that the design
reduced the solution time to three hours per load case. worked perfectly, so it went into production only three months
Engineers added constraints and loads to the model and sim- after the beginning of the design process — reducing design time
ulated its behavior. One of the more complex load cases involves by almost 89 percent from the hand-calculation build-and-test
jacking up the bogie to change one wheel. When the jack rises, method used for the original design.
the bogie flexes under bending loads, and the load is transferred Overall, the use of ANSYS simulation tools drastically reduced
along the lift jack opening. ThyssenKrupp engineers used a non- the time required to bring the new design to market. In just a
linear contact to maintain the contact at the right location as the few months, ThyssenKrupp engineers thoroughly evaluated the
part flexes. bogie’s behavior under many different load-case scenarios, far
Using ANSYS DesignModeler, the engineering team repro- more than could have been physically tested. The result is a very
duced stiffness and other behaviors of this complicated com- reliable product that far exceeds design specification at a manu-
ponent while defining many load cases and taking nonlinear facturing cost equal to the previous design.
material and contacts into account. It performed stress and buck-
ling analyses, to determine if interior ribs were required, then
validated the forward, backward and lateral movements of this Design time was reduced by
part. The interaction between the wheels and the apron surface
was analyzed for different combinations of material stiffness and
almost 89 percent from the
contact conditions. hand-calculation build-and-
ITERATING TO AN OPTIMIZED DESIGN
test method used for the
Once the analysis identified the part’s problematic zones, engi- original design.
neers added and removed material manually; they also used
the shape optimizer within ANSYS Mechanical to improve
the geometry. An envelope of the initial volume was defined

Model of original bogie design (left) and new bogie (right)
Global displacement results for new design Stress analysis results with mesh for wheel- Linear buckling analysis results for operating
replacement load case for new bogie design load case for new bogie
Stress simulation of original bogie Stress simulation for internal components of new bogie
Previous-generation bogie design New bogie prototype during testing

ANALYSIS TOOLS
Ready
for
Liftoff
Recent technology developments
from ANSYS help aerospace
engineers address pressing
engineering challenges.
T
he global aerospace industry faces many critical challenges centered around engineering and technology. R&D
teams are often tasked with balancing multiple, sometimes conflicting, priorities and developing new strategies
to address these challenges: decreasing aircraft weight, reducing noise and emissions, and maintaining passenger
comfort and security while keeping budgets in check. Whether they’re working on innovative new engine designs,
reshaping wings for better aerodynamics or exploring the use of composites materials in the fuselage, engineers are position-
ing this industry for a future in which all of these goals can be achieved. Just as the aerospace industry has advanced over the
years, ANSYS software has evolved to anticipate new engineering problems. In this article, ANSYS experts share some recent
technology innovations that benefit the global aerospace industry.
ANTENNA SYSTEMS: and shape, and digital space–time sig- world and to predict system performance
SIGNALING A NEW ERA nal processing — have added even more and electromagnetic effects with a high
By Lawrence Williams, Director of complexity. In addition, advancements degree of confidence.
Product Management, Electronics across the aerospace industry, including Recent technology advancements
the growing use of composites materials in ANSYS HFSS take advantage of high-
A ntenna performance has always

been a sophisticated engineering
topic, but recent advances — including
in radomes and airframes, affect antenna
performance in complicated ways.
For decades, ANSYS software has
performance computing resources and
novel numerical methods, making it
faster and easier to solve large problems.
specialized active antennas, microwave helped engineers to assemble 3-D antenna The software applies a domain decom-
circuits and devices, agile beam steering systems in a low-cost, low-risk virtual position method (DDM) to distribute

and more intuitive than ever. This capa-
bility is especially important for applica-
tions such as dielectric resonator filters,
which need to meet stringent design spec-
ifications, including those related to the
operating environment and the device’s
power-handling capabilities. By link-
ing HFSS and ANSYS Mechanical, engi-
neers can confidently predict electrical
performance under varying thermal and
structural loads — a solution that seam-
lessly brings together electromechanical,
thermal and structural analysis for the
New current-coupling capabilities enable modeling of an FEM region containing the feed antenna along
with an IE region on the reflector’s surface and feed-supporting struts that carry current from one region to first time. A filter’s bandpass frequency
another. Modeled radiation patterns are nearly identical to a traditional full-system FEM simulation. response can be quickly visualized as it
shifts from low-power, ambient-tempera-
ture conditions to a high-power, thermal-
deformation state.
By linking ANSYS
HFSS and ANSYS
Mechanical, engineers
can confidently predict
electrical performance
Complex radiation patterns of novel array shapes — including any missing elements — can be quickly under varying
identified using DDM computing techniques and a new array mask feature in ANSYS HFSS software.
thermal and structural
loads — a solution
that seamlessly
brings together
large electromagnetic problems across a
network of computers to solve in 3-D with
New features in ANSYS 14.5 advance
antenna modeling capabilities even fur-
electromechanical,
higher fidelity. Material and geometry para- ther. An important solver enhancement thermal and
metric sweeps, as well as solutions across for HFSS is the ability to current-couple structural analysis
frequencies, can be significantly acceler- FEM and IE regions. For a reflector antenna for the first time.
ated via the multiple-design-point license. system measuring 50 wavelengths in
ANSYS is a leader in developing hybrid diameter, the current-coupling method
solution techniques that accelerate solu- can speed solution time and reduce mem-
tions. As aerospace and defense engineers ory requirements by 81 percent — while
know well, some portions of an antenna modeling the system’s radiation pattern
problem are best solved by the finite ele- with the same degree of accuracy as a tra- COMPOSITES MODELING
ment method (FEM), while other portions ditional FEM simulation of the complete FOR RADOMES:
are best addressed using integral equa- antenna system. SHAPING A SOLUTION
tion (IE) or physical optics (PO) methods. Another significant HFSS enhance- By Sean M. Harvey, Senior Technical
For instance, once antenna perfor- ment is finite-sized phased array analysis, Services Engineer
mance has been optimized as a stand- using DDM and an array mask. With this
alone system, the next step is to assess
its performance when placed within a
radome or on a vehicle. Hybrid solution
capability, engineers can easily assemble
an array by drawing a single antenna ele-
ment, then applying an array mask to rep-
A s concerns over fuel efficiency
increase, lightweighting planes is
an ongoing concern — and composites
techniques — combining FEM, IE and resent the array shape and the possibility materials are an obvious solution. With
finite element boundary integral (FEBI) of missing elements. The full array radi- their light weight, relatively low cost,
methods — enable antenna engineers to ation pattern and near-fields can be cal- electrical transparency, strength and
quickly and intuitively simulate the elec- culated to examine edge effects. A novel structural stability, today’s innovative
tric fields of antenna, radome and vehicle composite excitation feature allows composites are revolutionizing the aero-
upon which it is mounted. Solving such highly efficient solutions for user-defined space and defense industry. For example,
large problems previously required very array-weighting functions. the latest generation of commercial air-
long and time-consuming engineering Thermal bidirectional links in ANSYS craft from Boeing and Airbus are made up
simulations. software make multiphysics studies faster of over 50 percent composites materials.

ANALYSIS TOOLS
CFD fluid–structure interaction analysis Radome deformations resulting from airflow Detailed solid composites mesh can be
reveals flow streamlines and pressure contours over the radome surface can be evaluated once incorporated directly into mechanical assemblies
on a radome surface. whole-aircraft analysis has been conducted. and post-processed. Laminate details such as ply
drop-offs and tapers are easily integrated into the
model for analysis.
While composites offer many benefits, geometry and then automatically pass Engineers can also incorporate the effects
they present significant engineering chal- the new shape into the solver, eliminating of thermal changes on the design.
lenges. Material layers must be stacked in intervention or rework. The integration and flexibility of the
different orientations, at varying thick- New in ANSYS 14.5, solid compos- entire ANSYS suite allow radome engi-
nesses, to ensure structural stability while ites geometries can be evaluated as solid neers to apply tools such as ANSYS HFSS,
creating the complex, curving shapes that 3-D mesh and integrated into ANSYS industry-standard simulation software
characterize aircraft. Perhaps nowhere Mechanical solid assemblies within for 3-D full-wave electromagnetic field
is this challenge more apparent than in Workbench, enabling more accurate pre- simulation. Used together, the suite helps
designing radomes, the curved weather- diction of material stiffness and strength. to ensure the structural strength of a
proof structures that house antennas. This feature complements the existing radome design as well as to verify that it
With more than two decades of expe- shell representation capability, and it was delivers high signal-transmission perfor-
rience in modeling composites, ANSYS designed with the very specific needs of mance — obviously a critical concern in
helps leading aircraft engineering teams aerospace and defense engineers in mind. the radome application.
to overcome the challenge of design- New workflows in ANSYS 14.5 make com-
ing radomes and other composites struc- posites design faster and more intuitive.
tures. ANSYS Composite PrepPost (ACP), The ANSYS Mechanical suite enables AIRCRAFT ENGINES:
a module in ANSYS Workbench, enables parametric analysis for composites RETHINKING INDUSTRY
engineers to import a radome model and designs, delivering increased speed and STANDARDS
perform ply stacking, draping and fiber insight for mechanical engineers. Teams By Brad Hutchinson, Vice President
orientation in an intuitive virtual design can perform what-if analysis to quickly Industry Marketing, Turbomachinery
space. They can determine where compos- gauge the effects of design alterations —
ite layers should start and stop as well as
design appropriate transitions between
thick and thin material sections.
for example, changing the fiber orienta-
tion, thickness or ply drop-off locations,
or even suppressing or including ply
F ew technology areas receive as
much attention and critical review
as aerospace engines. As concerns over
ACP also allows engineers to evalu- stacks parametrically. fuel burn, emissions and noise increase,
ate performance of a composites design, As aircraft engineers use these fea- aircraft engineers are rethinking every
assess its structural strength, and identify tures to make refinements, they can look aspect of the traditional engine. ANSYS
potential regions of failure. By iterating at multiple design points, applying aero- software is a key enabler of their efforts
this process, the team can easily optimize dynamic or inertial loads to assess mate- to develop cleaner, quieter, durable and
a design that thrives in real-world condi- rial strength and displacement. They more environmentally friendly designs
tions. ACP is completely integrated into can replicate mechanical impacts cre- that also fulfill critical safety and reli-
the Workbench platform, enabling air- ated by real-world forces, such as bird ability promises.
craft engineers to change the radome’s strikes or hail, to ensure radome integrity. Several recent developments in the
ANSYS suite reflect emerging trends in
aircraft engine design. For example, engi-
ANSYS Composite PrepPost enables engineers to neers are increasingly concerned with
import a radome model and perform ply stacking, raising turbine entry temperatures to
draping and fiber orientation in an intuitive improve fuel efficiency. However, these
virtual design space. rising temperatures push the limits of
traditional materials and engine technol-
ogies — and necessitate innovative hot-
section cooling strategies.

ANSYS Mechanical is invaluable for ANSYS software has the speed and powerful transient blade row (TBR)
studying the structural strength and dura- power to more realistically simulate simulation methods, known as the trans-
bility of promising new materials that turbulence via scale-resolving simula- formation family of TBR methods. These
may withstand higher combustion tem- tion (SRS) methods, such as large- and methods address the issue of unequal
peratures. Coupled with ANSYS Fluent or detached-eddy simulation (LES and DES), pitch between adjacent blade rows, pro-
ANSYS CFX computational fluid dynam- which are better suited to model the com- viding full-wheel solutions while simu-
ics (CFD) software, Mechanical helps engi- plex behavior of combustors than tradi- lating only one or a few blades per row.
neering teams analyze the effectiveness tional Reynolds–stress (RANS or URANS) The approaches deliver tremendous sav-
of engine cooling systems via a multi- models. The recent development of a ings in computational time and required
physics approach. New communication novel scale-adaptive simulation (SAS) disk storage space, yielding results files
and file handling improvements in ANSYS model provides a RANS solution in stable of a much more manageable size for
14.5 make it easier than ever to link fluid flow regions while resolving large-scale post-processing.
flows, heat transfer and other multiphys- turbulence structures in regions where One valuable enhancement in ANSYS
ics phenomena via ANSYS Workbench. such phenomena are significant, such as 14.5 couples ANSYS Mechanical with the
HPC-compatible features enable users to bluff body wakes or free shear layers. ANSYS CFX TBR transformation methods,
solve complex, numerically large cooling Turbine and compressor aerodynam- streamlining aeromechanical analyses
problems quickly, as well as to simulate ics are at the heart of engine design. After such as blade flutter and damping. Now
hundreds of cases for studying all possi- many years of development, these com- the highest-fidelity aero and mechanical
ble operating conditions. ponents are highly evolved, so realizing methods are linked and are more efficient
Today, aerospace engineers are additional performance gains is a chal- to use than ever. This enables rapid, accu-
rethinking even the most basic processes lenging task. However, much of the anal- rate investigation of aeromechanical phe-
that power engines, including combus- ysis has been steady-state, for practical nomena critical to developing safe and
tion. To support their efforts, ANSYS offers reasons. Developing additional insight durable engines.
advanced combustion models, including requires applying unsteady solution
thickened flame, improved spray and methods, because that is the real nature
fuel evaporation models. These simula- of the flow in the many successive rows TBR methods deliver
tion capabilities help engine designers of blades that comprise a compressor or tremendous savings in
capture complex phenomena, such as turbine. An impediment has been a lack computational time and
fuel–air mixing, heat release and emis- of availability of efficient, HPC-enabled required disk storage space,
sions, that have traditionally been diffi- unsteady blade row methods. Fortunately, yielding results files of a
cult to replicate in the virtual world. ANSYS has introduced and evolved much more manageable
size for post-processing.
Dynamic thickening Local thickening Flame/turbulence Accurate flame
in reaction zone factor as function of interaction: representation
mesh size efficiency function
ADJOINT SOLVER:
STREAMLINING CFD STUDIES
By Chris Hill, Principal Software
Developer
C omputational fluid dynamics analy-

sis is foundational to the aerospace
industry. From external component aero-
dynamics to the complex ventilation sys-
With ANSYS thickened flame models, the
flame is dynamically thickened to limit thickening tems inside commercial jets, CFD studies
to the flame zone only. An efficiency function are critical in many engineering tasks.
takes into account unresolved chemistry/ However, with so many complicated
turbulence interactions.
physical processes to consider, CFD sim-
ulations can be complex and numerically
intensive. If many design iterations are
TBR transformation methods in ANSYS CFX
dramatically reduce computational time and needed to identify a design that meets
resources, providing unsteady full-wheel blade requirements, then the overall compu-
row solutions — yet they require simulating only tational cost can be very high. Ongoing
one or a few blades per row.
improvements to the discrete adjoint
solver in ANSYS Fluent continue to make
these sophisticated problems easier and
faster than ever to model and solve.

ANALYSIS TOOLS
The adjoint CFD solver from ANSYS

enables engineers to focus on one spe-
cific aspect of performance — for exam-
ple, pressure drop across a system of
ducts — and optimize design inputs
based on that single performance mea-
sure. The adjoint solver accomplishes
the remarkable feat of tracking the effect
of changing hundreds of thousands of
design variables simultaneously, via a
single computation.
Fluent’s adjoint solver significantly
Aircraft geometry (left): The contour plot shows displacement that should be applied to the aircraft
speeds overall simulation time by tar- surface to achieve a 10 percent improvement in lift/drag. This displacement is extracted directly from
geting those design areas that are most the adjoint solution data set generated in ANSYS Fluent (right).
important in influencing drag, lift, pres-
sure drop or other critical performance
measures. Engineers can optimize designs
quickly and systematically by homing in
on the most influential parts of the sys- internal flows. Furthermore, they can without having to explicitly capture all
tem, instead of running a series of simu- optimize standard aerodynamic forces scales involved in turbulent flows — and
lations and optimizing the geometry via and moments, including lift-to-drag they are very accurate for the vast major-
trial and error. ratios, for external flows. ity of applications. This technology
Customers can use the adjoint solver enables aircraft wing designers to predict
in Fluent to redesign components such factors that affect lift, drag and ultimate
as heating, ventilation and air condition- TURBULENCE: STREAMLINED fuel efficiency — perhaps the industry’s
ing (HVAC) systems inside aircraft. The SOLUTIONS FOR A COMPLEX most pressing challenges.
solver has the scale and fidelity to model PROBLEM Wing lift for aircraft is better predicted
the complexities of these piping and duct By Gilles Eggenspieler, Senior Fluid via capabilities that model separation
work systems, which function as a life- Product Line Manager and re-attachment of fluidic flows during
support system for human occupants. flight. Understanding the upstream lam-
Engineering teams also can apply the
adjoint solver to external aerodynam-
ics problems, such as the drag effects
T urbulence modeling is critically
important in many aerospace appli-
cations, as engineers seek to continually
inar boundary layers that transition into
a turbulent flow is key to predicting wing
lift or even compressor performance.
caused by antennas, sensors and cam- improve performance of their designs. Some applications require more
eras mounted on the fuselage. An R&D Accurate prediction of a system’s aero- advanced unsteady models. Large-eddy
team recently modeled an unmanned dynamics, heat transfer characteristics, simulation turbulence models resolve the
aerial vehicle to assess design changes mixing performance and other factors large turbulent structure in both time and
that would increase its lift-to-drag ratio. is key to determining performance with space and simulate only the influence
ANSYS continues to refine this revo- high precision — so accurate and robust of the smallest, nonresolved turbulence
lutionary simulation capability with each CFD turbulent modeling capabilities structures. These unsteady models help
new software release. New in Fluent 14.5 are required for aerospace and defense aircraft engineers to reduce the exter-
are improved workflows for setting up and applications, as well as in many other nal noise generated by wheels and wings
post-processing simulations. Engineers industries.
can use this tool to manage many aspects Today, engineering teams use CFD with
of design performance, including flow advanced turbulence modeling to evalu-
uniformity, flow splits and variances for ate performance with maximum accuracy.
Only highly accurate predictions deliver
the performance improvements — which
The adjoint CFD solver from may be only a small fraction of a per-
ANSYS enables engineers centage — that differentiate good design
to focus on one specific from excellent design. Aerospace compa-
aspect of performance and nies rely on accurate simulation model-
ing capabilities to gain a competitive edge.
optimize design inputs Steady-state or RANS models reduce the
based on that single complexity of turbulent flow by averag-
performance measure. ing the velocity field, pressure, density
Detached delayed-eddy simulation of flow over
aircraft landing gear captures the turbulent flow
and temperature over time. These models structures created by the landing gear structure.
offer engineers a highly attractive solu-
tion to predict the effects of turbulence

during takeoff and landing — a grow- vital field. In parallel, ANSYS is making Accurate and robust
ing challenge as global noise regulations huge strides in high-performance com-
become more stringent. puting technology that enables engineers CFD turbulent
ANSYS, with its team of leaders in to make appropriate trade-offs relevant to modeling capabilities
innovative turbulence model develop- highly accurate results for complex simu- are required for
ment and application, continues to push lations within an acceptable time frame.
the envelope in this highly complex but Because high-definition LES simu- aerospace and
lation is often too time-consuming for defense applications.
fast-paced development cycles, ANSYS
has combined the best of steady and LES
approaches. These hybrid models deliver
high-fidelity results for the right computa-
tional price. HPC acceleration ensures that
results are delivered quickly, so engineers computational resources. ANSYS is a tech-
can evaluate many designs in a short time. nology leader in this area, offering a wide
Hybrid models, such as SAS and DES, use range of the most advanced models.
the power of RANS steady models to sim-
ulate flow in the vicinity of aircraft skin Reference
while unleashing the power of LES for the Menter, F. Turbulence Modeling for Engineering
rest of the flow, in cases for which resolv- Flows, Technical Brief, ansys.com/Resource+
ing large turbulent structures is critical to Library/Technical+Briefs/Turbulence+
the study of aircraft noise or blade flutter. Modeling+for+Engineering+Flows
As CFD applications become more
Capturing the turbulent flow structures allows
prediction of the sound pressure level created by complex, more sophisticated turbulence
airflow around the landing gear structure. models are needed. Choosing the right
turbulence model to match the applica-
tion results in accuracy and optimized

AIRCRAFT ICING
Ice
Shape
Fluid flow simulation that predicts ice formation
on aircraft can help maintain safety, reduce test
costs and decrease weight.
By Thomas Michon, Aerodynamics Engineer, and Didier Chartrain, former Aerodynamics and
Performance Group Leader, DAHER-SOCATA, Tarbes, France
B
uildup of ice during a flight manufacturers traditionally use simulated is growing demand for new methods of
alters aircraft aerodynamics by ice shapes that are affixed to the wing (or predicting droplet collection efficiency
increasing drag and reducing the other relevant parts of the aircraft). and ice shapes on complex 3-D geome-
airfoil’s ability to provide lift. Ice accumu- Although the experimental method tries. While this is a significant challenge,
lation also can affect an aircraft engine’s provides the most accurate results, it DAHER-SOCATA uses ANSYS Fluent CFD
functional efficiency. These are impor- is too expensive to use in conducting software to simulate/compute the local
tant safety considerations, so protect- design iterations to identify all modifica- droplet collection efficiency and time-
ing aircraft from ice accretion is a major tions to the system that might be needed. dependent ice shapes. When performed
requirement for manufacturers. European Since most aircraft manufacturers must as early as possible in the aircraft devel-
Certification Specifications and Federal reduce weight, cost (both recurring and opment process, this practice can avoid
Aviation Administration Regulations nonrecurring) and energy consumption costly changes for final certification.
(CS23 and FAR23 for general aviation to improve aircraft performance, many
category) precisely define requirements use numerical simulation to predict ice
in terms of ice protection. Requirements accretion and determine the most critical Most aircraft
for ice shape must be determined for a ice shapes as early as possible in the air- manufacturers use
45-minute flight in icing conditions for craft design process. This avoids the con-
numerical simulation
unprotected areas and for a 22.5-minute siderable time and costs required to carry
flight for protected areas to take into out both a complete flight test program to predict ice accretion
account potential failure of the de-icing or in known icing conditions and experi- and determine the
anti-icing system[1]. mental tests in a climatic wind tunnel. most critical ice shapes
The standard practice by which air-
craft manufacturers meet these require-
Experimental tests can then be reserved
mainly for validation.
as early as possible
ments is to present certifying authorities Dedicated tools, such as 2-D LEWICE in the aircraft
with flight test data confirming that air- from NASA, traditionally have been used design process.
craft performance and flight-handling in this industry to predict droplet collec-
qualities were not critically affected tion and ice accretion. With the continued
by ice accretion. To obtain such data, push to increase aircraft efficiency, there

for droplet collection and thermodynamic validated to experiment. Because there
balance as well as the shape change that were no experimental data for this case,
occurs with accumulating ice. engineers considered the LEWICE results
to be a suitable, highly respected refer-
FLOW SOLUTION ence, and, therefore, DAHAR-SOCATA was
The flowfield resolution is addressed by confident to move from simple 2-D to
the Fluent solver using state-of-the-art more complex 3-D applications.
aerodynamic methodology. An accurate After 2-D validation, the droplet col-
droplet trajectory prediction relies on a lection UDF is applied to different 3-D CFD
precise description of the boundary layer. applications. In the preliminary design
In addition to applying the classical exter- phase, as soon as a 3-D wing definition is
nal aerodynamic model, the simulation available, the team uses the UDF to iden-
must accurately describe the near-wall tify the impinging area that requires pro-
area using the appropriate prism layer to tection by a de-icing system. Therefore,
precisely reproduce local phenomena. A early in the design process, designers can
kω SST turbulence model is used, begin- provide an accurate specification of de-
ning with first order, then employing sec- icing system requirements.
ond order after hundreds of iterations, In another example involving a turbo-
and finally employing third-order MUSCL prop aircraft, the engine inlet must be
spatial discretization. Appropriate use of protected from ice accretion to ensure suf-
the under-relaxation factor allows flow ficient engine performance in all icing
convergence in 600 iterations for a typi- conditions. The inlet lip is usually heated
cal 3-D model. with a piccolo system, which uses hot air
from the engine. Since this air offtake has
DETERMINING DROPLET a direct impact on engine performance,
COLLECTION the system must be carefully sized for
The engineering team computes droplet the inlet geometry. Because of the com-
collection efficiency using a user-defined plex airflow around the inlet, classical ice
function (UDF) developed by ANSYS and accretion codes are not suitable for this
customized to DAHER-SOCATA’s specific type of analysis. Instead, DAHER-SOCATA
needs. Fluent is a flexible software that used the droplet collection UDF to size the
has the capability to incorporate com- heated area of the inlet lip. The team per-
PREDICTING ICE ACCRETION plex physics and enable customization. formed analysis for different critical flight
Ice accretion involves complex physics Strictly speaking, the local collection effi- cases within the icing conditions envel-
and phenomena, including aerodynam- ciency (β) is defined as the fraction of liq- ope. The range of simulations performed
ics, multiphase flow and thermodynamic uid water droplets that strike the aircraft allowed the team to assess the effect of
behavior, which are all time-dependent relative to the number of droplets encoun- aircraft angle of attack and airspeed on
with geometry deformation. DAHER- tered along the flight path. In areas in the impinged area.
SOCATA couples these physical phenom- which β is positive, ice is likely to appear.
ena using the moving deforming mesh The UDF is based on an Eulerian descrip- THERMODYNAMIC BALANCE
(MDM) capability and the unsteady solver tion of the droplet concentration in the To determine the thickness of the ice, the
in ANSYS Fluent. This process accounts atmosphere, given as a volume fraction team computes the thermodynamic bal-
(α). DAHER-SOCATA engineers specify ance using a dedicated UDF in Fluent
the droplet diameter as well as the drop- developed by DAHER-SOCATA. The ice
let free-stream velocity. Since the droplets surface is in thermal equilibrium among
are considered a second fluid phase, spe- external convective heat transfer, conduc-
cific continuity and momentum equations tion, mass transfer (from droplets) and
need to be solved to take into account the phase changes.
droplet drag as source term, which is pro- Except for convective heat transfer,
vided by the UDF. Once both the droplet which is computed by the ANSYS CFD
velocity from the above equation and the solver, thermal flux (conduction through
concentration are determined, the rate of ice, latent heat from transformation from
liquid water impinging the surface is com- liquid to solid, sublimation, and sensible
puted based on the local face normal. heat, which accounts for droplet temper-
The traditional method of determining ice
accretion is to affix artificial simulated ice shapes Engineers validated this customized ature change) was coded within the UDF.
to test aircraft — in this case, DAHER-SOCATA’s approach using ANSYS software for a The theoretical definition used for the
TBM700. series of 2-D cases by comparing the pre- UDF was established after extensive lit-
dicted results with those obtained from erature research to find the most suitable
NASA’s LEWICE, which has been highly model. When the thermodynamic balance

AIRCRAFT ICING
INITIAL GEOMETRY
(Clean aeroshape)
Collection efficiency
NACA12
0.8
Flowfield resolution
0.7 Fluent
LEWICE
0.6
Droplet collection Local heat

efficiency transfer coefficient 0.5
Next time step
Beta
0.4
0.3
Thermodynamic balance
0.2
New ice layer 0.1
0.0
Mesh morphing -0.08 -0.06 -0.05 -0.02 0.00 0.02 0.04
FINAL GEOMETRY Curvilinear abscissa
DAHER-SOCATA’s ice accretion simulation process using ANSYS software Collection efficiency computed by ANSYS Fluent and LEWICE
Cooling by Aerodynamic heating

incoming droplets
Kinetic heating
Convective
heat flux Latent heat
Evaporation/
sublimation
ICE
Conduction
Ice collection efficiency prediction on TBM850 air inlet Thermodynamic balance
is computed by the UDF, engineers can surface that accounts for the complex ice • Every five seconds, the droplet
determine if the incoming water freezes accretion shape. The team currently com- collection is updated through
or runs back, and from this information bines both mesh morphing and smooth- a β computation.
can compute a new ice thickness. ing as well as local remeshing. Because
mesh modification is time-consuming, • When a new flow solution is available
MESH DEFORMATION the remeshing frequency is a trade-off (every second, in the present
As the thickness of the ice changes, the between computation time and accuracy. example), the computation of the
aerodynamic shape is modified and the The mesh modifications were performed thermodynamic balance provides
aerodynamic flowfield is recomputed. using the moving deforming mesh model a new ice thickness.
Instead of creating a new mesh at each in Fluent.
time step, the team morphs the exist- • Using the newly computed ice
ing mesh from the previous time step. COUPLED SIMULATION thickness, the mesh is morphed and
Boundary nodes are moved normal to the STRATEGY smoothed every five seconds.
wall, with a distance corresponding to ice Using the different tools described above,
thickness as computed using the thermo- DAHER-SOCATA has established a com- The above sequence is repeated through-
dynamic model described above. plete unsteady simulation strategy: out the full required simulation time. This
Since the final result strongly depends chronology was developed specifically for
on the flowfield accuracy, mesh quality • The flow solution is updated every DAHER-SOCATA’s TBM850 and its flight
is a key parameter. The challenge is to second using a 10 time-step unsteady envelope. It may need to be adjusted if
maintain a suitably fine grid close to the computation. deployed on another application.

When the team compared an aca- CONCLUSION in this way should minimize the size and
demic test case (NACA0012) using ANSYS This methodology to simulate ice accretion the amount of the de-iced zone, leading
Fluent to LEWICE, results showed very using Fluent tools — in final development directly to weight and cost reduction for
good agreement. Some local discrepancy before deployment at DAHER-SOCATA — is this aircraft and eventually to aircraft per-
appeared after 200 seconds of simulation, very promising. Comparisons with icing formance improvement.
so improvement of the thermodynamic tunnel test data deliver increased con-
balance model will be performed to rem- fidence that this 3-D method is reliable Reference
edy this using LEWICE as well as experi- for use in future aircraft development. [1] Federal Aviation Administration Advisory
mental results. DAHER-SOCATA is currently investigating Circular 23.1419-2D on Certification of Part 23
In recent years, certification require- how to optimize the TBM850 de-icing sys- Airplanes for Flight in Icing Conditions and 20-73A
ments regarding icing hazards have tem. The company now has the capabil- on Aircraft Ice Protection.
become more and more demanding, mak- ity to identify critical ice accretion zones
ing aircraft icing certification more chal- early in the design process, which should
lenging. Using the droplet-collection UDF lead to a significant development cost
combined with the unsteady solver and reduction by limiting the tests required
MDM, ANSYS Fluent is a consistent global for certification (both wind tunnel and
tool for ice accretion prediction. flight) to a strict minimum. Using CFD
Grid morphing for several time steps
t=0s t=30s t=60s t=90s t=120s
t=150s t=180s t=210s t=240s
Simulation chronology developed specifically for DAHER-SOCATA’s TBM850
TIME(s) 1 2 3 4 5 6 7 8 9 10 ...
Flow computation X X X X X X X X X X
Beta computation X X
Thermodynamic balance X X X X X X X X X X
Morphing X X X X X X X X X X
Remeshing X X
Comparisons with icing

tunnel test data deliver
increased confidence in
this 3-D method for reliable
use in future aircraft
development.
Ice formation prediction comparison using LEWICE and ANSYS Fluent

ROCKET ENGINES
Testing
the Next
Generation
of Rockets
Structural analysis provides
additional thrust for analyzing
rocket engine test equipment.
By Vladimir Tkach, Head of Stress Engineering, Alexander Milov, Lead Stress Engineer
Alexander Loshkarev, Lead Stress Engineer, and Denis Merzljakov, Lead Stress Engineer
NPO Energomash, Khimki, Russia
Historic Vostock rocket engines

S
pace exploration is entering an exciting new era. With
active national space programs accelerating in India and
China and commercial entrants such as SpaceX showing
success, traditional space-faring nations can see that the race is
on to provide greater payloads at competitive costs.
Located northwest of Moscow, NPO Energomash has a pio-
neering history of successful research, development and deploy-
ment of liquid-propellant rocket engines (LPREs). The company’s
technology has powered launch vehicles since the dawn of the
space age. These include the rockets that carried the first arti-
Bisected view of force measurement system configuration of rocket
ficial satellite, Sputnik 1, in 1957; the first human space flight engine test stand, which shows load-bearing frames, rods, pins, belts,
aboard Vostok 1 in 1961; and the long-serving Soyuz vehicles springs and sensors
that are still in service today for transporting crew and cargo to
the International Space Station. NPO Energomash develops sev-
eral classes of engines, including the RD-170 and variants that
provide the highest thrust of any engine ever used on space-
launch vehicles.
The modern RD-171 engine is a four-chamber LPRE that
burns kerosene and liquid oxygen to power the first-stage strap-
on boosters for the Zenit category of vehicles for land- and
sea-based launches. The engine can provide up to 800 metric
tons-force (1.75 million pounds-force) of thrust, helping the
Zenit achieve a lower cost of payload weight per launch relative
to other types of launch vehicles. Future heavy-lift launch vehi-
cles, however, will require the ability to lift even greater masses
into low-Earth orbit and beyond.
ANSYS Mechanical model of NPO Energomash LPRE test bench. Engine
To meet these future needs, NPO Energomash has been (not shown) is represented at left by CERIG rigid connections for its center
designing a more powerful LPRE that will generate thrust of of gravity and four nozzles. SHELL181 elements (grey) represent the largest
1,000 metric tons-force (2.2 million pounds-force). Rather than load-bearing frames, with BEAM188 and LINK8 elements used for the smaller
frame legs, rods, pins and sensors.
spend millions of dollars to purchase and construct a new test
facility, the organization determined that it was more reason-
able to analyze the existing testing infrastructure; the company
had a horizontal test bench force measurement system that was
designed for the RD-171. Before test-firing a new prototype, stress
engineers there needed to evaluate the ability of the current test
bench to withstand forces generated by a higher thrust. The team
used ANSYS Mechanical APDL to perform a static strength analy-
sis of the test bench system.
For the purpose of the structural simulation, the test bench
was divided into many different load-bearing frames, belts, rods,
bolts and pins to determine whether the structures would meet
safety factors required by the Russian Federal Space Agency
(Roscosmos) when loaded by the more powerful engine. Two
Geometry of the frame connected to the ring permanently fixed to the
such loading conditions are emergency thrusting at 105 percent ground shows how the team used symmetry during the modeling process
of nominal (1,050 metric-tons force) and inertial back-blow, or to reduce the simulation domain by a factor of six.
recoil, which occurs at engine cutoff. The stationary part of the
bench consists primarily of the two largest frames, which are
Before test-firing a new prototype, NPO Energomash

stress engineers needed to evaluate the test bench’s
ability to withstand the forces that would be generated.

ROCKET ENGINES
connected by pins to a load-carrying ring permanently fixed to

the ground. The movable part of the bench makes up the remain-
ing frames, belts, rods and bolts, and is connected to the station-
ary part by springs and force sensor systems. These connectors
allow the movable part to slide along the stationary part in the
horizontal (x) direction.
The NPO Energomash team used SHELL181 elements to vir-
tually represent the largest frame parts and BEAM188 elements
for smaller frames and frame legs. LINK8 rod elements, which
allow only tension or compression, represented the engine frame,
belts, sensors, pins and most of the rods. The engineers used
CERIG rigid connections to represent the engine’s center of grav-
ity and the centers of the four nozzles that are connected by gim-
bals to the engine frame. From experience, the team knew that
The full thrust from the engine uniformly loads the center of the frame.
the highest strains and deformations are localized in areas where Using symmetry, one-sixth of the load impacts the area shaded in purple
the frames are welded, so it was logical to evaluate part strength (top). Since the adjacent frame (not shown) is stationary, an equal reaction
based on these regions. Since the weld seam metal is weaker than force is applied back to the flange (bottom, purple). The flange deflects
upon loading, so this load is shown increasing linearly with the height
the steel sheets that comprise the frames, conservatism dictated of the flange.
that two variants be used during the structural analysis: the first
with the frame modeled using steel properties and the second
using weld seam metal properties.
Using a two-step simulation process, the team first performed
stress analyses assuming a linear elastic state. After detecting
the regions with the highest stress concentrations, engineers per-
formed mesh refinement studies and then ran nonlinear elastic–
plastic calculations of stress–strain behavior in these areas.
Taking advantage of symmetry of several load-bearing parts
helped to reduce some of the computational expense. Using a
desktop machine with eight cores, a typical linear static analysis
could be completed in five minutes or less, while a typical non-
linear analysis accounting for material plasticity ran in under
30 minutes. Overall, the engineers spent three months perform-
ing the calculations and preparing reports for approximately 100
individual simulations.
Predictions using ANSYS structural mechanics software
Results of elastic–plastic analysis showing displacements (left) and
showed that all of the test bench parts studied would with- equivalent stresses (right) on the frame. Frame material is simulated as
stand the 1,050 metric tons-force of engine thrust and sub- weld seam metal. High stresses on the flange area (bottom right) lead to
sequent recoil forces at engine cutoff. However, many of the displacements that are shown compared to the original position of the frame.
parts did not meet the Roscosmos-mandated reserve factors

requiring a load capacity of 50 percent to 60 percent above the
maximum expected design load. NPO Energomash engineers
used this information to design reinforcement modifications
for specific frames, rods, bolts and pins in question. By analyz-
ing many potential reinforcement modifications, the team is
ANSYS technology is a
critical tool for modern
design engineers who
know the value of time,
money and prestige
to their organizations. Representation of a probable method for reinforcing the frame to increase
safety factor. Modifications are shown in red and yellow for the lower section
of the full frame from front (left), back (center) and side (right) views.

Displacement results of original (left) and modified (right) frames, Equivalent stress results for flange at the bottom of original (left) and
showing noticeable reduction in the new design modified (right) frames. Reinforcing fins are shown to provide increased
strength to this part of the frame.
optimizing the prototype test bench to be used for a future test- Simulation results gave bench
firing of the new LPRE.
designers confidence that
From the company’s point of view, the technology offered by
ANSYS is a critical tool for modern design engineers who know
their force measurement
the value of time, money and prestige to their organizations. The system could meet the needs
simulation results thus gave bench designers confidence that of testing the next generation
their force measurement system could meet the needs of testing of rocket engines.
the next generation of rocket engines.
NPO Energomash is supported in this work by CADFEM CIS.
Comprehensive High Performance Computing
Consolidation using Remote Visualization

Optimized HPC solutions for your Enterprise
Gompute On Demand - The CAE Cloud solution
Private, Public and Hybrid HPC Cloud computing
Discover GOMPUTE - visit www.gompute.com

WING DESIGN
Mesh morphing reduces the time required

to optimize an aircraft wing.
By Marco Evangelos Biancolini, Researcher, and Ubaldo Cella,

Research Partner, University of Rome Tor Vergata, Rome, Italy
Giorgio Travostino, CFD Manager, and Michele Mancini,
Aerodynamics Engineer, Piaggio Aero Industries, Genoa, Italy

M The morphing
ost business aircraft manufac- modifying the geometry, recreating the
turers use computational fluid mesh and running the modified wing
dynamics (CFD) and wind tun- geometry. The new method uses RBF operation can
nel testing to evaluate wing designs one
iteration at a time. However, with this
Morph software — an add-on to the ANSYS
Fluent CFD solver — to change the wing be executed in a
method, optimizing a design typically shape and surrounding computational matter of seconds,
even on very
runs into many months — time that lead- mesh to solve a series of design points
ing manufacturers try to reduce. To speed without having to manually create new
up the optimization process, Piaggio Aero
Industries, S.p.A., teamed with research-
geometry and mesh. Researchers defined
the sweep angles of the two leading-edge
large meshes,
ers at the University of Rome Tor Vergata sections as shape parameters that form by using the
to validate a new design optimization
method that generates a single mesh
the basis of the design space.
Engineers employed an advanced parallel-processing
as a starting point and morphs it to any approach to carefully control wing shapes capabilities of
high-performance
new geometry studied. The morpher tool while maximizing the quality of the vol-
RBF Morph allows engineers to change ume mesh after morphing. Surface morph
the locations of nodes in the computa-
tional mesh to alter wing shape, then it
ing was used to control sweep angles by
imposing a translation in the flight direc-
computing
matches the surrounding surface and vol- tion to the external wing tip while con- clusters.
ume mesh to the new shape. Researchers straining the kink area. The region altered
applied ANSYS DesignXplorer to drive the by the morpher was defined by a bound-
morpher to map out the complete design ing box with constrained boundary. The
space, identify the optimal design and morpher changed the locations of nodes
validate its robustness in just a few weeks. in the computational mesh originally cre-
Piaggio Aero produces business, ated using ICEM CFD to adjust the wing’s
transport and military aircraft such as the shape and to match the surrounding sur-
P180 Avanti business plane, which is the face and volume mesh to the new shape.
fastest turboprop available for commer- The morpher applies a series of
cial use. In the past, the company’s engi- radial basis functions (RBFs) to produce
neers had optimized wing efficiency for a solution for the mesh movement using
a new business aircraft using traditional source-point inputs and their displace-
design methods. The goal of the efficiency ments (two shape parameters in this
study was to maximize the lift-to-drag case). Computationally, RBFs are very
(L/D) ratio — the amount of lift generated lightweight compared to storing all the
by the wing divided by the drag it creates meshes that are created. The new shape
when moving through the air. A higher is applied within the solver without need-
L/D leads directly to better fuel economy, ing to reload the mesh, and the Fluent
climb performance and glide ratio. The solver continues from a previously con-
wing’s leading edge is divided into two verged solution. The winning set of shape
parts, each with a different sweep angle parameters can be selected and applied
(defined as the angle between the leading to the geometry source to morph the orig-
edge of the wing and a line perpendicular inal model. The morpher incorporates a
to the axis of the aircraft fuselage). A kink volume mesh smoother that preserves
wing section is located between the two volume mesh quality during morphing.
leading-edge parts to smooth out the dis- The morphing operation can be executed
Business aircraft wings were originally
continuity between them. The key design in a matter of seconds, even on very large designed using a combination of CFD simulation
variables are the sweep angles of each meshes, by using the parallel-processing and wind tunnel testing.
leading-edge part. As these sweep angles capabilities of high-performance comput-
are changed, the kink wing section must ing (HPC) clusters.
also be adjusted
PHOTO COURTESY PIAGGIO AERO INDUSTRIES.
AUTOMATING THE CFD MODEL the number of solver iterations needed

MORPHER CHANGES The Fluent model was automated using for all design points past the first. The
MESH SHAPE the ANSYS DesignXplorer design opti- baseline model was iterated for about
To validate the new method, University mizer to update its shape at initialization 1,500 iterations, while each subsequent
of Rome researchers began by using and its converged flow solution at base- run was resumed from previously con-
ANSYS ICEM CFD to produce a hex mesh line. The final solution data set from the verged solutions without memory loss, so
of the wing with 14 million cells. The tra- baseline case was used to initialize sim- they required only about 500 iterations
ditional approach involves solving the ulation of each subsequent design point. for full convergence. Drag and lift were
CFD model, evaluating the results, then Hence, the method significantly reduced exported as output parameters. Design

WING DESIGN
CFD mesh of the wing The morpher altered the

shows two key design locations of nodes in the mesh to
parameters: sweep angles of change the shape of the wing and match
the two wing parts. the surrounding mesh to the new shape.
of experiments (DOE) identified a small response surface enabled the researchers

sample of design points to represent the to visualize the wing’s aerodynamic per-
design space in such a way that when formance over the entire design space and
the aerodynamic performance was calcu- intuitively understand how output vari-
lated at this small set of points, it could ables, such as drag, are dependent upon
be interpolated with DesignXplorer to the chosen design parameters.
predict performance of any other design By evaluating the complete design
point within the design space with min- space, the optimization showed that a
imal error. number of points in the design space
The ANSYS Workbench platform pro- delivered a 1 percent improvement in
vided seamless interconnections for L/D. The design optimization took a cou-
the interapplication data transfer, and ple of weeks, less than a tenth of the time
The automation platform triggered the
morpher to morph the shape of the wing and to DesignXplorer sequentially simulated all required to optimize the design using con-
solve the new shape, then passed the results to of the design points and collated outputs ventional methods. Solving the complete
ANSYS DesignXplorer. from the simulation. When the “update design space made it possible to evalu-
all design points” button was selected, ate robustness of the various design can-
the platform sent the first set of param- didates. The designs that are most robust
eter values representing the first shape are those in regions of the response sur-
to the CFD solver. This triggered the mor- face with the least slope. Researchers
pher to morph the shape to match the identified an optimal design that delivered
new shape being explored. The CFD solver consistent performance even as design
then solved this new shape and passed parameters were varied over a 25 percent
the output results, such as drag force, to range. Piaggio plans to use these design
DesignXplorer, where the results were optimization methods to reduce time to
stored. The automation platform then market and wind tunnel testing expenses
sent the next set of parameter values to for next-generation aircraft.
The response surface made it easy to visualize
the wing’s performance over the design space. the morpher. The process continued until
CFD solutions for all the design points
were completed. The design
GENERATION OF RESPONSE
optimization took
SURFACE a couple of weeks,
less than a tenth of
ANSYS DesignXplorer created a response
surface using Kriging regression anal-
ysis from the collected data set of drag the time required to
and other aerodynamic forces. The multi-
objective genetic algorithm (MOGA) opti- optimize the design
mization method was used to define
design goals and constraints, to cre-
using conventional
The MOGA algorithm within ANSYS
DesignXplorer was used to simultaneously ate new design points, and to evaluate methods.
optimize lift and drag over the design space. their performance based on the response
surface. DesignXplorer offers direct
and hybrid optimization methods. The

ANALYSIS TOOLS
Flight
Simulator
Aeroelastic Prediction Workshop
assesses computational methods for
predicting unsteady flow fields.
By Thorsten Hansen, Customer Support, ANSYS, Inc.
T The designer
he early stages of aerodynamic high cost of maintaining these codes, the
design typically are based on the industry is actively exploring the develop-
assumption that the mechan- ment of new techniques. must take into
ical structures are rigid, a concept that
greatly simplifies the design process.
To address this challenge and assess
state-of-the-art methods and tools for the account the
As the process progresses, the designer prediction and assessment of aeroelastic elasticity of
must take into account the elasticity
of the mechanical structures, because
phenomena, the American Institute of
Aeronautics and Astronauts hosts the the mechanical
this often has a major impact on aero- Aeroelastic Prediction Workshop as a con- structures,
dynamic performance and structural
integrity. Aeroelasticity refers to the inter-
tinuing education event. It provides an
impartial forum to evaluate the effective- because this
action between the aerodynamic flow ness of current simulation tool codes and often has a
major impact
field and the mechanical structure. It is modeling techniques as well as to deter-
one of the most challenging aspects of mine whether or not they are able to
designing aerospace structures because accurately simulate challenging nonlin-
on aerodynamic
performance
it’s difficult to model the tightly coupled ear aeroelastic problems involving vor-
multiple physics involved in the phe- tices, shock waves and separated flow
nomenon. Traditionally, companies have that result from aeroelasticity. In prep-
and structural
integrity.
developed in-house methods that use cus- aration for the 2012 workshop, ANSYS
tom aeroelasticity codes. However, with used ANSYS Fluent and ANSYS CFX com-
the relentless demand for higher-fidelity putational fluid dynamics (CFD) software
information to support improved aero to solve the problems, then presented
dynamic efficiency, coupled with the the results at the meeting. A select set of

ANALYSIS TOOLS
results is illustrated in this article. The full data set is available

online (see endnote).
The Aeroelastic Prediction Workshop challenges the simu
lation community to apply best practices to predict unsteady
aerodynamics characteristics for rigid or weakly coupled aero-
elastic systems. The test configurations were selected to increase
in complexity, from fully turbulent with attached flow and weak
shocks to transient separation conditions with strong shocks and
significant interactions between flow features. For each configu-
ration, participants were asked to provide a convergence study,
steady-state analysis, and time-accurate response to forced oscil-
lations using best practice guidelines to quantify numerical and
model errors. The simulation results were compared with wind
tunnel test data. Aeroelasticity problems have a high computa-
Comparison of
tional intensity, so the ANSYS team used high-performance com- ANSYS Fluent CFD
puting (HPC) platforms with up to 2,048 cores on a Cray XE6™ to results to wind tunnel
solve the workshop challenges. experiment for RSW
dynamic test case at
10 Hz
RECTANGULAR SUPERCRITICAL WING
Rectangular
The first test configuration, called the rectangular supercritical supercritical wing in
wing (RSW), is a structurally rigid rectangular planform wing wind tunnel
that oscillates at a designated pitch amplitude and frequency. PHOTO COURTESY NASA.
The wing is mounted to a small splitter plate that is offset 6

inches from the wind tunnel wall. The RSW was originally tested
in the NASA Langley Transonic Dynamics Tunnel in 1983. The
study was performed at Mach 0.825 with a 4 million Reynolds
number. Steady-state data was obtained with the model held initial conditions and displacing the mesh to match the harmonic
at a fixed angle of attack. Separate dynamic data was acquired wing motion. Simulation showed this test configuration to have
by oscillating the model in a pitching motion about the 46 per- a very complex flow field driven by the shock-induced separated
cent chord line. Forced pitch oscillation frequencies of 10 Hz and flow. The dynamic cases were harder to predict, and some varia-
20 Hz were used in the two dynamic test cases. tions were seen between test results and simulation.
The RSW was meshed with ANSYS ICEM CFD using hexahedral
elements and a scalable grid that provides consistent mesh qual- HIGH REYNOLDS NUMBER AEROSTRUCTURAL
ity after grid refinement. This application appeared to be quite DYNAMICS MODEL
simple at first glance. However, the simulation conducted with The high Reynolds number aerostructural dynamics (HIRENASD)
both Fluent and CFX revealed modeling complications because test configuration is a 34-degree aft-swept, tapered clean wing
the splitter plate and wing model were enveloped in the bound- with a BAC 3-11 supercritical airfoil profile. The semi-span test
ary layer. Despite that fact, both the static and dynamic simula- article is mounted to the ceiling using a noncontacting fuselage
tion results showed general agreement with experimental data fairing connected to a turntable, balance and excitation system.
at higher span locations where the influence of the wind tunnel The HIRENASD wing is geometrically more complex and has a
boundary layer and splitter blade is low. small amount of structural flexibility that is used to oscillate the
wing in its structural modes to acquire unsteady aerodynamic
BENCHMARK SUPERCRITICAL WING data. The HIRENASD was excited at the second bending mode fre-
The benchmark supercritical wing (BSCW) is another structur- quency at approximately 80 Hz forces applied using piezoelectric
ally rigid rectangular planform wing that oscillates at a desig- stacks in the mounting hardware.
nated pitch amplitude and frequency. The BSCW has a NASA The structural flexibility in this test configuration increased
SC(2)-0414 airfoil. All data in this study was obtained at Mach the complexity of the simulation challenge. ANSYS Mechanical
0.85 with a dynamic pressure of 200 psi, fixing the Reynolds was used to simulate the structure and calculate mode shapes.
number at 4.49 million based on the wing chord. Dynamic data Static aeroelastic equilibrium results predicted by the CFX simu-
was obtained for the BSCW by oscillating the model in a pitch- lation closely matched experimental values. The dynamic cases
ing motion about the 30 percent chord. Steady-state information were solved by using the converged steady-state solution as the
was calculated as the mean value of oscillatory time histories. initial conditions and displacing the mesh to match the har-
This configuration was chosen because the experimental data monic wing motion predicted by structural finite element anal-
shows highly nonlinear unsteady behavior, particularly shock- ysis (FEA). Agreement was again good at the mid-span cross
separated transient flow. sections in the dynamic test cases.
Pressure coefficient results determined by the CFX solver show ANSYS software performed consistently in all three test
a significant influence on the turbulence model for shock predic- cases. The results with CFX and Fluent converged rapidly to
tion on the upper surface for the static solution. The dynamic sim- a single solution, both as a function of grid density and time
ulation was solved with a converged steady-state solution as the interval. Results were validated by running a grid and time

ANSYS CFX turbulence model error for BSCW at steady-state conditions Comparison of static aeroelastic equilibrium simulation predictions to
experimental measurements for HIRENASD
Fluent calculations Simulation

for BSCW show good predictions of
fit with experimental unsteady-state
measurements. Splitter Plate coefficient of pressure
for HIRENASD
BSCW correlate well to wind
tunnel experiments.
Fairing
BSCW setup for HIRENASD is 3-D

wind tunnel testing aeroelastic wing with
PHOTO COURTESY NASA. generic fuselage
model. PHOTO COURTESY
RWTH AACHEN.
step sensitivity study. Despite the complexity of the flow fields

in these test configurations, simulation results in most cases
showed good agreement with wind tunnel tests. As expected
with problems of this difficulty, some deviations were identi-
fied in the simulation results, and these are being investigated.
ANSYS continues work on improving capabilities to solve
these and other aeroelastic simulations. Publications sum-
marizing the data from the 2012 Aeroelastic Prediction
Workshop are being prepared for conferences in 2013. ANSYS Simulation results in most cases
is already planning to participate in the next Aeroelastic
Prediction Workshop and will use ANSYS Workbench
showed good agreement with
to seamlessly couple the CFD and FEA simulations to wind tunnel tests.
more accurately simulate aeroelastic behavior.
Thanks to Cray for the high-performance computing resources used by ANSYS Reference
for this project. Aeroelastic Prediction Workshop, https://c3.nasa.gov/dashlink/projects/47/

THOUGHT LEADER Simulation@work
A SYSTEMATIC
APPROACH
Oil and gas leader FMC Technologies is making a science out of systems-level simulation. Multiphysics Simulation
Manager Ed Marotta discusses the company’s unique approach — which includes certification for analysts and best-
practice sharing that spans the globe.
By ANSYS Advantage staff
W
ith a reputation for
innovation in the
oil and gas industry,
FMC Technologies
designs, manufac-
tures and services
technologically sophisticated systems
and products, such as subsea production
and processing systems, surface well-
head systems, high-pressure fluid control
equipment, measurement solutions, and
marine loading systems. The organiza-
tion was recently named by Forbes mag-
azine as one of the Ten Most Innovative
Companies in America. With 27 produc-
tion facilities in 16 countries, FMC has
more than 16,000 employees around
the world.
In keeping with its strong focus on
innovation, FMC operates three tech cen-
ters in the United States, Norway and and director of the freshman engineering physics and their interactions. So we
Brazil that leverage corporate knowl- program there. Marotta spoke with ANSYS formed a team of highly skilled analysts
edge to develop smarter product and Advantage about bringing a disciplined to look at very complex problems related
systems designs. Based at the U.S. Tech approach to engineering simulation at specifically to ocean environments.
Center in Houston, Ed Marotta directs one of the world’s leading innovators in
FMC’s Multiphysics Simulation Group. oil and gas technologies. How has this multiphysics approach
This team was formed in 2010 to max- helped FMC to emerge as a leader in
imize the impact of systems-level multi- What led FMC to create a team specif- systems-level simulation?
physics simulations at FMC, enabling the ically focused on multiphysics studies? Just as you cannot optimize overall per-
company to more quickly and efficiently Here at the U.S. Tech Center in Houston, formance by looking at a single physical
advance its products and technologies by we focus on optimizing energy production force, you need to consider many compo-
rapidly modeling, verifying and introduc- technologies for both subsea and ocean- nents to optimize an entire system. One of
ing industry-changing innovations. surface environments. Obviously, there our most critical systems in recovering oil
Marotta is eminently qualified to are many physical forces at work in these is the tree — an assembly of valves, pip-
lead this multiphysics effort, with a environments. We must consider external ing, spools and fittings that control flows
B.S. degree in chemistry, M.S. and Ph.D. factors such as water temperatures, sub- and pressures. The tree incorporates a
degrees in mechanical engineering, and sea ocean currents, hydrostatic pressures multitude of components that are sub-
post-graduate studies in chemical engi- and fluid–structure interactions — as ject to a range of structural, thermal and
neering. Prior to joining FMC, he was well as internal electromagnetics and fluidic phenomena. We have to consider
director of Texas A&M University’s ther- fluid dynamics within our equipment. varying operating pressures and temper-
mal conduction laboratory as well as It’s not enough to consider just one force; atures. If we looked only at one compo-
associate research and teaching professor we need to look at the impact of multiple nent in isolation, we would not be able to

Just as you cannot
optimize overall
performance by
looking at a single
physical force, you
need to consider
many components
to optimize an
entire system.
predict the performance impact of the

entire tree as we make design changes.
Instead, our analysts have the capability
to attach new components to the tree, to
make design modifications — for exam-
ple, to the insulation system — and then
to conduct systems-level simulations.
An example of this is simulating an
CFD model of tree under full environmental conditions
entire production tree to maximize ther-
mal insulation and slow down cooling.
When an oil recovery system has to be
shut down for a weather event or other
contingency, it’s critical to maintain a
warm internal temperature, despite the degree of confidence that our designs expansion to other FMC global engineer-
fact that the tree is submerged in cold will perform as expected in the real ing groups is our goal. In a fraction of
ocean water. By using ANSYS Fluent to world, delivering high levels of quality the time, we can arrive at an engineering
conduct computational fluid dynamics and safety. Having a clear understand- solution that comes within +/− 20 percent
(CFD) investigations of our subsea trees ing of functional and performance speci- of our high-fidelity models. This allows
and manifolds for cool-down predictions, fications is paramount to achieving high us to run a what-if scenario very quickly
we can make design changes that help to quality in our simulations. and cost effectively, so we can rule out the
mitigate hydrate formation, which would In terms of innovation, we have a bad design possibilities. For the designs
compromise operational performance of unique initiative called “compact model- that make the cut at the compact-model
the equipment. We can look at the ther- ing” that attempts to streamline the earli- stage, we then move on to higher-fidelity
mal contribution of each component est stages of design, allowing us to move simulations and higher computational
in isolation as well as the performance forward very rapidly. We have a strategic loads. We believe this compact-modeling
of the entire system. We can customize agreement with ANSYS that has enabled approach will allow us to introduce
water currents, boundary conditions and us to leverage special engineering simu- groundbreaking new technologies and
thermophysical properties for customer- lation software that we hope to eventually advanced state-of-the-art products very
specific sites. ANSYS software even enables run on an iPad® or some similar mobile quickly and efficiently.
us to simulate two-phase mixtures. device. These tools will be unique to the
Multiphysics Simulation Group; however,
What are the specific engineering
pressures in your industry — and how is
FMC responding? We are creating a high degree of confidence
There are three concerns driving the
industry right now: safety, quality and that our designs will perform as expected
innovation. We’re addressing the safety
and quality issues by focusing on robust
in the real world, delivering high levels
design at the systems level, as we’ve of quality and safety.
already discussed. By constructing
numerically large, complex simulations
of entire systems, we are creating a high

THOUGHT LEADER
Why is engineering simulation so are solving. We believe this is essential engineers have master’s and Ph.D.
important to your work at FMC? to ensuring the integrity of our simula- degrees, which means that they have used
Here in Houston, we simply can’t build tion results. ANSYS software in academic settings.
physical prototypes or run systems test- Here in the Multiphysics Simulation Most of our customers also use ANSYS
ing in a water tank. It would be prohibi- Group, we’ve created an analyst certi- software. By collaborating closely with
tively expensive to recreate conditions at fication program that ensures that our ANSYS, we believe that we are getting the
the depth of 10,000 feet (3,000 meters) engineers are well trained in the use of best of both worlds: We are employing
of ocean water. So obviously, we have to simulation software. Our goal is to work the most widely used simulation toolkit
rely on engineering simulation, and our with our global analysis teams to expand in our industry, but we are applying it in a
reliance on simulation keeps growing as that certification program to other sites. very customized way that sets FMC apart.
innovation becomes more and more crit- Working with ANSYS, we have developed We work in a very competitive industry,
ical. The same is true for our other global customized internal training classes for and we have great confidence that ANSYS
engineering centers, which are tackling our team, and we also seek out exter- will help us build and maintain our engi-
different but just as complex engineer- nal educational opportunities. All of our neering leadership.
ing challenges. efforts are focused on making simulation
an exact science at FMC, ensuring that our
Even though global teams are work- analysts have the right skill set and stan- We work in a
very competitive
ing on different problems, do you also dardizing our global analysis processes.
collaborate and share knowledge? This allows us to not only arrive at inno-
Collaboration is one of our core concepts
at FMC. It is important to emphasize that
vations rapidly, but also to have a very
high degree of confidence in our results. industry, and
globally we have a very large pool of
extremely talented analysts in Norway
In addition, the Multiphysics Simulation
Group has an internal engineering ini-
we have great
(Asker and Kongsberg), Brazil (Rio de tiative called the Smarter Design Space, confidence that
ANSYS will help
Janiero), Singapore, India (Hydrabad) focused on bringing all of our engi-
and Scotland (Dunfermline) who collabo- neers and analysts together to optimize
rate on a daily basis on our most complex
and pressing engineering problems. FMC
our design in virtual space and improve
the accuracy of our results, backed by a us build and
has well over 100 analysts with advanced
degrees who share knowledge and best
high-performance computing cluster and
shared software tools.
maintain our
practices to ensure that the most accu- engineering
leadership.
rate analysis is achieved. We share this How would you describe your rela-
information globally through an internal tionship with ANSYS?
online forum called “The Edge,” through In the past four years, the Multiphysics
which our engineers and analysts can Simulation Group has grown from two
ask and answer questions globally. We full-time simulation analysts to a team
have specific global design guidelines of 11 engineers. ANSYS has been crucial
that make certain everyone is perform- in supporting this growth by providing
ing analyses and deploying ANSYS soft- the required training, technical support
ware in the same way, no matter where and customized tools, such as those for
they are or what specific problem they compact modeling. The majority of our
brazil Structural simulation

of a gimbal that gives
Identifying Stress Points for support to risers with
Even Load Distribution rubber–steel pad
FMC engineers in Brazil are conducting structural analyses on a

gimbal system — which is used to reduce shock to protect criti-
cal sections of piping along with the module that boosts flow, by
accommodating the roll of the rig in the marine environment.
By using ANSYS Mechanical to conduct structural analysis, the
engineers identify areas of stress and ensure that loads are not
transferred to piping.

norway
Ensuring Pipeline Integrity During
Pressure Fluctuations
Flow assurance engineers at FMC Norway

use ANSYS CFD software to analyze
vibration-related issues caused by inter-
nal flows inside subsea piping using the
Reynolds stress model. To predict fluid
Frequency fluctuations downstream of single pipe bend
forces on the pipe structure, wall pressure
fluctuations of high Reynolds number
multiphase flows are determined via com-
putational fluid dynamics simulations. model. FMC engineers correctly predicted both identify and address potential flow-
For single-phase and multiphase flows, the frequency of the wall-pressure fluc- induced vibration issues for a new subsea
FMC analysts recently used ANSYS to per- tuations downstream of a single pipe production system.
form simulations with Reynolds numbers bend. Comparing this input to the natu-
up to 1 million, using the SST turbulence ral frequencies of the piping helps FMC to
singapore field development. The goal was to pre- challenging, as the complex geometry of
Defining Erosion Allowances Inside dict erosion rates in the tree to verify that the tree meant that fluid flows were unsta-
a Recovery Tree proposed erosion allowances in piping ble in certain sections. However, by using
and fittings were sufficient. This is vitally ANSYS CFX, engineers in Singapore accu-
The FMC flow assurance team in Singapore important, because insufficient erosion rately predicted fluid behavior — and
recently conducted a 3-D CFD erosion anal- allowances could lead to a breach of con- identified where increased erosion allow-
ysis for a subsea recovery tree for a gas tainment. This simulation was especially ances were required.
Analyzing erosion patterns within recovery tree section
WEB EXCLUSIVES
THE NETHERLANDS BRAZIL UNITED STATES
Perfecting Gas–Liquid Phase Separation Optimizing Steady-State and Transient Predicting Long-Term Fatigue in Piping
in Inline Separators Thermal Performance
These additional examples from FMC Technologies are available at www.ansys.com/exclusives/magazine.

INDUSTRIAL EQUIPMENT
MAGNETIC
ATTRACTION
Simulation aids a revolutionary magnetizing machine to produce magnets with precision-tailored magnetic fields,
forces and behaviors.
By Scott Evans, Director, Product Development and R&D Programs, and

Ron Jewell, Vice President, Sales and Marketing, Correlated Magnetics Research, Huntsville, U.S.A.
M
agnets have a large Air gap
number of commer-
cial applications in
a variety of devices,
such as clutches, bear-
ings, gears, fasteners,
motors, sensors, security devices, point-
ers, scopes and optics, to name a few. The
Load shaft Drive shaft
material in conventional high-field mag-
nets is typically oriented, or magnetized,
in a single direction, a condition that
results in fields that are far from ideal
for many electromechanical applications.
Magnet
Manufacturers of magnetic devices have support
been addressing this limitation by creat- structure
ing complex magnetization fixtures or by Magnetic
assembling groups of smaller magnets. material
Both of these methods increase fabrica-
tion challenges and result in field losses
Typical magnetic linkage
in the gaps between magnetic pieces.
The latest advancement in this field,
correlated magnetics, involves precise
and rapid magnetization of materials a
small volume at a time, making it pos- the first areas of application is improv- makes it difficult to keep shafts aligned.
sible to optimize the emission of fields ing the design of magnetic clutches. If the shafts go out of alignment, wear on
from magnetic materials — and even to Magnetic clutches are used in equip- bearings and shafts increases substan-
create field profiles that were not previ- ment in the aerospace, marine, med- tially. Ship engineers then must spend a
ously possible. This breakthrough abil- ical and chemical industries, among considerable amount of time maintain-
ity to accurately shape magnetic fields others, and in devices including indus- ing alignment and addressing problems
increases device efficiency, enables sav- trial ovens, pumps, compressors, meter- caused by misalignment.
ings in assembly and integration, and ing devices, controllers and hydraulic During the operation of a typical two-
establishes whole new areas of applica- machines. Magnetic torque-transfer plate magnetic coupling device, torque is
tion for magnets. Furthermore, the tech- devices are often used in applications in transferred by shear forces between mag-
nology permits material size and weight which it is difficult to maintain alignment netic materials interacting across an air
reduction by reducing any overdesign between driving and driven shafts. The gap. Transferring torque without physi-
intended to compensate for alignment devices also play a role replacing dynamic cal contact means that alignment is much
errors in assembly and integration of seals for lower torque applications within less of a concern; the process also elim-
magnetic materials. These correlated chemical or high-pressure processes. In inates wear while providing a barrier to
magnets are created using an innova- the marine industry, one application is vibration, torque overloads, electricity
tive new machine, called a MagPrinter™, connecting electric motors to pumps on and heat. This type of device traditionally
that renders complex magnetic struc- ships. A ship’s hull interacts with waves, uses several sections of uniformly mag-
tures in bulk magnetic materials. One of which causes flexing in the frame and netized material arranged as alternating

Shear force direction
stripes or as pie pieces to form the surfaces facing the air gap.
But the shear forces generated by conventional designs are rel-
atively weak, so conventional magnetic torque transfer devices
must be many times heavier and larger than direct mechanical
couplings to deliver the same torque. In other words, they have
t
a low torque density, which is directly proportional to the shear
S forces between magnets.
Feature size
Correlated Magnetics Research (CMR) is developing multi-
pole magnetic structures that offer higher torque-transfer per-
formance from smaller, lighter magnets. The company is also
Alternating blocks of magnetic material are used to produce magnetic
shear forces. exploring techniques for eliminating the non-torque-producing
tensile forces between the magnets. CMR has two key innova-
tions: the application of signal processing methods and coding
theory to guide the production of complex patterns of magnetic
elements and resulting magnetic fields, and the MagPrinter™
that is capable of rendering very complex magnetic structures
rapidly and inexpensively in bulk magnetic material. The CMR
R&D team found that ANSYS Maxwell electromagnetic field sim-
ulation software played a key role by assisting in the design of
these complex magnetic structures that deliver shear forces
40 percent higher than previously known magnetic structures
and, at the same time, halving magnetic material requirements.
BASIC RESEARCH INTO MAGNETIC LINKAGES

CMR engineers use Maxwell to investigate the shear force gen-
erated by various magnetic structures. For example, in the early
The effect of block width on measured shear stress and density
stages of research and development, the team performed an
investigation to determine how shear force varies with respect to
the aspect ratios — thickness to width — of the magnetic blocks
facing the gap between the two correlated magnetic structures.
The model showed that an aspect ratio of 1:1 yielded the highest
shear force per surface area and shear force per volume of mag-
netic material.
Next, square regions, 96 mm per side, were assembled in
the software from magnetized blocks: 8 mm, 4 mm, 2 mm and
1 mm wide. The thickness of each block was equal to its width;
each substrate had the same area. Simulation showed that with
a zero displacement across the gap between the magnetic blocks,
as the blocks are reduced in width from 8 mm to 1 mm, the shear
Magnetic field emitted by one-sided structure under zero shear condition.
The triangles are magnetic blocks, and the rest of the image is air. The per area decreases modestly and the force per volume of mate-
highest field intensity is orange, and the lowest is dark blue. The lines that rial rises sharply. So, while the highest forces come from larger
travel across the gap illustrate attractive forces. blocks, the best use of magnetic material comes from many
small blocks.
When the simulation was repeated with a small 0.5 mm
gap between the substrates, the results were very different. The
shear force per area falls more rapidly as the blocks are halved
in size; the maximum shear per volume occurs with a 2 mm
block width. Smaller blocks confine the field more strongly near
the surface, so the best use of magnetic material comes when the
blocks have a width that is about four times the gap between the
substrates. Other architectures will have different ratios, but the
team anticipates that the basic relationship will apply. Generally,
by knowing the gap that can be used in a particular application,
engineers can estimate both the best feature sizes for the mag-
Magnetic field emitted by one-sided structure in a maximum-shear
condition. The highest field intensity is orange, and the lowest is dark blue. netic structure and the maximum shear force per magnet vol-
The lines that travel across the gap illustrate attractive forces. ume that is achievable. Maximizing the shear force, in turn,
makes it possible to increase torque density and reduce the size
and weight of a magnetic clutch with a given torque capacity.

INDUSTRIAL EQUIPMENT
WORKING WITH ONE-SIDED DEVELOPING A WORKING

MAGNETS PROTOTYPE
To obtain guidance into scaling and CMR’s MagPrinter includes a unique elec-
design issues in more complex magne- tromagnetic coil that is placed near a mag-
tization patterns, researchers examined netizable material and energized with
magnetic structures that emit magnetic a very short high-current electric pulse.
fields predominantly on one side with The energized coil emits a very intense
several models. An example of a one- magnetic field in a confined region of
sided structure is a Halbach array, which space, magnetizing a small region of the
Two continuously varying substrates aligned
uses a spatially rotating pattern of mag- adjacent material. Each of these small have a large attractive force (top). The pattern
netization to augment the magnetic field regions is called a maxel (much like a in the upper magnet has been shifted to the side
on one side of the array while cancelling pixel, a shortened form of magnetic ele- and illustrates a maximum shear force condition
(bottom).
the field to near zero on the other side. ment). An array of maxels forms a com-
The first model was an array of triangles plex magnetization profile and, likewise,
whose bases face the opposing substrate emits a complex magnetic field. Guided
and that have magnetization vectors by the results of the simulation studies,
that alternate toward and away from CMR conducted a series of experiments
the opposing magnetic structure. The to find the best combination of magnetic
triangles with bases facing away from patterns and pulse energies for producing
the opposing substrate are magnetized shear forces. The knowledge from simula-
to the left and right with respect to the tion efforts and shear force experimenta-
opposing magnetic structure. Magnetic tion was applied to fabricating a magnetic
fields were plotted with the opposing clutch. The magnetization energies and
magnets aligned at zero shear and max- geometry of those codes were adapted to
imum shear conditions. This arrange- create a high-torque maxel code.
ment generated large forces between the The team divided a 3-inch outside
Pattern of tracks printed on magnet to
opposing magnetic structures. When the diameter magnet into tracks, and each produce high-performance magnetic clutch
two sides are aligned, those forces are track was printed to have an alternating
attractive or repulsive. When the two magnetization that approximated the con-
patterns are offset, a shear force is devel- tinuously varying pattern. Magnetization
oped, while attractive or repulsive forces was accomplished in about three min-
are reduced. At a maximum shear force utes with the fourth-generation CMR
condition, the attractive and repulsive MagPrinter machine. The system fea-
forces are zero. tures a computer-controlled platform
CMR examined shear force genera- that moves a magnetic material substrate
tion between magnetic materials that relative to a specialized print head that
had smoothly and continuously varying imparts a focused magnetic field onto
magnetization. This type of magnetiza- the substrate, resulting in a single, well-
tion profile also creates a strongly one- defined maxel at a prescribed location.
sided field emission; this is possible only The fifth-generation machine, which is
theoretically (so far). The shear forces now the standard, performs the same task
generated by smoothly varying magneti- in about 15 seconds. The tracks and their
zation followed the same patterns iden- partner tracks on the opposing magnet
tified with the triangle array described were oriented so that each track reached
above. When two continuously varying a maximum torque condition at the same The MagPrinter creates customized magnetic
substrates are aligned, there is a large orientation of the magnets. materials that can be widely employed in
industry.
attractive force. If the pattern in the upper With two 3-inch magnets, each one-
magnet is shifted to the side, a maximum eighth inch thick, this pattern with-
shear force condition occurs. The simula- stood 27 Newton-meters (N-m) of torque.
tion with smoothly varying magnetization At a gap of 1 mm, this clutch produced
also established an upper limit to shear about 300 N-m per liter of torque den- (SBIR) grant to develop high-torque mag-
forces that can be generated between mag- sity, about half the torque density of a netic gears and clutches utilizing CMR’s
netic materials. In effect, the simulations large-scale planetary gearbox, and triple Polymagnet™ technology. The two-year
using Maxwell help to quantify achiev- that of the highest-performing commer- contract is a collaborative effort between
able performance gains as manufacturing cial magnetic gear, which also has a gap CMR and MagnaDrive, Inc.
enhancements make it feasible to mag- between magnets of about 1 mm. These
netize smaller regions of material and to results helped to convince the Office of
render more complex magnetic patterns Naval Research to award CMR a Phase II
into magnetic materials. Small Business Innovation Research

ENERGY
HOT STUFF
NEM reduces cost and improves efficiency for concentrated solar power generation.
By Ingmar van Dijk, Solar Team, NEM Energy b.v., Leiden, The Netherlands
Power tower research facility in Spain where NEM tests its heliostats PHOTO CREDIT: © PLATAFORMA SOLAR DE ALMERIA / CIEMAT.
A
dvanced technology is
playing an important role
onto a receiver on top of a tower. The dif-
ference between CSP and the more widely A key design
as the world looks for effi-
cient and cost-effective
known photovoltaic form of solar power is
that PV converts sunlight directly to elec- challenge
for NEM is
sources of energy. Solar tricity using the photovoltaic effect, while
energy generation is grow- in CSP, concentrated sunlight is converted
increasing the
ing, especially in sunny areas such as to heat. The heat can be used to directly
Africa, the Middle East, the Mediterranean produce steam, or a heat-transfer fluid
stiffness of
and the southwestern United States. can be used to store some of the heat to
Photovoltaic (PV) energy has been a long- provide a buffer so that steam can be pro-
the mirrors to
time leader in this field, but concentrated duced after the sun goes down. The steam,
solar power (CSP) systems (which use in turn, is used in a conventional turbine
put as much
mirrors or lenses to concentrate a large generator to produce electricity.
area of sunlight onto a small area to drive A key design challenge for NEM is
a heat engine connected to an electrical increasing the stiffness of the mirrors to
power generator) have been around for a
long time and have now started to pick
put as much reflected light as possible
onto the target, called a receiver, with-
reflected light
up steam. The U.S. Department of Energy
(DOE) has offered roughly $5.89 billion
out paying a cost premium. The com-
pany uses ANSYS Mechanical software
as possible
in loans to four CSP projects, an amount
greater than what it has offered to devel-
within the ANSYS Workbench environ-
ment to evaluate the stiffness of large onto the
opers of photovoltaic projects[1].
CSP is experiencing rapid growth,
numbers of heliostat design alterna-
tives. The results are fed into a ray trac- target without
with about 740 MW of global generating
capacity added between 2007 and the end
ing program that determines the energy
generated by the design. This makes it paying a cost
of 2010, bringing the total installed capa-
bility to 1,095 MW. NEM Energy b.v. is
possible to determine the performance-
to-cost ratio of each design alternative premium.
developing a power tower system type of without having to build physical proto-
CSP that uses a field of sun-tracking mir- types. NEM is one of the top five produc-
rors called heliostats to concentrate light ers of steam-generating equipment in

ENERGY
NEM can determine the

performance-to-cost
Receiver
Steam turbine
ratio of each design

G
alternative without
having to build
Condenser
physical prototypes. Solar field Solar tower
Operation of power tower CSP plant shows how heliostats focus light on a
central receiver that transfers heat to a fluid that, in turn, produces steam.
the world, and the company’s experience in this area is a key

advantage in developing CSP.
HELIOSTAT DESIGN CHALLENGES

NEM’s heliostat designs provide two-axis tracking of the sun. One
axis rotates around a vertical pillar, which is supported by a foun-
dation. The other axis rotates around a horizontal tube called
the torque tube, centered on the top of the pillar. NEM’s proto-
type provides approximately 60 square meters of mirror area.
Cost is critical because a single power plant requires approxi-
mately 15,000 heliostats. The goal of heliostat manufacturers is
to reduce the cost of manufacturing while achieving 25 years of
service life and minimizing distortion to maximize energy gen-
eration. According to the Power Technology Roadmap and Cost
Reduction Plan published by Sandia National Laboratories in
2011, the U.S. Department of Energy set a goal to reduce level-
NEM heliostat prototype
ized cost of energy (LCOE) of CSP technology for a hypothetical
100 MW electric plant from today’s costs of approximately
$0.15/kWh to a value of $0.09/kWh or less in 2020. The DOE also
set an objective to reduce heliostat costs from the current base-
line of $200 per square meter to $120 per square meter.
Stiffness is a key factor in accomplishing these goals. Just
a 1-degree rotation error for a heliostat 380 meters away from
the tower results in a 6.6-meter tracking error, meaning that the
reflected light is delivered 6.6 meters from the intended target on
the tower. The structure also must meet building codes to ensure
that it will maintain integrity during storms and earthquakes.
NEM engineers apply ANSYS Mechanical in the ANSYS
Workbench environment to optimize heliostats from perfor-
mance and cost standpoints. The design concept is defined in
CATIA® V5 and imported into ANSYS DesignModeler. The helio-
stat contains 16 mirror segments, or facets.
As these facets are only a few millimeters thick and have large
surface areas, they are very well suited for the use of shell (mesh)
elements. An added benefit of shell elements over solid elements
here is that shell elements can export rotation angles, which are
very important to properly study the deformation behavior of the
heliostat. The rest of the structure is meshed with solid elements.
AUTOMATIC CREATION OF CONTACTS

Finite element model of heliostat created in ANSYS DesignModeler
With 200 different geometric bodies making up the heliostat, it is
an enormous job to manually create contacts, so NEM engineers use
ANSYS Workbench to automatically add more than 1,000 contacts

using a 5-mm tolerance value. They FUTURE PLANS
model bearings as general joints with NEM plans to introduce computational
translations fixed and rotation free. Some fluid dynamics (CFD) with fluid–structure
of the contacts that are automatically cre- interaction (FSI) to model the interaction
ated are removed to allow the bearings to of the structure with the wind. FSI can
rotate. The linear drive is a very complex improve analysis accuracy by applying
part whose internal operation is not rele the properly interpolated force on every
vant to the stiffness of the mirrors; it is node in the structure and by accounting
replaced by a very stiff spring to simplify for the impact of structural deformation
the model and reduce computational time. on wind flow. Furthermore, FSI will sub-
The edges of these mirrors can expe- stantially reduce the need for wind tunnel
rience large forces due to airfoil effects testing, which costs up to approximately
Mirrors were meshed as shell elements.
when the wind blows against the side of 10,000 euros per day. Physical testing of
the structure. NEM engineers slice the the final design, however, is required as
mirror surface into a number of sub part of the certification process.
areas and apply forces to each of these The company expects to increase its
areas based on the results of wind tun- use of bucking analysis to reduce the
nel testing. A gravitational load is also weight and cost of the heliostat support
applied and a boundary condition is used structure. It is no secret that a tubular
to model the support of the heliostat by structure becomes stronger and lighter
the foundation. when you increase its diameter and reduce
NEM uses plastic deformation calcu- its thickness, but eventually you reach
lations on small sections of the model to the point at which buckling becomes a
account for permanent deformations of critical failure mode. Buckling analysis
the structure. These effects normally are can help to determine how far NEM can go
very small and have minimal impact on in making the pillar and torque tube thin-
overall structure deformation, but they ner, with the goal of reducing the steel
require a considerable amount of com- mass of the structure while maintaining
putational resources to perform this its strength.
nonlinear analysis. NEM engineers typ- Driving the steel mass down and
FEA results show deformation of structure.
ically perform plastic deformation on designing stiffer optical systems is the
small sections of the structure to ensure key to delivering clean, cheap and reliable
that the permanent deformation is small energy from the sun. Finite element anal-
enough that it will not affect the accu- ysis is helping NEM engineers to improve
racy of the linear analysis used for struc- performance and reduce cost of the com-
tural predictions. pany’s heliostat designs at a much faster
NEM engineers use ANSYS Parametric pace than could be accomplished solely
Design Language (APDL) command snip- by building and testing prototypes.
pets to evaluate the model at different
angles and wind speeds as part of a batch Reference
process. In a typical case, they set up a [1] Wang, U. The Rise of Concentrating Solar Thermal
routine to run the model 50 times, which Power. Renewable Energy World, 2011, June 6.
generates perhaps 50 gigabytes of output.
The output is then exported to a ray trac-
ing routine, which NEM engineers wrote With 200 different
in the MATLAB® programming language, geometric bodies,
it is an enormous
that determines the impact of the struc-
tural deformation on the light reaching
the receiver. The results are presented job to manually
in the form of a flux map that shows
the amount of energy received in differ-
create contacts,
ent areas of the receiver. These results, in so NEM engineers
A MATLAB routine converted deformation into
turn, are used by another MATLAB rou- use automation
flux maps that show energy reaching receiver
in tower. An ideal flux map (top) is produced by
tine to calculate the potential amount of
electricity that can be generated by each
features in ANSYS
a perfectly stiff structure; flux map on bottom
shows a more realistic case.
heliostat design. The R&D team compares Workbench.
these results to the projected manufactur-
ing costs for the particular design to cal-
culate its return on investment.

ACADEMIC Departments
GATHERING
DUST
A new simulation approach helps to improve particle removal efficiency of wet scrubbers.
By Damian Pieloth, Physicist, Ph.D. Student; Gerhard Schaldach, Chemist, Scientific Assistant, and
Peter Walzel, Doctor of Technical Sciences, Professor, Chair of Particle Technology Lab,
University of Dortmund, Germany
W
et scrubbers are the Löffler model[2,3] of particle deposition principle is to capture particles in liq-
frequently applied on droplets; it was implemented employ- uid droplets, with the liquid absorb-
in industrial proc ing user-defined subroutines within ing the pollutants. The liquid droplets
esses, such as ANSYS computational fluid dynamics have much higher inertia than the par-
removing pollut- (CFD) software. These user-defined sub- ticles, which makes them relatively easy
ants from fur- routines permit the implementation of to separate from the outlet gas stream
nace flue gas and sticky particles from new user models and the extensive cus- using a mist eliminator or entrainment
process streams. The greatest challenge tomization of existing ones to provide separator. Efficient removal of particles
in designing wet scrubbers is to remove additional modeling capabilities, proprie- 5 micrometers or less in diameter gener-
the maximum amount of particles while tary data or specific boundary conditions. ally requires devices, such as jet scrubbers,
minimizing energy and washing-liquid The key advantage of the Euler–Lagrange with high gas–liquid relative velocities
consumption. Pilot plant experiments method in simulating wet scrubbers is that provide pressure recovery by proper
optimizing the design of scrubbers are that it requires far fewer computational duct geometries. An advantage of wet
time-consuming and expensive because resources than the alternate Lagrange– scrubbers is that, generally, they are the
of the high cost of building prototypes Lagrange approach of treating particle– only air pollution control device that can
and the multitude of design parame- droplet collisions explicitly. remove both particles and gases — and in
ters. Researchers from the University of some cases they can achieve high removal
Dortmund have developed simulation SCRUBBER DESIGN PRINCIPLES efficiency for both pollutants. In addition,
methods that accurately predict the per- Wet scrubber devices remove small wet scrubbers can handle high temper-
formance of a wide range of wet scrubber particles or small droplets from process atures with moisture simultaneously
designs[1]. The new approach is based on gases. Their underlying physical working cooling the flue gases, thus minimizing
Clean gas
Rotary sprayer
Droplets
Fresh water
Waste water
Raw gas
Diagram of rotary wet scrubber ANSYS CFX numerical modeling results (right) matched Löffler’s theoretical calculations and
experimental data (left).

The key advantage of the Euler– possible to simulate particle capture by
Lagrange method is that it requires droplets for different particle-size and

droplet-size distributions by using several
far fewer computational resources than Euler phases for the dust. One limitation
the alternate approach of treating of the Löffler model is that it assumes a

homogenous distribution of dust particles
particle–droplet collisions explicitly. within the gas flow. Larger dust particles
violate this assumption because they tend
to segregate from the suspension state
under the influence of mass forces such
as gravity or inertia. However, in most
industrial applications, wet scrubbers
the overall size of the equipment that is of dust particles on droplets increases. are applied to dust particles smaller than
required to clean the incoming gas stream. Bigger droplets decrease deposition rates 5 microns, so this simplification is widely
Physical tests of wet scrubbers play due to the weaker inertial forces on the justified. Also, strands of cleaned gas
an important role in the design process, dust, while smaller dust particles are behind individual droplets are assumed
but it is very difficult to design wet scrub- more difficult to deposit because they to equalize their concentration quickly.
bers solely with physical experiments. are more likely to follow the streamline
The design parameters for a typical wet around a droplet. The CFD simulation fol- SIMULATION CORRELATES WELL
scrubber include the number of sprayers, lows the trajectories of individual drop- WITH EXPERIMENTS
the droplet-size distribution they gener- lets. Droplets leaving the spray system The Dortmund team performed scrub-
ate, sprayer location within the scrubber, collect dust particles along their trajec- ber simulations with ANSYS CFD, com-
the particle-size distribution of particles tory through the scrubber, thus increas- pared them with results from literature,
to remove, the geometry of the scrub- ing the dust load of the droplets. and found a very good match. Next, they
ber, and the gas mass flow. The designer In the second step, the cleaned gas designed and built a prototype pilot-scale
needs to understand the impact of each volume is evaluated by tracking the drop- rotary scrubber to generate experimental
of these parameters on the performance lets through the gas flow and calculating data and also to compare these results
of the scrubber to optimize the design. the deposition efficiency in every control with ANSYS CFX simulations. Salt par-
Computer simulation using CFD has long volume passed by the droplets. Dividing ticles were powdered and screened in
been recognized as a good alternative to the cleaned volume by the volume of the order to fit typical dust-particle sizes as
pilot trials, but the traditional method droplets yields the specific cleaned vol- in industrial applications of wet scrub-
of explicitly simulating particle–droplet ume of the droplets. This calculation bers. By using de-ionized water, they were
collisions requires great computational defines the performance of the scrubber able to measure the capture of salt parti-
resources and has therefore been simply for a given volume of sprayed water. cles at droplets by reading the electrical
beyond the reach of most designers. The third step involves calculating the conductivity of the water after the scrub-
local dust concentration change by add- bing process. The droplet-size distribu-
NEW SIMULATION METHOD ing up the cleaned volumes of all drop- tion was varied by changing the rotational
Members of the Particle Technology (PT) lets crossing a given gas flow volume.
group at the University of Dortmund This model focuses on individual drop-
addressed these challenges by utilizing lets and their deposition efficiency so that The Dortmund
the Löffler model of particle deposition on
droplets. Small dust particles tend to fol-
the information obtained from the model
can be applied to a wide range of scrubber team performed
low the streamlines of the gas flow around
droplets but leave the streamline due to
designs and geometries.
The authors used an Euler–Lagrangian
scrubber
their inertia and collide with the droplet. formulation to simulate web scrubbers in simulations
The model used here divides the scrub-
bing process of particles by droplets into
ANSYS CFX. The Euler phase represents
the two components of the raw gas: air with ANSYS CFD,
three subprocesses. and dust. The Lagrangian phase, which
compared them
with results from
The first step is calculation of the maps the droplets onto the simulation,
deposition efficiency of dust particles on also has two components: water and dust.
a single droplet for a given flow regime.
The deposition efficiency of particles on
During the course of the simulation, the
dust component within the Euler phase is literature, and
a droplet is a function of process param- transferred to the dust component within found a very
good match.
eters, including particle- and droplet-size the Lagrangian phase. The interchange of
distribution as well as gas flow velocities. dust loading between the two phases is
With larger dust-particle sizes and higher calculated within a CFD user-defined sub-
relative velocities between dust particles routine based on the flow and particle–
and droplets, the deposition efficiency droplet characteristics. This makes it

ACADEMIC
Pilot-scale prototype
ANSYS CFD wet scrubber simulation
velocity of the Lamrot sprayer installed within the device[4]. The

specific cleaned gas volume was calculated from mass balance
at the scrubber. The team made considerable efforts to match
the narrow, well-defined particle- and droplet-size distributions
in the physical experiments under different scrubber operating
conditions. In addition to this effort, the team simulated a jet
scrubber used by an industrial partner and compared the simula-
tions with experimental data. In all cases, the simulation results
showed good agreement with experimental data.
The need to efficiently remove dust particles from exhaust
gases challenges the design and optimization of wet scrubbers.
Equipment must meet strict separation efficiency requirements
Comparison of CFX predictions to experimental data
while also minimizing water and energy consumption. The incor-
poration of the Löffler model into ANSYS CFD software allows
for a quick evaluation of a wide range of possible designs until
an optimal design is found. This is done without the expense of
building and testing multiple prototypes. The simulation pre-
dictions have been correlated with a series of physical experi- References
ments and found to accurately predict scrubber performance. [1] Pieloth, D.; Kohnen, B.; Schaldach, G.; Walzel, P. CFD-Simulation von
The approach described in this article requires considerably less Nasswäschern, Chem.-Ing.-Tech., 2012, 84(1-2), pp. 127–137.
computing power compared to algorithms that treat particle–
droplet collisions explicitly. It also can be applied to other appli- [2] Löffler F., Partikelabscheidung an Tropfen und Fasern, Chem.-Ing.-Tech.,
cations that require an accurate prediction of capture of particles 1983, 55(5), pp. 171–178.
by droplets, for example powder insertion into a spray or coating
of particles[5]. [3] Ho, C.; Sommerfeld, M. Modelling of micro-particle agglomeration in
turbulent flows, Chem.-Eng.-Sci., 2002, 57(15), pp. 3073–3084.
The incorporation of the [4] Mescher, A.; Möller, A.; Dirks, M.; Walzel, P. Gravity affected breakup of
Löffler model into ANSYS CFD laminar threads at low gas-relative-velocities, Chem.-Eng.-Sci., 2012, 69(1),
software allows for a quick

pp. 181–192.
evaluation of a wide range [5] Schaldach, G.; Pieloth, D.; Kohnen, B.; Großmann, M.; Walzel, P. CFD-
of possible designs until an Simulation der Feststoffeinbindung in einen Sprühstrahl, Chem. Ing. Tech.,
2011, 83(6), pp. 893–899.
optimal design is found.

TECH TIP
DESIGNING SOLID
COMPOSITES
Employing ANSYS Workbench workflow streamlines simulation of solid composites.
By Matthias Alberts, CEO, CADFEM US Inc., Greenville, U.S.A., and

Pierre Thieffry, Lead Product Manager, ANSYS, Inc.
T
raditionally, layered
composites struc-
tures are modeled as
thin structures using
shell elements. This
approach is valid when
designing thin parts, such as hollow
tubes for bikes, panels for airframes
and wind turbine blades. But when
the parts are more massive, such as
gas turbine blades or stringers for
Figure 1. Composites
pressure vessels, using shell elements pressure vessel with
is not appropriate. In such cases, both titanium caps
stresses in the direction of the thick-
ness and shear stresses out of plane
are significant, and solid models
are required. Solid models are also appropriate when loads are
applied in the direction of the thickness or when the structure is
subject to large deformations.
While defining thin-layered composites poses several chal-
lenges, the definition of solid composites is even more com-
plex. The shapes usually are not simple and require special
treatment — a turbine blade, for example. Composites products
generally include noncomposites parts that must be included
in the simulation. Consequently, the engineering team needs
an efficient workflow for the design of products made of lay-
ered solid composites and other parts. An effective process
Figure 2. Workflow for design of a composites pressure vessel
starts by examining the layer definition, based on the same
method as used for thin structures, then moves on to create
solid composites by extrusion. This is followed by the assembly
of composites and noncomposites parts, culminating in analy-
sis of potential failure of the overall structure.
To highlight this workflow, the example presented is a pres- The entire simulation
process is based on
sure vessel (Figure 1). The entire simulation process is per-
formed in ANSYS Workbench (Figure 2) using ANSYS Composite
PrepPost (ACP). The workflow begins by defining the geome-
try. The model is split into shell composites parts (A) and solid
a workflow in ANSYS
noncomposites parts (D), which are recombined as solids to Workbench using ANSYS
Composite PrepPost.
create the final description of the analysis (B). This combined
solid assembly includes connections between parts, loads and
boundary conditions, as well as results such as stresses or
deformations. The investigation of composites failure occurs
as the last step in this process (C).

TECH TIP
To start, define the layers on a surface The critical area in the creation of
mesh (Figure 3). The surface generally solid composites is generating a solid
will be the inner or outer surface of the model of the layup using a solid extru-
product being designed. Define the lay- sion, based on the previous surface
ers in the following sequence: defini- definition of the plies. Advanced capa-
tion of materials, fabrics (and, optionally, bilities, such as ply tapering, surface
stackups), orientation of the various sur- smoothing or extrusion guidelines, are
faces of the composites (possibly includ- available to deal with complex shapes
ing draping for highly curved surfaces) (Figures 3a, 3b and 3c). You may apply
and, finally, ply sequence. This approach ply tapering using cutoff rules. Surface
is very close to the actual manufacturing smoothing can be performed using
process. An analogy can be made between the snap-to-geometry feature to fit the
the initial surface and a mold, and the extruded model to a given CAD surface.
ply sequence defined within the simula- Extrusion guidelines help to extrude the
tion tool can be the same as the actual fab- surface model along arbitrary directions.
ric layup within the mold. However, the Another important aspect is handling
simulation tool obviously offers more drop-offs. Ply drop-offs can cause damage
flexibility in ply ordering, as plies can be and delamination in a composite layup.
swapped easily, modified or removed to In simulation models, drop-off elements
achieve the required stiffness, weight and are represented by degenerated brick ele-
cost requirements. ments. They usually are made of a homog-
enous material such as resin.
Once the solid model has been created,
it is automatically merged into the final
assembly along with the noncomposites
parts (Figure 4). Automated contact detec-
tion between parts, loads and boundary
conditions, and solution settings all can
be specified as you normally do for any
regular model in ANSYS Mechanical. The
transfer of the composites material defi-
nition in the full assembly is completely
automated. Once the model has been
Figure 3b. Surface smoothing using CAD
solved, standard results such as deforma- surface (in green)
Figure 3. Definition of ply sequence on inner
surface of vessel tions or stresses can be displayed on the
full model.
The transfer of
the composites failure and the potentially problem-
atic elements and layers in the assembly
material definition (Figure 6). You can even create a graph to
in the full assembly

show failure criteria values through all of
the layers at a given location on the model.
is completely Employing the Workbench-based
workflow for solid composites delivers
automated. benefits from additional capabilities. If
the composites parts are subject to pres-
sure from a flow environment, you can
easily add the fluid flow simulation to
the simulation, and pressures can be
mapped automatically on the struc-
Additional capabilities are available tural model. Users also have the abil-
to analyze potential failure of the compos- ity to parameterize a model to perform
ites parts. Three-dimensional failure cri- sensitivity or optimization studies based
teria (maximum stress, maximum strain, on geometry or composites variations
Figure 3a. Ply tapering
Tsai–Wu, Tsai–Hill, Puck, Hashin, Cuntze) (thicknesses, orientation, etc.) using the
are available. A typical failure plot gives failure criteria as a performance indica-
the analyst information on the risk of tor of the design.

Figure 3c. Extrusion guidelines Figure 4. Assembled model showing composites and noncomposites parts
(metallic caps in light gray)
Figure 5. Stresses on assembled model
Figure 6. Post-processing composites showing failure on full model (left) or through element layers (right) in charts

ANSYS, Inc.
Southpointe
275 Technology Drive
Canonsburg, PA U.S.A. 15317
Send address corrections to

AdvantageAddressChange@ansys.com
Shoots for the Moon

with ANSYS
Robot by Astrobotic. Simulation by ANSYS.

Realize Your Product Promise®
Space is unforgiving. One small mistake can spell disaster.
Using ANSYS simulation technology, Astrobotic is testing its lunar rover designs
virtually in a computer – since testing on the moon isn’t all that practical. Astrobotic
is delivering on its product promise by working to safely deliver a working rover to
the surface of the moon.
Guess the sky is no longer the limit.
Visit ANSYS.COM/Astrobotic to learn how simulation software can help you realize your product promise

AA V7 I1 Full Version

Uploaded by

Document Information

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

AA V7 I1 Full Version

Uploaded by

Copyright:

Available Formats

ADVANTAGE

Excellence in Engineering Simulation VOLUME VII | ISSUE 1 | 2013

Tackle your toughest ANSYS simulation tasks with

Discover the difference of Intel Xeon processors

ANSYS, Inc. For ANSYS, Inc. sales information,

leaders create mind-boggling

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 1

All Systems Go Ready for Liftoff Shaping Up

14 Testing the Next Generation

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 2

A Systematic Approach Gathering Dust

Magnetic Attraction Designing Solid

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 3

Simulation in the News

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 4

simulating simply kem.de, January 2013

radars and onboard SGB designs power transformers so that

devices and simulate they are highly reliable and durable —

Requirements and System verification

architecture and validation

don’t just run a

Detailed design and verification

– Thierry Marchal, Director, Industry Project definition

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 5

By Robert Harwood, Aerospace and Defense Industry Director, ANSYS, Inc.

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 6

simulation tools in designing the most EFFICIENT MULTIPHYSICS

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 7

Complex A&D applications must be simulated with deep

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 8

THE SKY’S THE LIMIT

support development of innovative

aerospace and defense technologies defense technologies and deliver solu-

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 9

By Gregory Mooney, Systems Software Engineering Lead — Landing Systems

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 10

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 11

Since implementing SCADE Suite in

MOVING BEYOND CODE

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 12

TAKES OFF Aerospace to help trol systems offer an automatic braking

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 13

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 14

design methods meant that prototypes

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 15

 Rotunda and column  Lift-and-drive system

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 16

 Previous-generation bogie design  New bogie prototype during testing

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 17

A ntenna performance has always

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 18

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 19

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 20

C omputational fluid dynamics analy-

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 21

The adjoint CFD solver from ANSYS

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 22

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 23

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 24

© 2013 ANSYS, INC. ANSYS ADVANTAGE Volume VII | Issue 1 | 2013 25

Droplet collection Local heat

New ice layer 0.1

Mesh morphing -0.08 -0.06 -0.05 -0.02 0.00 0.02 0.04

FINAL GEOMETRY Curvilinear abscissa

Rotunda and column Lift-and-drive system

Previous-generation bogie design New bogie prototype during testing

Ice collection efficiency prediction on TBM850 air inlet Thermodynamic balance

Grid morphing for several time steps

Simulation chronology developed specifically for DAHER-SOCATA’s TBM850

Historic Vostock rocket engines

CFD mesh of the wing The morpher altered the

Fluent calculations Simulation

BSCW setup for HIRENASD is 3-D

brazil Structural simulation

Analyzing erosion patterns within recovery tree section