Desktop Engineering - 2015-08

August 2015 / deskeng.
com
CAD Workstation Q&A P. 32
Design in Virtual Reality P. 23
Linear & Nonlinear FEA

Buckling P. 50
Design
TECHNOLOGY FOR DESIGN ENGINEERING
with
WORKSTATION REVIEWS P.37
FACTORIES OF THE FUTURE P.44
DESIGN FOR COMPLIANCE P.48

Ferrari Takes a Victory Lap
With ANSYS
Victories by Ferrari. Simulation by ANSYS.

Realize Your Product Promise®
Winners demand the best. Ferrari would know. It has one of the best
racing records the world over.
Using ANSYS simulation technology, Ferrari is realizing its product

promise by optimizing critical aspects of its race cars, like brake
cooling systems and full-body aerodynamics, to better handle the
twists and turns of the racing world.
Sounds like Ferrari is in the driver’s seat.
Visit ANSYS.COM/Ferrari to learn how simulation software can help you realize your product promise.
Rapid Manufacturing with a
Polite Disregard for Tradition
Tech-driven injection molding, CNC machining and
3D printing for those who need parts tomorrow
Proto Labs uses proprietary software and a

massive compute cluster to accelerate manufacturing
of prototypes and production parts for every industry.
Got a project? Get 1 to 10,000+ plastic, metal or
liquid silicone rubber parts in as fast as 1 day.
Request your free copy of
Digital Manufacturing for Dummies
at go.protolabs.com/DT5GJ ADDITIVE MANUFACTURING
ISO 9001:2008 Certified | ITAR Registered CNC MACHINING

Major Credit Cards Accepted | © 2015 Proto Labs, Inc. INJECTION MOLDING
Degrees of Freedom by Jamie J. Gooch
Direct Data, Design

and the Future
W
hen my kids were younger, they liked to play product development time and expense.
the game of Telephone with their friends. If That’s one of the exciting promises of the Internet of
you’re not familiar with the game, it begins Things (IoT). Predictions for the future include products
with one person telling something to some- “phoning home” when they are about to fail, or to provide
one else, then that person repeating it to someone else and a stream of actual usage data that can become the basis of
so on until the information has gone through the group and improving the next generation of products. It sounds a lot
back to the person who originally provided the informa- like what happens when a software application crashes and
tion. After going through so many people, the information a window pops up asking you to send a report back to the
would invariably be some giggle-inducing, garbled version vendor. But what happens to that data once it’s submitted?
of the original. How is it collated and acted upon?
It’s not just kids who have a tough time communicat-
ing what they’ve heard. We played the same game in a Not So Fast
college journalism class to underscore the importance of We can assume that such real-world data would be worth the
first-hand information and not relying on a single source time and expense involved to make some products “smart”
when reporting a story. No matter how hard we tried to with various sensors to detect issues and the technologies
needed to communicate those issues. Let’s also assume that
the security concerns threatening the IoT will eventually
Design engineering teams should be mitigated. Even after those two high hurdles are cleared,
consider the source of outside data. there is still the issue of making sense of all that data.
There is already an over-abundance of data, much of it
not being effectively harnessed to improve product design.
accurately relay what we heard from one student to the Why? Because collecting, filtering, disseminating and ana-
next, the information would always come back differently lyzing massive amounts of data is still a huge challenge. It re-
than what was originally conveyed. quires a workflow, IT infrastructure and advanced algorithms
For reporters, the phenomenon can lead to some embar- that don’t yet exist in most organizations. Adding more data
rassing headlines. For design engineers, it can mean the dif- to the mix without an efficient system in place to handle it
ference between producing a product your company’s cus- would only exacerbate the issue.
tomers will love and something that they won’t buy. It might Many of the titans of technology are focused on grow-
mean producing something that is over-engineered for how ing the IoT and bringing the benefits of Big Data to more
it’s really being used, or something that fails when used in and more organizations. Depending on your industry, the
actual field conditions. use of real-time, real-world data may only be a matter of
time. However, especially in these early days, design engi-
Real-Time, Real-World Data neers should consider the source of the outside data coming
Simulation, prototyping and testing catch most of these is- in. The filters applied to the data that are intended to make it
sues before a product design makes it to the factory floor. useful should also be thoroughly vetted. Just like in the early
Engineering teams make extraordinary efforts to ensure the days of simulation solvers, you shouldn’t trust the algorithms
products they’re designing are safe, reliable and marketable to be perfect right out of the gate. An automated process of
— investing considerable time and expense in the product relaying Big Data may not be plagued by the same issues
development process. But what if design decisions could be humans are when playing a game of Telephone, but the old
based on more real-time, real-world data? Instead of getting computer programming adage “garbage in, garbage out” still
product use information filtered through a Telephone-like applies. DE
game of sales figures, maintenance records, focus groups,
test data and more, a direct data route would not only help Jamie Gooch is the editorial director of Desktop Engineering.
ensure the integrity of the information, but could also save Contact him at [email protected].
2 DESKTOP ENGINEERING August 2015 /// deskeng.com

Who will prevent downtime
and equipment failure?
You and NI will. With an integrated platform that combines flexible,

rugged hardware with intuitive software, NI helps organizations improve
operational efficiency by providing systems that monitor and analyze
rotating equipment. With advanced I/O, complex signal processing, and
data analytics and visualization capabilities, NI puts you on the cutting edge
of the Industrial Internet of Things and connects equipment, people, and
technology like never before. See how at ni.com/mcm.
800 891 8841

©2015 National Instruments. All rights reserved. National Instruments, NI, and ni.com are trademarks of National Instruments.
Other product and company names listed are trademarks or trade names of their respective companies. 22782
August 2015 VOLUME 20/ISSUE 12
34 New Horizons
in Data Storage
The ubiquitous hard disk drive’s
days are numbered, both for single-
user devices and group data storage.
By Randall S. Newton
37 Extreme Power and Performance

The new Lenovo ThinkStation P900 delivers
industry-leading results.
By David Cohn
12 Dealing with the

Big Data Deluge
IoT products can deliver valuable
40 Big, Bright, Fast and Loud
Origin PC delivers the massive ultra
high-end Millennium Pro X2 workstation.
By David Cohn
information that can help evolve
future designs — but engineers need INTERNET OF THINGS
to know how to use it.
By Beth Stackpole 44 Manufacturing in the World of Industrie 4.0
Manufacturers everywhere can learn from this German initia-
tive supported by the Siemens Digital Enterprise suite.
PROTOTYPE By Pamela J. Waterman
ENGINEERING SERVICES
23 Blending
Reality with Pixels 48 Design for Compliance
The age of augmented and Service providers help the medical device
virtual reality-integrated design industry accelerate time to market and
tools has humble beginnings. reduce risk.
By Kenneth Wong By Lauren Gibbons Paul
28 VR Speeds Design SIMULATE

The technology enables a
hands-on look for engineers
in a host of industries.
50 Linear and Nonlinear Buckling in FEA
Ensure your structures can withstand outside
By Frank J. Ohlhorst and
stresses with these methods.
Jess Lulka
By Tony Abbey
ENGINEERING
COMPUTING FOCUS ON BIG DATA
32 Answers to Your 16 Execs, Engineers
Workstation Questions and IT Prepare for Big Data
Insights on how to assemble the right hardware As Big Data matures, so too will
for your design engineering needs. technology and security to support its use.
By Brian Albright By Jim Romeo
19 Simulation Lifecycle
Management’s New Mission
ON THE COVER: Big Data begins to make its way into designs. The discipline takes on new challenges from IoT, apps and the cloud.
Images courtesy of iStockphoto. By Kenneth Wong

DEPARTMENTS 54 Editor’s Picks PUBLISHER
Tom Cooney
Products that have grabbed the
EDITORIAL
editors’ attention.
2 Degrees of Freedom By Anthony J. Lockwood
Jamie J. Gooch | Editorial Director
Kenneth Wong | Senior Editor
Direct data, design and the future. Anthony J. Lockwood | Editor at Large
By Jamie J. Gooch Jess Lulka | Associate Editor
Sarah Petrie | Copy Editor
6 Virtual Desktop CONTRIBUTING EDITORS

Tony Abbey, Brian Albright, Mark Clarkson,
Researchers simulate hidden Tour de David S. Cohn, John Newman, Frank Ohlhorst,
France advantages, Dr. Ahmed Noor’s Beth Stackpole, Pamela J. Waterman
2015 NAFEMS World Congress ADVERTISING SALES

603-563-1631 • Fax 603-563-8192
Erich Herbert | Sales Manager (x263)
Chris Casey | Sales Manager 847-274-5476
55 Advertising Index ART & PRODUCTION

Darlene Sweeney | Director 603-563-1631 (x257)
55 Fast Apps A PEERLESS MEDIA, LLC PUBLICATION

Brian Ceraolo | President and Group Publisher
Sometrix uses Simpleware software to Tom Conlon | Vice President
model medical devices and Siemens
presentation on cognitive computing leverages Teradata platforms to ADVERTISING, BUSINESS, & EDITORIAL OFFICES
Desktop Engineering ® magazine
and the Cradle to Cradle Product Design streamline Big Data. Peerless Media, LLC
Challenge winners are announced. 1283D Main St., PO Box 1039 • Dublin, NH 03444
By Kenneth Wong and Jess Lulka 603-563-1631 • Fax 603-563-8192
E-mail: [email protected]
www.deskeng.com
9 Rapid Ready Tech
3D printing could speed up the Internet,
Ford invests in Carbon3D, Dubai plans
a 3D-printed office building and the www.peerlessmedia.com
Ares system goes for crowdfunding. Kenneth Moyes | President and CEO, EH Publishing, Inc.
51 Spotlight 56 Commentary SUBSCRIBER CUSTOMER SERVICE

Directing your search to the PLM to support supply chain Desktop Engineering ® magazine
PO Box 677 • Northbrook, IL 60065-0677
companies that have what collaboration. 847-559-7581 • Fax 847-564-9453
you need. By Tony Christian, Cambashi E-mail: [email protected]
Desktop Engineering® (ISSN 1085-0422) is published
monthly by Peerless Media, LLC, a division of EH
Publishing, Inc. 111 Speen St., Ste. 200 Framingham,
MA 01701. Periodicals postage paid at Framingham, MA
and additional mailing offices. Desktop Engineering® is
Visit Deskeng.com distributed free to qualified U.S. subscribers.

SUBSCRIPTION RATES: for non-qualified; U.S. $108
one year; Canada and Mexico $126 one year; all other
• Navigation countries $195 one year.
Quickly find the content you’re looking for Send all subscription inquiries to Desktop Engineering 111
in dedicated Design, Simulate, Test, PLM, Speen St. Ste. 200 Framingham, MA 01701
Prototype/Manufacture, Engineering LIST RENTALS: For information on list rentals, contact

Statlistics, Danbury, CT: 203-778-8700.
Services and Engineering Computing
Postmaster: Send all address changes to Desktop
sections. Drill down deeper via individual Engineering®, PO Box 1496 Framingham, MA 01701-
topic areas, technology vendors and 1496. Reproduction of this magazine in whole or part
authors within each section. without written permission of the publisher is prohibited.
All rights reserved ©2015 Peerless Media, LLC.
• Integration Address all editorial correspondence to the Editor,

Desktop Engineering. Opinions expressed by the authors
All of Desktop Engineering’s content are not necessarily those of Desktop Engineering.
— from videos to Engineering Services Unaccepted manuscripts will be returned if accompanied
by a self-addressed envelope with sufficient first-class
Directory listings, to our blogs and social postage. Not responsible for lost manuscripts or photos.
media posts — appear in our news feeds
throughout the site.
Check out deskeng.com and let us know

what you think.
deskeng.com /// August 2015 DESKTOP ENGINEERING 5

Virtual Desktop by Kenneth Wong & Jess Lulka
Can Simulation Predict the

Tour de France’s Outcome?
P
rofessor Bert Blocken, PhD care, consumer product and con- tional fluid dynamics (CFD) simula-
student Yasin Toparlar, struction at ANSYS, have spent a tion in ANSYS software. We then
and Thierry Marchal may good part of their careers recreat- confirmed our numbers with a wind
have uncovered something ing physical phenomena — like the tunnel test.”
that could affect the regulations of aerodynamics of moving cars and The 3D geometry of the cyclist
the Tour de France. The real race trains — in pixels inside computers. came from a laser scan from a pre-
took place along the 3,360-km route At first glance, their professional vious project dating back to 2006.
that stretches from Utrecht, Nether- expertise seems to have nothing in The 3D model of the vehicle was
lands to Sèvres, France, but Marchal common with the cycling cham- purchased online, from one of the
pointed out: “Part of the race can be pions who rely on their muscles, 3D content suppliers. The simulated
won in front of the computer.” training and stamina. But it does. scenario covers a 50-km distance.
Blocken, the chair of Building “We have begun working with The researchers incrementally re-
and Urban Physics at the Eind- athletes going for the Olympic Gold,” duced the distance between the car
hoven University of Technology says Marchal. The same ANSYS tech- and the cyclist to see what happened.
in the Netherlands and at Leuven nology used to simulate and study the “When the car was following the
University in Belgium, and Marchal, airflow around a passenger jet or a cyclist at a 5-meter distance, the
global industry director for health- Formula One racecar can be applied drag reduction gave the cyclist an
to a swimmer or a cyclist. advantage of about 24 seconds over
“In a cycling race, the frontrun- the car following at 30 meters,” says
ner is protecting the follower from Blocken. “Conducting CFD (com-
the wind effect, so it’s generally putational fluid dynamics) simula-
understood it’s easier to be behind tions allowed us to put the cyclist
someone,” says Marchal. “But Bert and the car extremely close to each
Blocken and his team found out the other. You wouldn’t be able to do
follower may also be influencing the that in a real experiment because it
frontrunner’s performance by push- would be dangerous to the cyclist.”
ing the air toward the front.”
In the case of a vehicle following CFD as a Predictive Tool
a cyclist, Blocken and his PhD stu- CFD should be used where it counts
dent Yasin Toparlar discovered that most, Marchal suggests. “For cer-
the effect is significant enough to tain segments of the race, like the
change the outcome of the race. stages involving mountain climbing,
you might not want to do CFD for
The Race Inside the Computer the climb, where everyone slows
Currently, a car following a cyclist down. But for downhill, it makes
in the race is required to keep a more sense,” he says.
distance of 10 meters for safety rea- Some sports equipment manufac-
sons. But based on the research from turers now work with CFD experts
Eindhoven University of Technol- to improve product performance.
ogy with the ANSYS software, this For example, bike makers and hel-
The use of digital simulation in distance should be re-examined. met makers could reshape their
ANSYS software allows researchers “We were simply trying to un- products to improve airflow and
to incrementally change the distance derstand the physics of the influ- reduce drag. Even the cycling and
between the vehicle and the cyclist. ence of a car following a cyclist swimming suits are subject to scru-
Image courtesy of ANSYS and on the drag of this cyclist,” says tiny under CFD for possible aero-
Eindhoven University of Technology. Blocken. “We first did a computa- dynamic improvement. (For more,

Preparing for the Era of Cog-
Assisted Product Development
I
n his keynote address at the NA- nator” and “The Matrix.” Noor has heard
FEMS World Congress (June 21- some of his peers express similar con-
Researchers from Eindhoven 24, San Diego), Dr. Ahmed Noor, cerns. “With all due respect, I don’t share
University of Technology a researcher and scholar from their paranoia,” he said in an interview
conduct simulations to Old Dominion University, brought up with Desktop Engineering. He believes
understand the aerodynamics “cogs,” among other things. it’s possible to achieve “human-machine
of a vehicle following a bicyclist These cogs are not the wheels and symbiosis,” where “humans are always in
in a race. Image courtesy of gears that once drove the steam engine- the loop and both work together so the
ANSYS and Eindhoven University era industrial activities; they’re at the overall system functions at its best.” He
of Technology. heart of what could very well be the new looks to wearable computing devices as a
IoT-era industrial revolution. They are precursor to the symbiosis.
“Cognitive computing for engineering
read “CFD Making Waves in Olym- analysis and design,” says Noor. Cognitive Computing Predictions
pic Swimwear Again,” deskeng.com/ “We are witnessing a new revolu- He envisions “cog-assisted product de-
virtual_desktop/?p=6002.) tion in computing and communica- velopment technologies, where every
But it would be a mistake to tion, brought about by the synergistic phase is aided by cognitive computing
think of CFD as the ultimate pre- couplings of a number of technologies, and predictive analytics.” At the moment,
dictor of competitive sports. “We including cognitive computing, Inter- however, he acknowledges such tools are
can boost the athlete’s performance net of Things (IoT), mobile and wear- almost nonexistent in in the commer-
with equipment, optimize his or her
situation, but the athlete is the one
who must deal with tiredness, keep
up morale and peddle the bike,”
Marchal says.
Blocken anticipates skepticism
on his research from some, but
he’s confident in his findings. “The
agreement between the digital
simulation and the real wind tunnel
test was even better than expected.
And we have 17 years of experience able devices, big data prescriptive ana- cial market; promising versions of it are
in performing accurate and reliable lytics, among others,” he says. In his difficult to find even among university
CFD simulations with the ANSYS talk, he identified “some of the char- research projects. “IBM is working very
software,” he says. acteristics of the coming intelligence hard to develop this type of intelligence,”
Eindhoven University of Technol- / convergence / mass customization he says. “Engineering and simulation
ogy and ANSYS formally notified era, the future cognitive cyber-physical software vendors should seize the mo-
the cycling community and the regu- engineering systems (beyond today’s ment and work to integrate this feature in
latory organizations of their find- autonomous systems), and their impli- their next-generation offerings.”
ings. “We feel we need to go public cations on the product creation tools, Considering the fast pace of change,
with the findings because we are facilities and environments.” Noor suggested we should learn to
concerned with the fairness of the The human-machine interaction so become much more nimble and agile.
game,” says Marchal. “We also want far has been marred by anxiety and dis- “We should revisit our plans and strate-
the UCI (Union Cycliste Internatio- trust — the fear that automation and AI gies once in a while, perhaps every two
nale) to be aware of that so it could (artificial intelligence) threatens human months, and revise them in response to
modify the regulations.” supremacy. It’s a dystopian vision we have the changes we see,” he says.
— K. Wong seen depicted in movies like “The Termi- — K. Wong

Virtual Desktop by Kenneth Wong & Jess Lulka
Sustaining Product Design

I
n March, the Cradle to Cradle
Jerri Hobdy’s stool design implements
(C2C) Products Innovation Institute
recyclable materials for a healthier piece
challenged designers everywhere to
of furniture. Image courtesy of Jerri Hobdy
contribute to the world of sustain-
able product design. The institute part-
nered with Autodesk for its first Cradle
to Cradle Product Design Challenge.
The winners announced this spring are
certainly bringing in a new era of sus-
tainable design. Entries were submitted
from a variety of sectors, including home
and furniture, apparel, electronics and
automotive. The contest selected winners
from three different categories: profes-
sional, student and Autodesk Fusion user.
With her winnings, Hobdy will fin- toward public restroom users by helping
Have a Seat ish prototyping her chair and stool be- them learn how to properly wash their
Jerri Hobdy, a furniture designer, created fore submitting it for official C2C cer- hands in addition to reminding them of
the PUREIIFIHIDE, a chair and stool tification — then she’ll start offering their environmental impact.
collection. The product uses recyclable the first commercially available C2C Smith originally thought a waterfall
materials like solvent-free, vegetable certified residential home furnishings design might be ideal for educating
tanned leather and steel. The growing across the U.S. users on environmental impact. How-
market for non-toxic furniture and her ever, he soon discovered that current
goal to provide future customers the Sustainability in the Bag waterfall faucets on the market weren’t
healthiest, best-designed product on For consumers everywhere, student successful in doing so — but a more vi-
the market inspired Hobdy’s design. winner Tjitte de Wolff from the Uni- sual design with graduated cylinders and
This objective is what drives Hobdy versity of Twente, designed the Venlo measuring cups did.
to incorporate reusable materials and Bag. Made from 99% recycled materi- This led Smith to modify the Fi-
sustainable design into product creation. als, the bag is suited to address pollu- nite Faucet to include a clear upper
“I am interested in sustainable design tion caused by plastic bag use. De Wolff cylinder to act as a visual monitor for
because I am interested in healthy design was inspired to make this design after water usage. Once it empties, it must be
… for the planet but even more so for viewing the documentary called “Plas- turned off to refill. Aside from using sus-
people. In the U.S., especially, we are tic Planet,” and learning more about tainable design to create a more effective
on the edge of a huge and continuing the Great Pacific Garbage Patch. faucet, Smith also sees it as something
paradigm shift in how we view access to “The demand for cheap, single-use that can bring innovation and smart
health and healthy choices in a vast range shopping bags will still exist,” says de business practices into product design.
of purchase decisions. The furniture Wolff. “Therefore a Cradle to Cradle “From the fun side of things, sus-
industry is a very chemical ridden, under- solution for plastic bags not only elimi- tainable design often involves some re-
regulated puzzle. It will take people that nates pollution, but also contributes ally neat and innovative material use, in
have vision and passion to see the long- to a healthy ecosystem following the order to recycle materials in new ways,”
term and immediate benefits of cleaning credo: do more good, rather than less says Smith. “On a more serious note,
up furniture manufacturing. I am one of bad.” He plans to use his winnings for It’s easy to see sustainable design as just
those people. I am encouraged by [the his studies and for creating prototypes another part of the green movement,
fact that] C2C and Autodesk are willing of the Venlo Bag. but it’s also a really intelligent busi-
to partner with individual designers from ness decision. The savings in energy,
around the world like myself to make Watching Water Usage cost and labor of intelligently recycling
positive changes. Sustainability is impor- Cole Smith, the Autodesk Fusion 360 products can be immense compared to
tant because it affects the big guys and winner, submitted his concept for the designing from stock material.”
the small [ones],” Hobdy says. Finite Faucet. This product is geared – J. Lulka

Making Digital Designs Physical Rapid Ready Tech
3D Printing the Internet
The backbone of the Internet is miles and Dubai will be Home
miles of fiber cable, running through the
ground, along the sea bed and so forth.
to First 3D-Printed
The basic method for creating optical Office Building
W inSun Global and the Museum
of the Future have partnered to
bring on the revolution. The Chinese
company, which has already proven its mastery of AM (additive manufacturing) construction,
has agreed to erect an office building in Dubai purely through 3D printing. It sounds
innocuous enough, but once the project is completed, WinSun will have set the foundations
for a whole new industry with the first commercial sale of a building made through AM.
The modest 2,000-sq.-ft. building will be constructed near the Museum of the Future
and act as temporary office space for its staff. Printing will be accomplished by a 20-ft. tall
AM system, which will construct the basic components of the building. These components
fiber is still basically the same as when will then be assembled in a few weeks to finish the building.
someone first stretched out a piece of MORE ➜ rapidreadytech.com/?p=8950
glass. Researchers at the University
of Southampton are researching new controlled)
group vice president of Global Product
methods of optical fiber production using engraving/
Development and CTO. “We are excited
additive manufacturing (AM). milling and
to further our relationship and look
The new process will use ultra-pure 3D scanning.
forward to innovating together to make 3D
glass powder to build optical fiber using It isn’t quite a
manufacturing a reality.”
AM, allowing researchers to specifically makerspace,
design each part of the fiber. In order but it’s closer MORE ➜ rapidreadytech.com/?p=8928
for the project to succeed a number of to a true utility
hurdles need to be overcome, including,
“… the high melting temperature of the
device than U.S. Government Moves
a 3D printer alone.
glass (over 2000˚C in the case of silica); During the Indiegogo campaign, Against 3D-Printed Guns
the need for precise control of dopants, backers can get the full Monty for less 3D printing
refractive index profiles and waveguide than the cost of many 3D printers alone. became
geometry; and the need for transitions MORE ➜ rapidreadytech.com/?p=8933 part of
between the layers to be smooth, the national discussion
otherwise the properties of the resultant about gun control following the
fiber will be altered.” Ford Partners successful firing of the “Liberator”
MORE ➜ rapidreadytech.com/?p=8957 with Carbon3D designed by Cody Wilson of Defense
Distributed. Wilson quickly found his
Although prototyping alone was enough
design and his company under fire from
Ares All-in-One Printer to attract automobile manufacturers
to 3D printing, companies have begun
the U.S. Department of State, which
Heads for Crowdfunding to demand more from new technology
issued an order to remove the plans from
his website.
EasyArts has launched the Ares on investments. Companies have not only
Although Wilson complied, he
Indiegogo. Ares is a fairly standard Fused begun to expect the systems to produce
launched a lawsuit against the
Deposition Modeling-style 3D printer, but parts faster, but also expect a certain
agency claiming the order violated his
it also has the potential to transform into level of practical use for a 3D printed
rights to freedom of speech. Regardless
a number of other tools, including a laser part. The need for better and faster 3D
of Wilson’s suit, the U.S. Department
engraver and 3D scanner. printing led Ford to form a partnership
of State has begun to move ahead with
In the case of the Ares, easy access to with Carbon3D.
stricter rules regarding the availability and
the print head is intended to allow users to “Carbon3D’s CLIP technology has
legality of electronic files that can lead to
swap different tools into place. Along with allowed us to realize our need for high-
the unauthorized production of firearms.
3D printing, the system is capable of laser speed, high-quality printing of actual
engraving, CNC (computer numerically automotive-grade parts,” said Raj Nair, MORE ➜ rapidreadytech.com/?p=8972
➜ For more coverage of rapid technologies, visit RapidReadyTech.com

BOXX Technologies /// Visual Collaboration Sponsored Report
The Universal Language

Show, don’t tell when it’s time to explain design intent.
I
f you’ve ever visited a foreign country where you didn’t
speak the language, you have an idea what it must be like
for non-engineers to understand design intent. Speaking
more slowly or loudly doesn’t help. Showing a layman sche-
matics, Excel tables or even simulation visualizations is the
engineering equivalent of a tourist asking directions to the
airport by stretching out their arms to pantomime a plane.
You might get your idea across, but it will take awhile — and
time is a luxury most design engineers don’t have.
Design engineers speak the same language. Tolerances,
loads, stresses, thermal envelopes and more make it easy
for engineers to communicate requirements, development
issues and engineering change orders. However, not all
RENDERING brings an idea to life in a way a 2D schematic
clients and colleagues speak that language, which stymies
collaboration and results in inefficiencies and lost oppor- or even a 3D CAD model cannot. Image courtesy of
tunities. If only you could show them exactly what you had Daniel Simon.
in mind in a way anyone could understand. That’s what
rendering allows you to do. and ultimately better and more satisfying designs. Rendering
Rendering can be the universal language of manufactur- can help you quickly lead colleagues and clients to better de-
ing by enabling true visual collaboration among designers, cisions with easily understandable visual information. From
clients, executives and other key stakeholders. Where in the initial concepts, through multiple design changes and to the
past rendering was a complex, time-consuming task that re- final presentation, rendering provides all stakeholders with a
quired expert assistance and expensive computing resources, clear view of the project.
software and hardware tools now exist that put powerful Equipped with the right hardware, including overclocked
rendering capabilities into the hands of designers. A picture workstations, powerful GPUs and render
maybe worth 1,000 words, but rendering can be worth much nodes built to enable faster, iterative visual-
more as it allows you to fully express the intent of your de- ization of designs, engineers can work faster,
sign, help win bids, market products, identify flaws and im- provide better results and truly collaborate
perfections, and explore future concepts. with other team members, customers and
colleagues in various disciplines in ways that
Communicate Early and Often weren’t previously possible.
The design process has been compressed. Customers expect
more complex products in an increasingly short timeframe, Making the Case
which has made communication throughout the process even
more critical. for Visual Collaboration
Prototypes, while extremely useful in communicating the
look and feel of a design, are expensive to build and often aren’t
created until later in the design process when most of the key
A new paper produced by Desktop Engineering on
behalf of BOXX Technologies makes the case for
investing in the hardware, software and workflow to
decisions have already been made. Lacking the ability to quickly support visual collaboration. It explains how rendering
and effectively communicate design intent and the impact of can improve the design process, enhance collabora-
change orders to all stakeholders inhibits collaboration. tion and enable better customer interactions through
Quality rendering, however, can cut through the techni- real-world examples.
cal noise early in the design process, allowing colleagues and Download “Making the Case for Visual
clients to experience the “ah-ha” moments that can lead to Collaboration” for free at deskeng.com/de/visual.
new ideas, faster change approvals, clearer communication,

Visual Collaboration Workﬂow
BOXX APEXX 2 and APEXX 4
The APEXX 2 workstation for CAD design, 3D modeling
DQGDQLPDWLRQZRUNÁRZVIHDWXUHVVXSSRUWIRURSWLRQDO
RYHUFORFNHGSURFHVVRUVDQGWZRIXOOOHQJWKSURIHVVLRQDO
*38V,WLVDYDLODEOHZLWKRSWLRQDOHLJKWFRUHSURFHVVRUVIRU
LQFUHDVHGUHQGHULQJSRZHUDQGPXOWLWDVNLQJ7KH$3(;;
LVFRQÀJXUDEOHZLWKXSWRWZR,QWHO;HRQSURFHVVRUV
for total of 36 cores (72 threads), allowing it to handle
UHQGHULQJDQGVLPXODWLRQWDVNVZLWKHDVH,WDOVRVXSSRUWV
IRUDQRSWLRQDORYHUFORFNHGVLQJOHSURFHVVRUDQGXSWRIRXU
SURIHVVLRQDO*38V
BOXX renderPRO
7KHUHQGHU352LVDFRPSOHWHFRPSXWHUSRZHUHGE\GXDO
,QWHO;(21SURFHVVRUVWKDWKDVEHHQVWULSSHGRIDOOWKH
IHDWXUHVQRWQHFHVVDU\IRUUHQGHULQJVRLWFDQIRFXVRQ
&38FRUHFRXQW,WZDVEXLOWIRUPD[LPXPZRUNORDG
RSHUDWLRQ7KHDOODOXPLQXPFKDVVLVZDVGHVLJQHGWRVLW
GLUHFWO\RQDQ$3(;;DQGWREHVWDFNHGVHYHUDOXQLWVKLJK
IRUDGHVNVLGHUHQGHUIDUPVROXWLRQ
BOXX renderBOXX
UHQGHU%2;;LVSXUSRVHEXLOWWRSRZHUUHQGHULQWHQVLYH'
JUDSKLFVDQGDQLPDWLRQZRUNÁRZV)HDWXULQJGXDO,QWHO;HRQ
SURFHVVRUVXSWRSK\VLFDOFRUHVDVLQJOHUHQGHU%2;;
FDQDFWDVWKHFRUQHUVWRQHIRUDUHQGHUIDUP7HQPRGXOHV
ÀWLQWRD8[LQVSDFHZKLOH,30,WHFKQRORJ\
IHDWXULQJNH\ERDUGYLGHRDQGPRXVH.90RYHU,QWHUQHW
3URWRFRO,3HQDEOHV\RXWRHIÀFLHQWO\PDQDJHDUHQGHU
farm network.
BOXX 5HQGHU)DUP2Q:KHHOV
5HQGHU)DUP2Q:KHHOVLVDWXUQNH\UHQGHUIDUPVROXWLRQ
LQDFRPSOHWHKDUGZDUHSDFNDJH7KHXQLWLQFOXGHVUDFN
PRXQWHGGXDO&38UHQGHUQRGHVLQDPRELOHHQFORVXUH
52:FDQEHFRPSOHWHO\FXVWRPFRQÀJXUHGWRPHHW\RXU
QHHGV7RDVVLVW\RXU52:GHSOR\PHQW%2;;RIIHUVRQVLWH
VWXGLRVHWXSZLWKXQSDUDOOHOHGH[SHUWLVHDQGVXSSRUWWR
JHW\RXU52:XSDQGUHQGHULQJLQQRWLPH,WLVDYDLODEOH
LQDUDQJHRIVL]HVDQGH[SDQGDEOHWRRYHUPRGXOHV
FRUHV

Focus on Big Data /// Internet of Things
Dealing with the

Deluge of Big Data
IoT products can deliver valuable information that can help evolve
future designs — but engineers need to know how to use it.
BY BETH STACKPOLE
I
nternet of Things (IoT) data is the among other variables — is massaged, egy for ColdLight, a provider of an
new currency driving strategic deci- potentially mingled with external data, automated, advanced and predictive
sion making, including helping en- and ultimately mined for insights that analytics platform recently acquired by
gineers gain insight into what to build could direct any number of product PTC. “There needs to be some kind of
and how to build it more effectively. Yet development decisions. handshake between man and machine
in order to reap the full impact, orga- This scenario, while still nascent, in order to find value in IoT data.”
nizations need a strategy for managing has huge potential to transform the In recognition of both the IoT’s
the data deluge so it doesn’t swamp ef- engineering and design process. Yet the potential and the Big Data problem,
forts to optimize product designs. problem, according to experts, is that traditional design tool vendors are
Big Data’s role is to deliver in- the data coming off of smart connected taking aggressive measures to aug-
sights that can steer higher quality products is so vast that it’s next to im- ment their product portfolios with
product designs and be leveraged for possible for engineering teams to come new capabilities that address emerg-
predictive maintenance to more ef- up with any of these killer insights ing requirements. Some, like PTC and
fectively service products once out without help from an additional tech- Siemens PLM Software, have made
in the field. The concept works by nology platform that will facilitate data acquisitions in the Big Data analytics
equipping products of all types with collection and analysis. space planning to meld what has his-
an assortment of sensors that collect “The data has gotten so big that torically been an unrelated enterprise
and transmit critical data in real-time the ability for the human mind to find information silo into the traditional
over the Internet to a central reposi- points of value within the data set is engineering software workflow. Oth-
tory. From there, the data — recorded extremely difficult,” says Rob Patter- ers like MathWorks are positioning
temperature, stress points and speed, son, vice president, Corporate Strat- their machine learning capabilities to
Instead of requiring the user to define a formula or

The Classification Learner app lets you interactively equation, machine learning algorithms automatically
train models to classify data using supervised fit widely applicable models to data.
machine learning. Images courtesy of MathWorks Inc.

help solve this challenge, and non- plenty of data — in PLM (product life- never have considered before to get to
traditional players such as analytics cycle management), ERP (enterprise answers they never would have found,
companies are also jumping into the resource planning) and voice of the he explains.
fray and tuning Big Data solutions to customer systems — he says the prob- Cutting-edge organizations might be
address the needs of engineers. lem of fragmented silos creates infor- doing some of this kind of analysis on a
“I don’t know any great answer to mation chaos in companies, and tradi- very rudimentary scale, but it’s next to
the problem out there today,” says Mike tional business intelligence tools don’t impossible to see patterns between data
Campbell, PTC’s executive vice presi- help span the gap. “Today, companies stored in siloed systems without some
dent of CAD products. “All I know is are making decisions on a pretty short sort of automated, analysis tool, Boswell
that design engineers trying to develop set of data,” says Boswell, senior direc- says. To that end, Siemens recently re-
the next great products are starved for tor of Cloud Services Marketing and leased Omneo Performance Analytics,
real-world information. They recog- Business Strategy. “Traditional busi- part of the Omneo quality management
nize that they’ve been making assump- ness intelligence isn’t taking advantage platform that came through its acqui-
tions about requirements and real- of Big Data.” sition of Camstar, best known for its
world conditions … and they’re very To see how Big Data can change the manufacturing execution system (MES).
excited about gaining insights about equation, Boswell uses the example of Omneo Performance Analytics, avail-
how products are used or whether they an electronics manufacturer trying to able as a Software-as-a-Service (SaaS)
are under or over designed.” identify a quality problem causing re- offering, monitors data across the entire
calls on a hard disk drive. By combin- supply chain and customer experience,
Closing the Information Gap ing engineering data, PLM data, field analyzing billions of data combinations
Bill Boswell at Siemens PLM Software quality data, returns data, customer to uncover hidden intelligence that can
agrees that the traditional process often experience data and what Boswell dubs proactively pinpoint the source of prod-
leaves engineers making critical design “call home” IoT data, engineers can ask uct issues well after a product has been
decisions in a vacuum. While there is a different set of questions they may released to the field.

Focus on Big Data /// Internet of Things
in minutes — not weeks — the source
of the problem and resolve it.
“All of a sudden, the light switch
went on and the engineer realized it
wasn’t the processor or the mother-
board, but the integration between
the two,” Lokas says. “The number
of permutations you’d have to go
through to uncover that would be hu-
manly infeasible.” Instead, Omneo’s
ability to automate that discovery
process was essential to finding the
patterns in Big Data and jumpstarting
the root cause analysis for the service
engineer, he explains.
MathWorks customers facing simi-
Omneo runs billions of variable combinations to uncover hidden data trends lar challenges are leveraging a variety
that can used to mitigate product quality issues in future designs. Image of tools to find patterns and gain in-
courtesy of Siemens PLM Software. sights from an ever-expanding lineup
of data sources, including data acqui-
sition hardware, data warehouses, live
and historical industrial plant data,
and of course, IoT. Via its family of
Toolbox offerings — the Data Acqui-
sition Toolbox, the Database Toolbox,
the Image Acquisition Toolbox, the
Neural Network Toolbox to name a
few — MATLAB users are attacking
diverse and dynamic Big Data sets to
gain insights that can be leveraged to
improve existing designs or build new
products and services, says Seth De-
Land, MathWorks’ product manager
for Data Analytics.
DeLand uses the example of a large
off-highway equipment manufacturer
that had 40 years of stored data that
was continuously collected from ma-
chines in the field. Using the MAT-
LAB Toolbox offerings, the manufac-
turer could develop more accurate test
models to understand how its equip-
The monitoring tool in Omneo Performance Analytics can be used to track ment is actually performing in the
and observe trends for all Big Data sources in a single place for more field and whether it can hold up under
comprehensive analysis. Image courtesy of Siemens PLM Software. a full range of operating conditions.
Rather than use subjective tests that
Having this 360° view of the sup- cording to Rami Lokas, senior direc- an engineer devised years ago based
ply chain, coupled with the ability to tor, R&D for Omneo. The customer on assumption, engineers can leverage
churn through billions of data combi- was experiencing on-going issues re- the treasure trove of historical data to
nations in seconds, was a critical asset lated to laptop memory, and using the create models for hardware-in-the-
for Dell, an Omneo user, to proac- Omneo platform to connect the dots loop (HIL) testing and test beds that
tively identify and service a problem between the various data silos, a Dell are more reflective of real-world con-
for one of its major customers, ac- quality engineer was able to determine ditions. “They might find that a sub-

assembly was never designed for the that a human person couldn’t do.” comes into play, says Patterson. “For
use case that it’s under in the field and Helping human engineers make the first time from a design engineer
although it’s still working, they can de- sense of Big Data is where ColdLight standpoint, we have enough data out
liver feedback to the design team they fits into PTC’s product portfolio. The there … to provide an unbiased view
never would have gotten because there software, which also has a machine- of the world,” he says. “The ability to
were no field failures,” he says. learning component, among other data process that data and build predictions
MathWorks’ machine learning analysis capabilities, parses through with no predisposition to the business
capabilities also have potential for incoming data, finding patterns within will open possibilities for people to do
helping engineers get a handle on data sets to help with predictive ana- things in a different manner than has
Big Data. Engineers need to be able lytics, ColdLight’s Patterson says. been done in the past.” DE
to explore and visualize data, elimi- “The amount of data available is very
nate noise and outliers, and develop large, but it’s not all relevant for actual Beth Stackpole is a contributing editor to
predictive models, but the problem is use or the goal an engineer is trying to DE. You can reach her at [email protected].
they are engineers — not statisticians, achieve,” Patterson says. Manually de-
says Paul Pilotte, MathWorks’ tech- tecting patterns in such a large data set
nical marketing manager for Data is next to impossible for highly trained INFO ➜ ColdLight: ColdLight.com
Analytics. “When you have a lot of statisticians, let alone engineers. Cold- ➜ MathWorks: MathWorks.com
data, you need a more efficient way Light automates that process, using ➜ Omneo: Omneo.com
of developing models and getting data machine learning or neural networks,
to surface to create accurate predic- among other data analysis capabilities, ➜ PTC: PTC.com
tive models — machine learning is a to detect patterns and build predic- ➜ Siemens PLM Software:
powerful approach for that,” Pilotte tions to only send engineers the data Siemens.com/PLM
says. “It gives you the ability to build that they need. For more information on this topic, visit
highly accurate predictive models That’s when the true value of IoT deskeng.com.

Focus on Big Data /// Security
Execs, Engineers and IT

Prepare for BIG DATA
As Big Data matures, so too will technology and
security to support its use.
BY JIM ROMEO
F
or a major oil and gas explora-
tion company using complicated
algorithms to compute millions
of data points taken with sophisticated
sound waves, technology is vastly im-
proving their ability to predict natural
resource opportunities in their ex-
ploration efforts. They conduct these
investigations from a ship at sea and
use supercomputers to analyze it in a
way that provides them with a predict-
able course for their exploration and
discovery.
Welcome to the strategic age of
Big Data — a growing method of
building business. Big Data computes
A MATLAB dashboard for visualizing data from vehicle fleets, which
huge amounts of data points for ap-
incorporates real-world driving data with engine maps created in test cells.
plications that range from condition
Image courtesy of MathWorks.
monitoring at a sprawling chemical
process plant, to calculating weather
patterns in regions where shipments ing sure that the tests they run are Internet of Things
travel. Big Data is on every busi- capturing all of the conditions the The Internet of Things (IoT) is ac-
ness leader’s mind as companies use design may encounter in the field,” celerating the use and application of
the power of information to better says Seth DeLand, product market- Big Data. Sensor technology and its
compete globally. Tantamount to this ing manager of Data Analytics at reliability has improved greatly over
effort is a parallel concern of main- MathWorks. “More data is going the past decade; data attained may
taining data integrity and security in to be available to design engineers. be fed into user interfaces and soft-
an often risky global environment of They’re going to have more and ware computing programs all the way
hacks and cybercriminals. more opportunities to leverage real- around the world via the Internet. As
“Big Data can be fed into many world data to drive decisions during sensor technology becomes more ad-
parts of the design process. For ex- the design process. Engineers are vanced, so too do the possibilities for
ample, data from real-world opera- also going to see their companies Big Data applications.
tions of equipment is being used as looking to build new services based “One area where it is becoming
inputs and boundary conditions for on their data. For example, equip- prevalent though is in sensor sys-
simulations. We’re also seeing orga- ment manufacturers are looking to tems,” says Mike Haley, senior direc-
nizations looking to improve their provide predictive maintenance as a tor, Emerging Products and Tech-
test coverage with Big Data by mak- service to their customers.” nology at Autodesk. “In this case,

physical products — buildings, ma- software engineers. Customers will connected devices are going to gen-
chines, batteries, etc. — are instru- aggressively seek solutions to capital- erate a tremendous amount of data,
mented with sensors that track as- ize on this. in some cases hundreds of terabytes
pects of their performance over their “The last component of the IoT or petabytes. Some of our custom-
lifetime. By downloading and cor- infrastructure is Big Data and the ers are looking at new ways to store,
relating all of this information using cloud,” says Milhe. “Whether per- index and search vast amounts of data.
Big Data technologies, it is possible formance data is sampled every mil- Some companies are using predictive
to give designers more informed de- lisecond or every hour, 200 billion analytics and trending and will apply
sign choices based on real-world per-
formance statistics.
Haley points out that certain in-
dustries adopt data picked up via IoT NEW VERSION
better than others. The utilization of
such data and its incorporation into
business strategy isn’t fully mature.
There’s still a way to go.
“Specific industries where this is
becoming prevalent include: build-
ings — sensors in concrete and
beams as well as components such as
HVAC (heating, ventilation and air
conditioning) systems; battery tech-
nology — understand battery chem-
istry and performance in difference
usage scenarios; and automotive,
which probably has the longest his-
tory of this,” says Haley. “That said,
true Big Data analysis of sensor data
is in its earliest stages as a large vol-
ume of data has to be acquired over
time then appropriately analyzed to
discover patterns and predictors that
will be useful to designers.”
For design engineers and compa-
nies relying on technology to drive
success, IoT and Big Data relies on
several things. “To understand the
implications of Big Data on engineer-
ing firms, it’s useful to first examine
the components of IoT and how each
will affect design engineers,” says Ray
Milhe, vice president of Enterprise Over 100 new features & Over 500,000 registered users
Solutions for ANSYS Inc. “Design- improvements in Origin 2015! worldwide in:
ing products within this framework ◾ 6,000+ Companies including
will have unique implications on en-
FOR A FREE 60-DAY EVALUATION, 120+ Fortune Global 500
GO TO ORIGINLAB.COM/DEMO ◾ 6,500+ Colleges & Universities
gineering firms.” AND ENTER CODE: 7853 ◾ 3,000+ Government Agencies
Milhe says that by 2020, industry & Research Labs
experts estimate there will be as many
as 200 billion wireless connected de-
vices. This will provide an enormous
onslaught of data. Harnessing this 20+ years serving the scientific
data is both a challenge and an op- & engineering community
portunity for design engineers and

Focus on Big Data /// Security
it to simulation and test data, which policies and management.” of embedded software. Engineering
they can use internally or as part of a Big Data and cloud storage envi- firms will not be able to create, test
service to their customers — for ex- ronments, as Lowans notes, are likely and certify this volume of code using
ample — preventive maintenance. to fuel IT investments in the short traditional, manual methods and still
run. Research recently presented by get to market on time. They will have
Security Concerns CompuCom based on its survey re- to adopt model-based software devel-
The IoT and more prevalent data has sults revealed that 26% of IT profes- opment tools to design and simulate
cultivated a concern about all data, sionals expect their organizations to their embedded systems and then use
including Big Data, in cybersecurity focus most of their 2015 technology certified code generation to produce
circles. Data in all forms is vulner- investments on cloud technologies, the actual embedded software.”
able to hacker attacks and breaches followed by security (24%), and Big Ultimately, leveraging Big Data as
of security. Security poses a serious Data/analytics (23%). part of the design process is some-
business risk for those enterprises “Enterprises understand the enor- thing firms should start preparing
hoping to capitalize on Big Data in mous potential of the cloud to inte- for now on a smaller scale so their
the future. A breach of security can grate social, mobile and Big Data,” resources can evolve.
lead to a tacit hijacking of a compa- says Sam Gross, chief technology “The astute engineer and progres-
ny’s intellectual property; a company officer at CompuCom. “With the sive design firm will start small, invest
may fall target to industrial espionage agility challenges that IT faces, we in currently available comparative
and other cybercriminal acts that en- are not surprised that cloud [comput- information about critical assets or
danger their business operations. ing] is top of mind, even as security equipment and engage experts in nu-
Businesses are struggling to keep up remains an ongoing consideration. merical and statistical methods to put
with the needed security to protect Secure cloud-scale application and that information into appropriate de-
volumes of data across its many dis- mobility solutions both drive and sign context,” says Roy Whitt, senior
parate data silos. consume Big Data, and it is no sur- vice president and general manager
According to research firm Gartner prise that the focus on cloud, security for Asset Answers for Meridium Inc.
Inc., more than 80% of organizations and Big Data technology investments Whitt posits that his company
will fail to develop a consolidated data rises to the top.” will ease into Big Data computing
security policy across silos, leading to allow for development of effective
to potential noncompliance, security Continuing Trends techniques to utilize the ever-increas-
breaches and financial liabilities over As Big Data investment continues to ing amounts of data sure to come over
the next few years. At its Security and rise to the top of IT spending, tech- the next decade.
Risk Summit last year, Gartner stated nology should become even more “Identifying a few key or critical
that while much data is at risk, the ad- advanced and reliable continuing a pieces in a design that can be im-
vent of Big Data and its interchange trend where smart devices, remote proved if optimized in the short term
via cloud computing is fueling the monitoring and data transmission will offer an early competitive advan-
overall concern for data security. have been vastly improved. This will tage, while positioning the individual
“Businesses have traditionally fuel design engineers to seek more engineer and design firm for success
managed data within structured and advanced software tools and technol- in the future,” he says. DE
unstructured silos driven by inher- ogy to utilize these capabilities and
ent requirements to deploy relational optimize the performance of data on Jim Romeo is a freelance writer based in
database management systems, file the road ahead. Chesapeake, VA. He can be reached via
storage systems and unstructured According to Milhe, the future JimRomeo.net. Send comments on this ar-
file shares,” said Brian Lowans, prin- computation of more data and better ticle to [email protected].
cipal research analyst at Gartner. data will rely on software develop-
“However, the advent of Big Data ment tools more than ever. “Over the
and cloud storage environments is past 10 years, we’ve all seen devices INFO ➜ ANSYS: ANSYS.com
transforming the way in which data get smarter. Embedded software is ➜ Autodesk: Autodesk.com
is stored, accessed and processed, what makes devices smart, and there ➜ Gartner Inc.: Gartner.com
and CISOs need to develop a data- has been an explosion of the lines of
centric security approach. Unfortu- code in today’s products,” he says. ➜ MathWorks: MathWorks.com
nately this is not common practice “We see this trend continuing. It ➜ Meridium Inc: Meridium.com
today, and its planning is critical to will be commonplace for discrete For more information on this topic,
avoid uncoordinated data security products to have millions of lines visit deskeng.com.

Simulation Lifecycle Management /// Focus on Big Data
Simulation Lifecycle
Management’s New Mission
The discipline takes on new challenges from IoT, apps
and the cloud.
BY KENNETH WONG
M
atteo Nicolich, product manger for Enterprise
Solutions at ESTECO, identified one of the
hidden costs of democracy. “By democratizing
simulation, you let more people run simulation, so more
data will be generated,” he says.
The push to democratize simulation and the spread of de-
sign optimization has fundamentally changed simulation lifecy-
cle management (SLM). It may have initially been developed as
a version- and history-tracking tool, with some job queue man-
agement utilities on the side. But the sheer size of the source
files, the number of iterations involved and the simultaneous
evaluation of hundreds of design variants quickly transformed
SLM’s mission. It’s now tasked with IP (intellectual property)
At its LiveWorx conference this May, PTC demonstrated
guardianship, process automation, HPC (high-performance
the digital twin of a Santa Cruz mountain bike. This virtual
computing) management, and remote visualization (for a start).
model incorporates data streamed from sensors attached
When the Internet of Things (IoT) arrived on the scene,
to the physical model. Image courtesy of PTC.
it also brought its own Big Data headaches. Industry watcher
Gartner predicted “4.9 billion connected things will be in
use in 2015, up 30% from 2014, and will reach 25 billion by assumptions for input, for values like loads, electrical
2020.” The volume of data from these connected devices and charges and aerodynamics. With industrial IoT, we have an
products — heart-rates reported by smartphone health apps, opportunity to use real-time data, measured and collected
climate data uploaded by installed wind turbines and engine by thousands of products in the field. At best, analytics and
performance data collected from moving vehicles, to name measured data from a device can only predict when some-
but a few — represent new challenges for SLM. thing is going to happen, given the existing configuration
SLM’s new challenges are also its greatest opportuni- of the device and its operating environment. They can’t tell
ties. The real-time data available from connected devices, what will happen if you change these parameters in a sub-
well-defined processes captured as apps, and the (almost) stantial way, or how to optimize the performance of your
infinite computing power available on demand is about to product. This is where simulation fits in. It elevates Big
catapult simulation to new heights. Data from being predictive to prescriptive,” he says.
The source data for simulation tends to come from
The Impact of IoT controlled physical tests, covering only a limited use of
“Simulation data is massive; device data is much more the product under idealized conditions. For example, an
transactional. SLM software has to address that difference. automotive simulation might be based primarily on a pro-
They’ll need other database strategies that are more trans- fessional driver’s maneuvers in a closed course. However,
actional,” says Todd McDevitt, marketing director, ANSYS. with real-time data, the range and variety are almost in-
The database challenges notwithstanding, McDevitt finite. “Now, with real-time data, we can simulate what
foresees new types of simulation made possible by device happens with when a weekend warrior or a soccer mom is
data. “Today, simulation users use lab data and sometimes behind the wheel,” says McDevitt.

Focus on Big Data /// Simulation Lifecycle Management
conducted in software programs like
Modelica. In such system models repre-
sented in high level abstraction, the elec-
tromechanical assemblies simulated in
CFD (computational fluid dynamics) and
FEA (finite element analysis) programs
are considered subsystems. SLM (which
governs CFD and FEA data) and system
modelers have in the past remained apart
from each other. But the pursuit of IoT
may bring the two closer.
“We don’t believe these two types of
simulation need to be kept in their own
silos,” says Shankar.
Similarly, at ANSYS, “We see them
[system modeling and FEA simulation] as
coupled simulation. We have a technol-
ogy that lets you create a reduced order
model out of your detailed 3D model, and
use that for your system level representa-
Visualization of results without the need for specialized tion,” McDevitt says.
post-processing tools enables non-experts to quickly ANSYS’ system modeling program Simplorer is de-
access critical insights that simulations can provide. scribed as “an intuitive, multi-domain, multi-technology
Image courtesy of Siemens PLM Software. simulation program that enables engineers to simulate
complex power electronic and electrically controlled sys-
But there’s also the possibility that real-time IoT data may tems.” ANSYS customers may use ANSYS Workbench as
spawn “digital twins that mimic the real physical entities the the integration platform to bring together system model-
customers are using,” says S. Ravi Shankar, director of Global ing and finite element analysis.
Simulation Product Marketing at Siemens PLM Software. “As
the real product goes through its lifecycle, the IoT data com- Juggling Private and Public Cloud
ing from the field and reflective of wear and tear can be used Among small- and mid-sized businesses (SMBs), SLM
to keep the digital twin synched to the real object,” he says. may also be in the midst of a transition from in-house
With such a setup, manufacturers could, for example, data centers to remote clusters and on-demand cloud re-
periodically conduct fatigue analysis on the digital twin sources. In the long run, the public cloud levels the play-
of a bike, not only to understand its current state, but to ing field with its pay-as-you-go pricing and no-upfront-
predict when it might break or fail. The concept has been investment proposition. But in the transition period, SLM
prototyped and presented by PTC at its LiveWorx confer- may need to straddle both on-premise clusters and on-
ence; an event devoted to the company’s IoT offerings. demand cloud. The hybrid approach lets companies har-
In a live demonstration at the event, Mike Campbell, ex- vest the hardware they have already invested in, but also
ecutive vice president of CAD products at PTC, showed a access additional horsepower from outside to address peak
digital twin replicating the movements and mechanical behav- demands and overflows.
iors of a mountain bike from Santa Cruz Bicycles, using wheel “We make our solvers available on Rescale, an on-de-
speed, pedal cadence, pressure on suspension and steering mand cloud simulation platform. Customers have the op-
angles reported by sensors mounted on the bike in operation tion of using their existing licenses for the Siemens solvers
in the field. The setup used real-time sensor data from PTC’s or using licenses provided via Rescale on a pay-per-use
ThinkWorx to animate the 3D CAD model of the bike con- basis. The advantage for the customer is the scalability of
structed in PTC Creo Parametric software. The digital twin the IT platform,” Shankar says.
was presented in an iPad augmented reality viewer app. Though when looking at SLM, companies should
also consider its broader implications. “You can’t talk
Feeding into about SLM without talking about process management
System-Level Simulation and HPC management. HPC clusters have to be man-
The sensor data that must be incorporated into simula- aged and monitored to keep track of the jobs’ progress.
tion is usually the domain of system-level simulation, Our customers simulate very large sub-systems and com-

Simulation Lifecycle Management /// Focus on Big Data
plete products. Workflows can involve several groups dis- Capturing the Process in an App
tributed around the world. Organizations need tools to In the past, developing repeatable simulation processes
coordinate the process as well as manage the data,” says and protocols was just a prudent way to conduct business.
McDevitt. But there’s an added benefit. “Mapping out your process
In May, ANSYS released ANSYS Enterprise Cloud, and codifying your simulation protocols takes some time
which allows businesses to integrate public cloud re- and energy away from regular work, so you might think
sources into their simulation workflow. Users may access it’s secondary, but to develop a SimApp and gain its ben-
and manage the in-house and remote computing resources efits requires a streamlined process,” says Juan Betts, man-
through ANSYS EKM, configurable for both individual aging director of Front End Analytics.
users and collaborative teams. A simulation app — or SimApp, in Front End Ana-
lytics lingo — is usually built on top of general purpose
Remote Visualization is Essential simulation solvers. With limited input fields and guided
Because of the size of the models involved in large-scale steps, they have become one of best ways to make soft-
simulation runs, most experts recommend an IT setup ware-driven simulation more accessible to the non-ex-
and workflow that avoids or minimizes data movement. perts. Front End Analytics specializes in developing and
“Downloading and uploading simulation data involves a deploying SimApps. “There have been cases where we had
lot of wasted time and resources. So you want the data to to diplomatically tell our customer that their processes are
reside in once place where people can access it and view it. too ad-hoc to create a SimApp. We have then worked with
Remote visualization is critical,” McDevitt says. these customers to mature their processes to the point
This method can also minimize data transfer. “When where we can create a SimApp for them,” says Betts.
you take on design optimization, you need a lot of com- The company can build apps directly on enterprise
puting power and also generate lots of data. With powerful SLM systems such as SimManager (from MSC Software)
remote visualization tools, there’s no need to transfer the or SIMULIA SLM (from Dassault Systèmes). It also uses
data back to your local systems,” says ESTECO’s Nicolich. EASA’s app-building platform to serve the clients. “If you
DISCOVER BETTER DESIGNS. FASTER.

FLOW − THERMAL − STRESS − EMAG − ELECTROCHEMISTRY − CASTING − OPTIMIZATION
REACTING CHEMISTRY − VIBRO-ACOUSTICS − MULTIDISCIPLINARY CO-SIMULATION
[email protected]
www.cd-adapco.com
Desktop Engineering advert August 2015.indd 1 deskeng.com /// August 2015 DESKTOP ENGINEERING 2114:05
25/06/2015
de0815_Focus_Simulation_Data_Wong.indd 21 7/16/15 1:35 PM

Focus on Big Data /// Simulation Lifecycle Management
have processes that don’t change a lot, your simulation “For design optimization or design of experiment, you
steps are established, and data management is important may not need to keep the resulting models. You can just keep
to you, then we can probably build our apps directly on the input parameters, output responses, and key performance
your SLM system,” says Betts. “But the drawback with indicators. If you need to, you can run the simulation job
that is, the ‘app-ification’ capabilities in SLM systems are again. Otherwise you end up keeping lots of useless data
still very limited. The variety and range of apps you can you that you won’t look back,” says Nicolich. In ESTECO’s
build in something like EASA is much higher.” simulation optimization platform, “You can set an expiration
The best-case scenario, Betts pointed out, is where date on a data set or a model. So even if you forget to delete
the app-building exercise leads to well-defined simulation it, when the expiration date comes, the system automatically
processes in a company. “Because when you’re architect- removes the data,” he says.
ing an app, you’re also architecting a process,” he says. Balancing the archival needs and the preference for data
cleanliness is more difficult for those who manufacturer
Keeping the Data Clean and Lean products with longer lifecycles. “With aircraft, you may need
As a way to curb the exponential growth of SLM data, to support the analysis of a model designed 50 years ago.
many experts recommend carefully selecting the type of With a wearable consumer device, the data is obsolete two
data to archive. “Even if your model is just a few hundred years later,” says Siemens’ Shankar.
megabytes, in a complicated process that involves a few Cloud storage offers a viable alternative to those who
hundred simulation runs, you will generate terabytes of prefer to archive SLM data for decades. But for some, secu-
data,” says Nicolich. rity concerns still remain a barrier. “The reluctance to go to
Even for large enterprises with considerable financial the cloud for a long time had to do with data security. That
muscle, accumulating and maintaining terabytes of data attitude may be changing now, especially among small- and
for every project is not a pragmatic solution. The cost and mid-sized companies. But ironically, the big guys are the
IT burden of storage aside, the sheer volume of data would ones who generate those large-volume simulation data, and
make analysis and reuse impossible. they are usually the most reluctant to go to the cloud,” says
Christine Wolfe, lead product manager for multiphysics of-
ferings at ANSYS.
SLM is relatively young, certainly younger than its
cousins PLM (product lifecycle management) and ERP
(enterprise resource planning). The rapid expansion of
SLM offerings in the last five years or so reflects manu-
facturers’ increased reliance on digital prototyping and
simulation. Deployed with tools for HPC management,
remote visualization and IP security, SLM could help
consolidate ad-hoc simulation jobs into a well-defined
company-wide infrastructure for simulation-driven de-
sign. It should be treated as part of your simulation strat-
egy, not a necessary evil. DE
Kenneth Wong is Desktop Engineering’s resident blogger

and senior editor. Email him at [email protected] or
share your thoughts on this article at deskeng.com/facebook.
INFO «ANSYS: ANSYS.com

«ESTECO: Esteco.com
Take just a minute to answer 5 quick «Front End Analytics: FEASol.com
questions to test your 3D Printing «Modelica: Modelica.org
Knowledge for a chance to win $1,000! «PTC: PTC.com
Go To: www.deskeng.com/de/best «Rescale: Rescale.com
«Siemens PLM Software: Siemens.com/plm
Sponsored by: For more information on this topic, visit deskeng.com.

Augmented and Virtual Reality /// Prototype
Blending Reality
with Pixels
The new age of
augmented- and
virtual reality-
integrated design
tools has humble
beginnings.
BY KENNETH WONG
T
wo years ago, while attending zSpace’s display tablet with stereoscopic glasses and virtual cutting tools
Dassault Systèmes’ SOLID- allow those outside engineering and manufacturing to interact with 3D
WORKS World user confer- technology. Image courtesy of zSpace.
ence, Bob Conley saw a demonstra-
tion of the eDrawings mobile app, a engineering tool, not a purposeless acquisitions to cater to early adopters
CAD file viewer for tablet users. What novelty. The adoption of augmented of AR and VR.
won him over was the app’s augmented reality may be slow among small
reality (AR) function, which enabled businesses with limited expenditure Augmented Reality to Go
users to project CAD models into a for experimentation. However, ac- For Conley, AR is more than show-
live feed from the iPad’s camera view. cording to CCS Insight, an analyst manship and razzle-dazzle. Some
So he promptly purchased an iPad and firm that tracks mobile communica- parts are physically too big or bulky
the eDrawings Pro app from iTunes. tion and Internet commerce, AR is for him to haul to the client’s site dur-
The $810 investment ($800 for the de- “starting to gain traction in indus- ing design reviews. For those, Conley
vice, $10 for the app) laid the ground- trial and business arenas … Europe is used to rely on onsite measurements
work for AR in his design consulting currently the largest test bed for this and some educated guesses to deter-
business, Interactive CAD Solutions. technology, with numerous blue-chip mine how the designed part would
“In the week after I got the app, companies across all sectors evaluat- fit into the client’s facility or existing
I generated more than $20,000 in ing its capabilities.” machinery setup. These days, he uses
sales,” Conley says. On his business’ CCS Insight predicts devices his business card as the digital replica
home page, Conley offers a down- powered by AR and its twin VR of the new part.
loadable PowerPoint deck with a bul- (virtual reality) could become a “$4 “My two-fold business card has
let list of services. Augmented reality billion-plus business in three years.” a global marker for [the] eDrawings
is on the very top of that list. With the promise of such real money, iPad app,” he said. “I just put down my
Conley belongs to a small but some hardware and software vendors card where the part was supposed to
growing group of CAD users who are already shifting their R&D strat- go, and check interferences.” The app
see AR as a legitimate design and egies and IP (intellectual property) recognizes the marker as the place-

Prototype /// Augmented and Virtual Reality
holder for a specific digital model;
therefore, through the tablet’s camera
view, the app projects the digital part
into the real world. Through XYZ co-
ordinates, the digital part’s orientation
is aligned to match the physical envi-
ronment where the marker sits.
In early experiments with eDraw-
ings’ AR tools, Conley learned a valu-
able lesson about the computer-rec-
ognizable markers. “When I printed
the global marker on my card the first
time, I shrank it to fit the card,” he re-
called. That, he found out, caused his
digital model to appear in a smaller
scale in the app’s window.
Being portable and lightweight,
the tablet offers a tremendous advan-
tage to Conley. It allows him to deploy
AR at any given location. On the other
hand, the device’s limited memory ca-
pacity proves to be a current draw- Bob Conley from Interactive CAD Solutions uses iPad-powered augmented
back. “One of the biggest challenges reality to check how SOLIDWORKS models would fit into the client’s
is working with large models,” says machinery. Image courtesy of Interactive CAD Solutions.
Conley. “I haven’t found a way to
overcome that yet.” clients give Conley files that are not The Heart Beats Faster in Stereo
eDrawings displays SOLID- supported by eDrawings. Conley’s so- At the NAFEMS World Congress in
WORKS models and industry stan- lution is to use GrabCAD Workbench, June, Steve Levine, Dassault Systèmes’
dard neutral 3D formats (such as an online collaboration platform, to SIMULIA chief strategy officer, stood
STEP or STL). But sometimes the convert them to the right file format. in an exhibit booth with a pair of ste-
reoscope glasses dangling on his nose.
He invited visitors to dissect a virtual
heart, known as the Living Heart Proj-
Augmented Reality vs. Virtual Reality
A
ect, loaded into the zSpace virtual re-
ugmented reality (AR) uses special displays to inject a ality display as a 3D model.
layer of digital objects and information into the user’s According to Levine, what stereos-
observable reality. Google Glass is a good example of copy offers is a sense of scale and the
AR-capable hardware. It could enhance your view of the surround- ability to experiment without jeopar-
ings by streaming digital data relevant to the site, such as interac- dizing a real patient. “If you show sur-
tive maps and Wikipedia articles. geons a computational fluid dynamics
Virtual reality (VR) uses special displays to let you experi- (CFD) model or a finite element anal-
ence an artificial digital environment. A good example of this is ysis (FEA) model, they have a tough
the Pure Land exhibit from the City University of Hong Kong. The time grasping it,” he said. “But if you
exhibit lets you take a virtual tour of an ancient cave adorned show them this, they got it instantly.
with Buddhist murals. In this case, the artificial environment is They’d take the heart apart, look at
a digital replica of the real site in Dunhuang, China; however, VR its cross-sections, and talk about how
environments could also be completely fictional. they might have operated on a particu-
VR content is created by capturing physical environments in lar patient’s heart.”
digital form or modeling digital worlds from scratch. VR envi- Dassault Systèmes, a 3D software
ronments might also be incorporated into AR experiences. For powerhouse, is betting heavily on the
example, while a factory worker inspects the physical plant, the notion that, in the future, consumers
detailed CAD model of the same plant could be streamed to his or will want to pay for experiences (a
her Wi-Fi connected AR eyewear. combination of hardware, software,

Set your product apart.
apps and services); therefore offering you would feel with another medium. Make it smart.
them products alone will no longer be If you’re designing a car, for example,
sufficient. Life sciences is one of the you’d be able to judge if the door FlexiForce® sensors
new markets the company is actively handles are in the right places,” says are a great fit for your
courting. Levine and his colleagues Christopher Nichols, creative director product design!
recognize a high-end CAD program at Chaos Group.
like CATIA would be more an obstacle You can explore VR environments Our sensors provide
than a tool for the surgeons and medi- in regular flat panel displays, using the force feedback that
cal professionals. The learning curve mouse and keyboard for path find- enables you to design
alone would be discouraging to those ing. However, experiencing VR con- smart, differentiated
who are outside the engineering and tent through immersive devices like products.
design disciplines. The Living Heart the Oculus Rift or Samsung Gear VR
on zSpace is a tailor-made solution, gives you a more natural way to inter-
comprising a detailed 3D heart model act with the environment, also a better A20
loaded on a portable stereoscopic tab- understanding of the volumes and dis-
1 Se
let with virtual cutting tools. It’s the tances represented by the 3D models. n
so
type of offering that can help Dassault Using VR for product develop-
r
Systèmes break into new markets be- ment, however, adds a greater burden
yond traditional manufacturing. on the content creator, the one who
produces the digital prototype or en-
From Entertainment vironment that will be uploaded to the
to Engineering VR goggles. The dimensions of the
“VR is one of the most exciting things objects in the VR world have to be
to happen to CG in the last decade,” geometrically precise, not approxima-
says Vladimir Koylazov, co-founder tions (as they might be in VR-powered
and lead developer for V-Ray render- videogames and movies).
ing software at Chaos Group. He and Currently, there’s no straightfor-
his colleagues are laying the ground- ward way to create VR content in
work for the era of stereoscope VR CAD software; therefore, VR con-
content. In June, the company an- tent will most likely be created by
nounced the release of V-Ray 3.2 for importing CAD-authored 3D con-
3ds Max, a free upgrade that includes tent into rendering programs like
two new VR camera types to “render 3dx Max or Maya. FlexiForce Force
stereo cube maps and spherical stereo For the novice VR content cre- Sensors are:
images for VR headsets such as Oculus ators, Nichols offered a tip. “You have
Rift and Samsung Gear VR.” to pay attention to the intraocular • Paper Thin
The technology offers a more in- distance [the distance between the
• Flexible
depth look and added element of in- virtual observer’s eyes, represented by
teractivity that traditional designs may the distance between the two cameras • Lightweight
not. “With VR, you can make informed in a stereoscopic rendering program]. • Economical
decisions about whether something Getting that distance wrong is like • Customizable
would work correctly,” says Lon Grohs, shrinking or expanding the [virtual
chief commercial officer at Chaos observer’s] head. If that distance is too
Group. “Virtual prototypes in VR are big, then the world looks a lot smaller; VISIT WWW.TEKSCAN.COM/ES
so much more realistic than what peo- if it’s too small, the world looks too
ple have seen before. I could inspect the big,” he says.
virtual prototype as if it were the real
thing, before the real thing is available.” Cognitive Computing Coming Up
Additionally, stereoscopic content “We’re moving from cloud comput-
offers a sense of scale, a critical factor ing to cloud, mobile and wearable
in all engineering judgments. “If you computing — the convergence of the
walk into a VR atrium, it’s far closer three,” said Ahmed Noor, a scholar
to walking into the real atrium than from Old Dominion University, as

Researchers may use Dassault

Systèmes’ 3DEXCITE technology to
view an accurate heart model inside
a virtual cave environment.
Image courtesy of Dassault Systèmes.
part of his keynote speech at the NA- “Gesture control with AR appli- position for its current and soon to be
FEMS World Congress 2015. cations and managing the ambient released products,” says Paul Travers,
He envisioned a future where light, especially in optical see-through CEO of Vuzix, in his explanation of
designers and engineers might work glasses, is critical in the operation of the IP acquisition.
with “cognitive systems that mimic wearable display technology and es- According to the patent office’s pub-
the way humans work through natu- pecially augmented reality in smart lic records, Patent 6243054 is described
ral language processing, data min- glasses. Vuzix acquired these addi- as “a computer system [that] stereoscop-
ing and pattern recognition.” This, tional patents to secure a stronger IP ically projects a three dimensional object
according to Noor, is the precur-
sor to the next phase, “anticipatory
Gaming Reality
V
computing, in which cyber-physical
systems can recognize the user’s irtual and augmented reality technologies are being driven, in large
needs by watching the user, and pro- part, by the computer gaming industry. AMD and NVIDIA are no
vide them with tools without being stangers to migrating technologies from the entertainment to the
asked.” Whereas today’s human- professional design industries. Both have virtual reality efforts underway.
machine interaction is limited to ro- According to the company’s website, “LiquidVR is an AMD initiative
botics and software, Noor thinks we dedicated to making VR as comfortable and realistic as possible.” The
will eventually interact with “multi- company is focusing its VR efforts on what it calls “presence,” or how
model virtual holography.” immersive the virtual world seems to be.
Some groundwork for Noor’s vision NVIDIA’s approach is called GameWorks VR, which it describes as
may already be on the way. This April, “a set of APIs, libraries, and features that enable both VR headset and
Vuzix Corporation, which supplies video game developers to deliver amazing VR experiences.”
eyewear and smart glasses, snatched The companies’ solutions could go a long way toward reducing the
up U.S. Patent numbers 6243054 and hardware requirements needed for immerseve VR and allowing soft-
6559813 for an undisclosed sum. ware developers to make it more accesible to other industries.

having an interface image in a space ob- product HP Sprout, HP introduced

servable by a user. The user controls the another term — blended reality (BR). Kenneth Wong is Desktop Engineer-
movement of a physical object within At its core, HP Sprout may be an all- ing’s resident blogger and senior editor.
the space while observing both the three in-one multi-touch computer, but the Email him at kennethwong@deskeng.
dimensionally projected object and the mounted projector and a camera, a com or share your thoughts on this ar-
physical object.” paper-like surface for projection and ticle at deskeng.com/facebook.
Patent 6559813 is “a virtual reality sensors offer multiple methods of in-
system [that] stereoscopically projects teracting and capturing content, both INFO ➜ AMD: AMD.com
VR images, including a three dimen- in 2D and 3D.
sional image having an interface image The emerging AR, VR and — to ➜ CCS Insight: CCSInsight.com
in a space observable by a user. The borrow HP’s vocabulary — BR de- ➜ Chaos Group: ChaosGroup.com
display system includes a substantially vices will likely colonize the general
➜ Dassault Systèmes: 3ds.com
transparent display means, which also consumer market and entertainment
allows real images of real objects to be industry before they find acceptance ➜ HP: HP.com
combined or superimposed with the in the more conservative engineer- ➜ Interactive CAD Solutions:
virtual reality images.” Both patents ing market. But when they do, design IntCADSol.com
promise “a way to contact user inter- and engineering software vendors may ➜ NVIDIA: NVIDIA.com
face images without contacting a key- confront one of their biggest chal-
board or a mouse or the display itself.” lenges. Adapting legacy CAD software
➜ SOLIDWORKS eDrawings:
eDrawingsViewer.com
code that relies on the mouse and
Rethinking the Human- keyboard to a new breed of portable, ➜ Vuzix Corp: Vuzix.com
Machine Interface wearable, head-mountable computers ➜ zSpace: zSpace.com
Finding the prevalent terms AR and that invite touch, natural speech, and For more information on this topic,
VR inadequate to describe its new gesture input won’t be easy. DE visit deskeng.com.
We Said We Meant
Fast Fast
Instant Mold Quotes

®
ICOMold’s instant injection mold quoting system
is so fast your parts will be racing to the finish line.
www.icomold.com © Copyright 2015 ICOMold®

VR Speeds Design
The FIVE Lab uses a virtual reality headset in

conjunction with a specialized glove to con-
trol designs. Images courtesy of Ford.
The technology gives engineers at Ford and other companies an early look at designs.
FRANK J. OHLHORST AND JESS LULKA
F
ord Motor Company is leveraging the precise location and orientation of the ence as “uncanny, and feeling just like
virtual reality technology to elimi- wearer. The goggles allow users to observe sitting in a real car.” She says the level
nate many of the design challenges realistically interpreted CAD models of a of detail is extraordinary. Ford uses the
faced by today’s automobile manufactur- vehicle, which can be placed in a variety virtual rigs to test for quality, engineer-
ers. The company, which says it is the first of different environments. The users can ing issues, fit and finish of the CAD
automaker to use ultra high-definition walk around the virtual vehicle, as if it was designs long before engineers move to-
immersive virtual reality, has made im- parked in an actual environment, such as a ward creating a clay model or any other
pressive strides in speeding up the design new vehicle showroom. physical prototype.
process and prototyping products without “We now have Ford designers and
having to physically construct them first. engineers around the world working A New Approach to Design
The secret behind the company’s suc- together virtually — inside and side- Baron describes the virtual lab as a col-
cess comes in the form of the Ford Im- by-side — on the same product,” says lection of high-performance computing
mersive Vehicle Environment (FIVE) Elizabeth Baron, Ford Virtual Reality (HPC) components, which are integrated
Labs. The FIVE Labs function as virtual and Advanced Visualization technical with advanced simulation software ele-
vehicle prototyping rooms. They include specialist. “By using this technology, de- ments to create a 3D virtual world that
a dummy car rig with a seat and steering signers and engineers can quickly tran- allows designers and engineers to expe-
wheel, an 80-in. 4K monitor and a com- sition from one car design proposal to rience automotive designs. Simply put,
puter rig. another, and they can study and identify by using a totally immersive virtual envi-
Users wear a pair of VR (virtual real- which is the best option.” ronment, Ford’s designers and engineers
ity) goggles and a glove, each of which are Users can also sit down in the test can experience a vehicle the same way a
monitored by 19 motion tracking cam- rig and experience the interior of the customer would in a showroom.
eras dotted around the walls to capture vehicle. Baron describes this experi- Beyond the immediate benefit of vi-

MEASUREMENT COMPUTING
High-Speed Conditioned
Measurements with
sualizing design changes before actual several other solutions to design prob- Channel-to-Channel
production, the virtual lab offers many lems before any production or physical Isolation
other advantages. For example, engineers prototyping took place. Those improve-
collaborate in real-time, simultaneously ments ranged from the placement of • Easy to Use
experiencing the same point-of-view. pedals in right-hand drive vehicles to the • Easy to Integrate
By creating a shared 3D virtual reality, integration of entertainment systems and • Easy to Support
Ford’s designers and engineers are able environmental controls.
to quickly overcome traditional design
challenges. The lab also allows multiple
The benefits derived from the system
have led Ford to expand the capabilities NEW
designers and engineers from different of the FIVE labs since their introduction SC-1608 Series
geographic locations to work together in as a proof of concept. The transition from
Highlights
real-time on vehicle design, without hav- concept to reality actually took place very
ing to travel to a central location. Ford quickly, even though Baron faced several • Signal conditioning for
has accomplished that by integrating the internal challenges. Thermocouple, RTD,
virtual immersion capability of the virtual “One of the most difficult goals was Strain Gage, Frequency,
lab across multiple locations. getting buy in from various departments. Direct Voltage, and Current
“Thanks to a host of recent upgrades We had to convince stakeholders of the • USB or Ethernet interface
to the lab and its establishment in other viability of the technology and overcome
• Up to 500 kS/s sampling
Ford development centers around the the resistance to adopting virtual design
world, Ford became the first automaker concepts,” says Baron. “However, users • 16-bit resolution
to use the technology on a global level, have quickly realized the benefits offered • Channel-to-channel isolation
thereby enabling simultaneous input and have had a profound realization that • Isolated analog output
from designers and engineers working observing a vehicle’s design using a 1:1 available
on the same product at the same time,” relationship is far superior to looking at
• Isolated digital
says Baron. numbers on an Excel spreadsheet.”
input/output
Throughout 2013, Ford designers
and engineers examined over 135,000 Constantly Improving Technology
details on 193 virtual vehicle prototypes, Ford continues to improve and distrib-
all of which were virtually built in the ute FIVE Lab technology by adding
Immersion Lab. “Such an achievement it to design centers in Australia, Ger-
would have been impossible just a few many, China, India, Brazil and Mexico.
years ago,” says Baron. Recent enhancements to the system
The benefits of 3D immersive design include 4K “power-wall” displays, From Only
validation have become easy to identify which offer four times the resolution
and have been demonstrated by deliver- of a high-definition television, enabling $ 999
ing improvements for the customer in Ford designers and engineers to better
such vehicles as the Ford Fusion and 2015 evaluate and review 3D models of ve- The SC-1608 Series is the
Ford Mustang. For example, engineers hicles in real time, globally. perfect platform for systems
were able to examine the placement of the “We moved to the global One Ford needing flexible conditioning
side view mirror and rear brake light in plan so that international collaboration at a low cost per channel.
the Ford Fusion before creating a physical could lead to the development of globally
representation of the vehicle. appealing vehicles,” Baron says. “With this mccdaq.com/SC1608
In the case of the side view mirror, the technology, designers and engineers can
goal was to optimize visibility without enhance their ability to achieve that goal
impacting the design of the car. The posi- — while also improving vehicle quality.”
tion of the side view mirror went through New software enables engineers to
several virtual iterations before the team study hundreds of elements inside and
decided that a door-mounted design was outside of a vehicle, ensuring that ve-
optimal. The LED center brake light un- hicles coming to market “have been
derwent a similar process, with engineers painstakingly inspected for usability, con-
determining that an ultra-thin unit maxi- sistency and design effects that are now Contact us
mized visibility without losing style. discernible in the sophisticated shadows 1.80 0.234.4232
Baron and her team were able to offer and real-world lighting conditions the ©2015 Measurement Computing Corporation
[email protected]

virtual experience almost indistinguish-
able from a real vehicle.”
In its quest to improve the immersive
experience, Ford has deployed a tech-
nology called the Programmable Ve-
hicle Model, which allows engineers to
be immersed in a lifelike virtual vehicle.
Content evaluators are able to sit in the
Programmable Vehicle Model, which
consists of an interior laid out with a
steering wheel and seats, and be able to
both visualize and feel components.
Engineers can test steering wheels,
door handles and other major touch
points for placement relative to the driver’s
Engineers can place the virtual model in a variety
seat. The Programmable Vehicle Model
of realistic environments, including roads, urban
supports many virtual 3D capabilities, al-
landscapes and vehicle showrooms.
lowing designers who are located outside
of the vehicle in the Virtual Space to col-
tools provide,” according to Ford. designers and engineers with virtual laborate with others, who may be inside
“Other enhancements, such as the movement, regardless of their physi- the Programmable Vehicle Model.
new 4K-resolution power-wall, give cal location,” says Baron. “The system Furthermore, the Programmable
Ford engineers a life-sized view of is- links Ford workers globally to simul- Vehicle Model is integrated with a cave
sues that arise in vehicle development. taneously analyze and inspect the same automated virtual environment (CAVE),
That life-sized view is further enhanced virtual vehicle on a scale not possible which provides a wide field-of-view
by pairing the actual movement of before. The newest software provides a virtual environment, allowing Ford en-
gineers to rapidly test design and place-
ment components to study visibility and
Virtual Reality on the Go customer preference.
W
hile the RAVE CAVE and Ford FIVE Lab are examples of completely Beyond the FIVE Lab
immersive prototyping environments, you don’t necessarily need Even though the equipment used in
an entire building to bring virtual and augmented reality into the Ford’s labs is highly specialized, the ap-
product design process. This year, Microsoft announced the HoloLens — a plications for virtual reality-based prod-
holographic computing platform that is designed to integrate holograms within uct design are vast. Rave Computer,
physical spaces. With this technology, the company hopes to usher in a new era which provides some of technology for
of design that enables engineers to interact directly with the projected models. Ford’s FIVE Lab, is helping companies
The goal of this headset, according to Microsoft, is to provide a product that and government organizations create in
allows users to size, shape and manipulate models with their own hands and a more immersive environment with its
voice recognition. While some of these features are under development, the RAVE CAVE — also known as the Re-
HoloLens is already being implemented in projects with the NASA Jet Propulsion configurable Center for Automated Vir-
Laboratory to explore 3D renderings of Mars and bring virtual reality to the tual Environments.
International Space Station. Located in Sterling, MI, the RAVE
Virtual reality is also finding its way into the consumer market, with compa- CAVE is run as a non-profit organiza-
nies such as Oculus VR and Sony developing their own headsets for gaming and tion through a research partnership
entertainment. This summer, Oculus revamped its Rift headset to have a more between Rave, TARDEC and DC3S.
ergonomic design, AMOLED low-persistence displays and attachable hand con- Through this collaboration, the agencies
trollers. Additionally, Sony’s Project Morpheus promises a completely immersive promote modeling, simulation and visu-
gaming experience with surround audio and encompassing visuals. And while alization technology, creative R&D and
these headsets are focused on enhancing the consumer experience — they also STEM education. In addition to giving
represent increasing availability of the technology — which could mean a greater small- and mid-sized firms access to vir-
potential for eventual integration into the engineering workflow. tual prototyping capabilities, the CAVE
— J. Lulka also provides software companies the

chance to develop and perfect their own
platforms. Current software providers in-
clude Autodesk, Intel, ESI Group, RTT
(now Dassault Systèmes’ 3DEXCITE)
and Mechadyne.
By using the CAVE for product de-
sign and testing, users have access to
scenarios that might not be possible with
a physical prototype. For example, engi-
neers can place an automobile in many
different environments. Within the
CAVE, they can see how the car reacts
to different weather conditions, a process
that would traditionally require exten-
sive physical testing over a long period
of time. However, the CAVE enables To project designs, Ford uses a device that
teams to view all of these scenarios within has a seat, steering wheel, computing
several hours. “It lets engineers get a lot technology and a 4K display.
further along without creating a physical many of the engineers are request-
prototype than they otherwise would be ing additions to the experience, creating Ford plans to build a team of 125 pro-
able to,” says Matt Moy from product de- a rapidly evolving virtual working envi- fessionals under Maciuca, which would
velopment at Rave Computer. ronment. make the company one of the largest
Additionally, since the technology “One of those additions came in the dedicated automotive research teams in
at the CAVE is all encompassing, users form of adding ‘X-ray vision,’ which gives the valley.
can see designs in a whole new light. users of the immersion lab the ability to “We want to create a collaborative
The company recently showcased a see inside and through a vehicle structure environment, an environment that’s
model of the human body to a group of to study how various structural, mechani- very unique to Silicon Valley,” says Raj
physicians. “These physicians were not cal and electrical systems interact within Nair, vice president of global product,
only able to see the body, but place their the architecture. Our new technology is Ford. Indeed, some of the projects that
head [in it] and view the heart from the more refined, and allows us to see and un- are already underway include collabo-
inside out,” says Rick Darter, president derstand complex engineering issues while rations with several universities and
and CEO at Rave Computer. “[The considering aesthetics and design,” Baron technology companies. DE
technology] gives you that immersive says. “We are incorporating light and
capability of seeing underneath, inside shadow calculations in real time, which Frank Ohlhorst is chief analyst and freelance
of, as well as the actual figure of what adds more depth and authenticity.” writer at Ohlhorst.net. Jess Lulka is associate
you’re trying to demonstrate.” Baron is quickly able to highlight the editor of Desktop Engineering. Send e-mail
benefits of the system, including what she about this article to [email protected].
Looking Ahead calls “a holistic approach to vehicle design”
Moving forward, this type of technol- and “a profound realization of vehicle
ogy is expected to become more main- design.” As the system stands now, new
INFO ➜ 3DEXCITE: 3ds.com/3dexcite
stream as firms find more uses for im- vehicle designs can be imported and visu- ➜ Autodesk: Autodesk.com
mersive visualization environments. “If alized in a matter of days, accomplishing ➜ DC3S: Defensec3s.com
you’re an engineer and you’re designing what once took many weeks to create clay ➜ ESI Group: esi-group.com
something, it’s pretty much a no-brainer models and mockups of features.
➜ Ford: Ford.com
to take the opportunity to look at [the Those benefits have led Ford to fur-
product] before building something to ther expand its technological design ➜ Intel: Intel.com
look at. And so as the cost of the technol- capabilities, including the recently cre- ➜ Microsoft: Microsoft.com
ogy comes down, the use of this technol- ated Ford Research and Development ➜ Oculus VR: Oculus.com
ogy is increasing,” says Karl Rosenberger, Center in Silicon Valley. The new center
director of product development at Rave is helmed by Dragos Maciuca, a former ➜ Rave Computer: Rave.com
Computer. Apple engineer with experience in con- ➜ TARDEC: Army.mil/tardec
As Ford comes to rely on virtual im- sumer electronics, semiconductors, aero- For more information on this topic,
mersion as part of the design process, space and automotive technology. visit deskeng.com.

Engineering Computing /// Workstations
Answers to Your Workstation Questions

Insights on how to assemble the right hardware for your design
engineering needs.
BY BRIAN ALBRIGHT
T Q: Are there any common misconceptions about hard-

here are a variety of factors to consider when selecting
the optimal engineering workstation configuration.
ware needs you routinely run across with CAD customers?
Desktop Engineering questioned leading CAD value-
added resellers (VARs) to discover how they guide their cli- A. Teague: People confuse the number of CPUs with single-
ents through the hardware selection process. thread performance. With CAD, even with FEA, single-thread
performance is more important. I’ve seen people get a huge
Q: What factors (such as model size, use of simulation number of cores with a low clock speed. There are very few
and photo-realistic rendering) go into determining a hard- applications that take advantage of a huge number of cores.
ware recommendation for CAD users?
A. Burke: Yes, that all machines are the same. I also think cus-
A. Adrian Fanjoy, technical services man- tomers don’t understand that part complexity is as important
ager, Computer Aided Technology (CATI), as assembly size with regards to performance. I once had a cus-
SOLIDWORKS Reseller: Photo rendering tomer who didn’t understand why his part with over 250 unique
and simulation have a big impact because they individual holes, all at unique angles, all with unique features,
would determine whether or not you go with a was causing a significant degradation in performance.
system with more than four cores. That drives up
the cost of the machine drastically. Q: Do you recommend a multi-core processor for CAD?
After that, we generally look at the size and complexity of the
models and the assemblies themselves. A. Fanjoy: Yes, four cores, and only four cores. No more than
The number of cores when dealing with just SOLID- that. The caveat would be if they are going to use other applica-
WORKS is pretty standard; you go with four cores no matter tions simultaneously with CAD, they might need more cores.
what. What we are really doing is adjusting the amount of RAM
and the processor speed. No matter how big the assembly is, A. Teague: I do recommend multi-core, but not in large num-
you will benefit from an increase in processor speed. bers. For most CAD users a quad-core at a higher clock speed is
all you need. Most users don’t need that many cores. Even with
A. Chris Teague, senior applications engi- FEA, only some of those systems need large numbers of cores.
neer, Saratech, Solid Edge Reseller: Model size Often they are better off with a faster SSD (solid-state drive),
and the size of the assemblies. And then, of course, faster memory and a higher clock speed.
whether they are going to run FEA (finite element
analysis), detailed rendering or complex analysis Q: How do you determine the amount of RAM a CAD
work. Those customers will need more computer horsepower. user should invest in?
A. Matthew Burke, principal and CEO, A. Fanjoy: It’s not too detrimental to over-buy. We don’t rec-
M2 Technologies, Autodesk Reseller: The ommend anybody get under 8GB, because it’s almost a given
two most obvious factors are how large the as- anyway. But do you need 16GB or 32GB? Typically for large
semblies are and how complex they are. Beyond assemblies, 16GB is enough, and for super-large assemblies they
that, are they doing FEA, CFD (computational may need 24GB to 32GB, but those are gigantic assemblies of
fluid dynamics) analysis, or using a CAM package on the same 15,000 to 20,000 components or more.
workstation? More subtle factors include how often they up-
grade their engineering hardware. If a customer has a regular A. Teague: The rule of thumb on that is we don’t recommend
schedule, you might not recommend the highest end machine; anything less than 8GB for anybody. We try to push everybody
but if a customer only upgrades hardware every three to five to 16GB, because that’s a general purpose number. If they are
years, they really need to purchase as much horsepower as they running larger assemblies, then 32GB on the desktop is good.
can afford today. More than that is for guys running dedicated servers.

A. Burke: We try and keep this simple: put as much RAM in
as your budget will allow. Certainly there are published guide- Learn More
I
lines available for both minimums and recommendations, but
they generally don’t take into account the other applications f you’d like to learn more about the best hardware
that are going to be up and running. setup for SOLIDWORKS, download the “Maximize
SOLIDWORKS Performance” white paper pro-
Q: What processes do you recommend be carried out on duced by Desktop Engineering and sponsored by Intel.
an SSD vs. a traditional spinning hard disk? This free, 12-page paper is based on CATI’s in-depth
research and benchmarking of the software.
A. Fanjoy: At this point, the only thing we recommend a tradi- Download “Maximize SOLIDWORKS Performance”
tional hard disk for is massive storage. If you use a PDM (product here: deskeng.com/de/maximizesw
data management) system, and you’re operating with local cache,
and you keep that clean and keep maintaining that so you don’t
overburden it, a small 240GB SSD should be fine for any user. A. Fanjoy: The only time we tell people that a consumer-class
PC is sufficient is if they are using a home-use license to build a
A. Burke: In particular, FEA and CFD benefit a great deal deck at their house. Everything else should use quality graphics.
with an SSD drive. SSDs have come down considerably in
price, so they almost always tend to be worth the investment A. Teague: We do recommend professional workstations, be-
in a CAD workstation because they are simply so much faster. cause of the OpenGL issue. Consumer machines use consumer-
class cards, but DirectX drivers aren’t as well tuned for OpenGL.
Q: In your opinion should a CAD user invest in a top-
of-the-line, mid-range or entry-level graphics card, or are A. Burke: CAD is expensive, and designers and engineers are
integrated graphics sufficient? expensive — give them a machine with the architecture capable
of allowing them to perform their jobs efficiently. To put it an-
A. Fanjoy: That’s really going to depend on what you are doing. other way, don’t ask Jeff Gordon to win at Daytona in an SS
There is a point in just a couple of circumstances where entry- sedan from the local Chevy dealer.
level graphics aren’t going to do what you need. The majority of
our users would get away with onboard graphics just fine, at least Q: Is there a rule of thumb for how long a CAD user
that’s what my testing is showing. Areas where you’d need to look should wait before investing in a new workstation?
at a mid-range card would be with highly faceted models that
have intricate details, and I mean immense amounts of small de- A. Teague: The general rule of thumb is three years, but that
tail. What we’ve found is that as assemblies get larger, the ability can depend on a lot of things. Did you buy at the high end three
of the graphics card to keep up with the data becomes a problem. years ago, or did you get a cheap workstation? After three years
If there’s degradation, you can always lower the level of detail in look around and see what the technology is now vs. what you
the model to accommodate for that and still get smooth rotation. bought. Even if the hardware isn’t obsolete, a fresh install of
Onboard graphics will work for even more situations if you are Windows might make things faster.
willing to sacrifice that level of detail while spinning the model.
A. Burke: Technology evolves at a very rapid pace, but for
A. Teague: We don’t ever recommend integrated graphics, most folks the decision on hardware investment is happening on
because we have had so many problems with them not running a corporate IT level so they have little control over it. If they do,
OpenGL well. All the tools are running OpenGL, and the only we try and tell customers they should be on a two- or three-year
cards that are tuned to that are the professional cards. There’s upgrade cycle. Much longer than three years, we start to see the
no reason to not go top-of-the-line for graphics except for bud- hardware becoming a constraint to the users. DE
get. That’s an easy thing to upgrade later.
Brian Albright is a freelance journalist based in Columbus, OH. He
A. Burke: This is difficult to answer without additional context. is the former managing editor of Frontline Solutions magazine, and
Generally speaking our typical recommendation is a mid-range has been writing about technology topics since the mid-1990s. Send
card unless we know, through qualifying, the type of design and e-mail about this article to [email protected].
work occurring would substantially benefit from a top-of-the-line
card. Usually we say, save money here and invest in more RAM. INFO «Computer Aided Technology: CATI.com
Q: Do you recommend professional workstations for «M2 Technologies: M2T.com
CAD or are consumer-class PCs sufficient? «Saratech: SaratechInc.com

Engineering Computing /// Storage Options
New Horizons
in Data Storage
The ubiquitous hard disk drive’s days are numbered, both for
single-user devices and group data storage.
BY RANDALL S. NEWTON
T
he various technologies that de-
fine modern computing do not
mature at equal rates. For exam-
ple, the first smartphones caught many
people by surprise; today’s intense
competition keeps smartphone ven-
dors pushing the boundaries to exceed
expectations. By comparison, watching
PC operating systems mature feels like
watching paint dry.
Data storage technology falls some-
where in between: For those who pay
close attention, improvements in data
storage hardware and standards are incre-
mental and obvious. But if the only time
one thinks about data storage is when you
equip a new office or update a worksta-
tion, it can be like visiting a 12-year-old The new Samsung SM951 leads a new generation of ultra
whom you last saw at age seven: The change is astonishing. low-power, high performance solid-state drives (SSDs).
When talking about storage technology, it is important Image courtesy of Samsung.
to distinguish between single-user data storage installed in
a PC or notebook, and group or enterprise data storage of thumb for pricing HDDs is $0.06 per GB; $0.10 per GB
installed in servers. Group storage technology continues to for SSDs. Many computers for professional use now combine
improve at a steady pace, but there has been a recent in- both technologies to achieve a cost-effective balance between
flection point for single-user storage technology — or two size and speed; high-performance consumer and professional
inflection points if it has been a while since you last looked. computers are quickly moving to SSDs only.
The newer inflection point in single-user data storage
Single-user Data Storage is a shift in how the storage is connected to the com-
The older change in single-user data storage is the rise of puter. For years the SATA design standard stood as the
the solid-state drive (SSD) as an affordable option to the common standard for storage interface, supported by
mechanical hard disk drives (HDD) we’ve been using for RAID for data integrity in workstations and servers. In
the last 30 years. SSDs are a permanently installed version recent years, the SATA design has been showing its age,
of USB thumb drives, which have largely replaced our need becoming perceived as a bottleneck to improved perfor-
for CDs and floppy disks for spontaneous and personal data mance. Instead of inventing a new standard, the com-
sharing. Because SSDs have no moving parts, they retrieve puter industry has decided to switch to another interface
and store data much faster than HDDs. For years the cost already found inside computers, the PCIe standards used
per megabyte difference between HDDs and SSDs was sub- for graphics cards and custom devices.
stantial, but the gap has closed considerably. The current rule SSDs using PCIe instead of SATA first came to market

works (SAN). There is no single answer to the question:
“Which is best for my workgroup?”
Below is a summary of each major storage system.
• Direct-Attached Storage (DAS): A low-cost, low-main-
tenance approach to data storage, in which hard drives are
connected directly to a host computer. Best for small busi-
nesses and sole proprietors with a small budget and limited
or no IT support. Not a good idea for sharing data among
more than a small workgroup.
• Network-Attached Storage (NAS): Dedicated data
The new HP Z Turbo Drive G2 can increase data I/O storage and sharing through a network. NAS can provide
transfer speeds by up to four times, when compared to features like RAID and swappable drives. A good approach
previous SSDs. Image courtesy of HP. for small businesses that have access to some IT support.
• Storage Area Network (SAN): A dedicated storage sys-
in 2014. Samsung rocked the market in January by intro- tem offering high performance for a large number of devices
ducing an SSD with read/write speeds roughly four times or a high volume of data transfer. Compared with DAS or
faster than the fastest SATA-based SSDs currently on the NAS, SANs are sophisticated and complicated; their instal-
market. At first the SM951 was only sold to computer ven- lation and maintenance is best left to dedicated IT personnel,
dors, but in May it became available for retail purchase. not a staff engineer who helps out with “computer stuff.”
The SM951 requires a new physical connection standard, There are several important criteria to consider when
M2, which means you can’t run out and buy the new Sam- evaluating group data storage systems:
sung SSDs for all of the old workstations in your office. • Budget. How much can you spend?
(Though, keep reading, there is a fix.) • IT Staff. Do you have dedicated IT staff to manage the
The move to running SSDs on PCIe not only increases storage system?
speed, but it also requires less space to install and less energy
to operate than using SATA with SSDs or HDDs. Those ben-
efits make it ideal for high-performance mobile workstations,
where the space and power savings can be put to good use. Personal CNC
There remains a trade-off when comparing such a cutting
edge storage unit like the SM951 to mature HDD tech: the Shown here is an articulated humanoid
Samsung SM951 is available in 128GB, 256GB and 512GB robot leg, built by researchers at the Drexel
Autonomous System Lab (DASL) with a
options. Many workstation storage units today are 1TB or
Tormach PCNC 1100 milling machine. To
larger, and remain significantly less expensive than SSDs. read more about this project and other
Not long after Samsung introduced the SM951, HP in- owner stories, or to learn about Tormach’s
troduced a desktop SSD using the new technology. The HP affordable CNC mills and accessories, visit
www.tormach.com/desktop.
Z Turbo Drive G2 puts the SM951 on a PCIe card that can
be added to any desktop workstation. The G2 uses four chan-
PCNC 1100 Series 3
nels of I/O to move data around, and completely bypasses the
SATA connection. The speed increase remains at approxi-
mately 4x when compared to using HDDs, a speed bump that
looks even better when dealing with large datasets.
Part of what gives the SM951 and products built from it
such increased zip is a new hard drive controller interface
— NVMe (non-volatile memory express) — designed
specifically for SSDs.
Group Data Storage Trends

The growth path in single-user storage is pretty straight-
Mills shown here with
optional stand, machine
arm, LCD monitors,
forward, moving from HDD to SSD technology for faster and other accessories.
I/O and lower power consumption. But, the technology

landscape for group data storage is not so simple. There are
three major options today: Direct-Attach Storage (DAS), PCNC 770 Series 3
Network-Attached Storage (NAS), and Storage Area Net- www.tormach.com/desktop

Engineering Computing /// Storage Options
• Backup and Recovery. Manual or automated? Where is
the backup stored? What happens if you lose files?
• Performance. How many users will be accessing the
system? Are users all on-site or is remote access necessary?
• Scalability. Can the system easily expand to support more
users? How much storage will you need in five years?
Ten years?
• Capacity. How much data do you generate each month?
What is the rate of data growth?
Beyond these considerations, there is the basic issue of
The Intel DC P3700 is a 16TB SSD using NVMe
which hard drive connection technology to use. The vast
technology designed for Xeon-based servers or
majority of PCs (business or consumer) use the SATA I/O
workstations. Image courtesy of Intel.
standard to connect both HDDs and SSDs. There is also
the more expensive and more reliable serial attached small
computer system interface (SAS) protocol, popular in en- speed will not increase just because they are on a SAS con-
terprise-class computing environments. If your organiza- nection. SATA is still more popular overall as a data storage
tion is large enough to have dedicated IT staff and you want throughput standard, but the new SAS specification could
state-of-the art data throughput for long distances (as in a mean additional market share in coming years, along with
campus environment), SAS becomes a serious contender. the possibility of SAS systems more suitable to small- to
The SAS 2.0 specification was recently published, offer- mid-sized business (SMB) environments.
ing the possibility of 12Gbit per second multipath through- Hitachi and Seagate are the most popular vendors of
put in the coming years as manufacturers adapt. The new HDD SAS drives, in sizes up to 4TB, while SanDisk, Intel,
spec also supports adding SATA drives, although SATA HGST and Micron are the most popular vendors of SSD
SAS drives. SAS technology still commands a premium
price, but will fall as SAS 2.0 becomes more common.
Bigger-Faster-Cheaper Still Reigns

Demand will never stop growing for faster data access,
higher read/write speeds, greater reliability, and lower
prices. As long as Moore’s Law means anything, storage
options will continue to improve in all aspects. As with all
facets of the fast-moving computer hardware industry, the
time to buy is not about when a new standard or model
comes out, but when your business can gain significant ben-
efit from an upgrade or expansion.
There is still a price difference between HDDs and
SSDs that will throw off some buyers, but prices continue
to fall, especially as SSDs become more common in enter-
prise environments. DE
Randall S. Newton is principal analyst at Consilia Vektor and a

contributing analyst for Jon Peddie Research. He has been part of
the computer graphics industry in a variety of roles since 1985.
INFO ➜ HP: HP.com

➜ Hitachi: HGST.com
➜ Intel: Intel.com
➜ Samsung: Samsung.com
➜ Micron: Micron.com
➜ SanDisk: SanDisk.com
➜ Seagate: Seagate.com
For more information on this topic, visit deskeng.com.

Workstation Review /// Engineering Computing
Extreme Power and Performance

The new Lenovo ThinkStation P900 delivers industry-leading results.
DAVID COHN
L
enovo unveiled its redesigned ThinkStation P Series
last August at SIGGRAPH 2014. We previously re-
viewed the P300 entry-level workstation (deskeng.com/
de/?p=21756) but had to wait awhile before we could get our
hands on the company’s new flagship workstation: the P900.
To say that it was worth the wait would be an enormous un-
derstatement.
The Lenovo P Series has been redesigned from the ground
up with the new ThinkStation P900 packing a tremendous
amount of power in a large tower case that measures 7.8 x 17.6
x 24.4 in. (W x H x D) and tips the scales at 52 lbs. Thankfully,
the new case includes three large handles — top front, top
rear and bottom front — making it very easy to move around.
Each handle is marked with a red touch point, a feature that
becomes even more prominent once you open the case.
The front panel has a new FLEX module that gives users The Lenovo ThinkStation P900 has been redesigned from
the option to only add the components they need, allowing the ground up. Image courtesy of Lenovo.
customized configurations that can incorporate ultra-slim
optical drives, 29-in-1 media card readers and iEEE 1394 For example, the ThinkStation P900 offers a 1300 watt
FireWire and eSATA ports. In our evaluation unit, the FLEX power supply with a built-in self-test. The power supply can
module contained the power button, a 9-in-1 media card be swapped out in seconds by simply pulling on its red touch
reader, a headphone jack and four USB 3.0 ports, including point. On the motherboard, just below the power supply, are
one always on and capable of charging USB devices. Below the first of the 8 PCIe expansion slots: a pair of x16 Gen3 slots
this are three vertical 5.25-in. drive bays, one of which con- as well as an x4 Gen 2 slot. Cards inserted in these slots are
tained a standard 5.25-in. DVD+/-RW dual-layer optical held in place by a rear panel bracket with its own red touch
drive. An optional Blu-ray drive is also available. The rest of point. In front of these slots is one of several cooling fans,
the front panel surrounding the FLEX module consists of a which can also be removed without any tools.
perforated metal screen with the ThinkStation logo. Along the front portion of the case are four drive bays,
The rear panel provides PS/2 keyboard and mouse ports, separated by the FLEX module, and identified by red touch
a 9-pin serial port, four more USB 3.0 ports, four USB 2.0 points. Each bay contains a cleverly designed tray with blind
ports, two RJ45 network jacks and audio jacks for line-in, line- connectors that can accommodate a 3.5- or 2.5-in. drive, or
out, and microphone. The NVIDIA GPU (graphics process- a pair of 2.5-in. drives. Lenovo offers lots of drive options,
ing unit) in our evaluation unit added its own pair of DVI including standard 7200 rpm drives ranging from the 500GB
ports and two DisplayPorts. drive in the base configuration to a 4TB drive, as well as hy-
brid and SSD (solid-state drive) options. You can also order
Totally Tool-less systems configured with multiple drives in RAID 0, 10, 1 and
Accessing the interior of the P900 could not be easier. A lever 5 arrays. Our evaluation unit came with a single 512GB SSD
releases the left side panel, which then slides off to reveal an installed in the lower-most bay, adding yet another $432 to
interior reminiscent of the HP Z Series workstations. Red touch the total system cost. PCIe SSDs are also available, for a total
points give a clear indication of where users can grasp compo- of 14 drive options.
nents that require no tools to remove. In fact, we found that A large direct cooling air baffle occupies the center of the
the only components inside the ThinkStation P900 that would case. This baffle conceals the two CPUs and 16 memory sockets
require a screwdriver to remove were the CPU heat sinks. Even and ensures that these crucial components receive ample fresh
the motherboard itself can be removed without tools by press- air. The baffle is actually divided into three separate channels so
ing its unlock button. This makes customization easy. that each CPU and the memory sockets receive their own direct

Engineering Computing /// Workstation Review
An air baffle conceals the CPU and memory sockets Lenovo’s modular design features red touch points,
and ensures that fresh air reaches crucial components. making it easy to remove and replace components.
Images courtesy of David Cohn.
flow of air. This baffle is also easily removed by pulling on its 2687W v3 processors, each with 10 CPU cores. This Haswell
red touch point, as is the dedicated 4.5-in. cooling fan. processor has a maximum turbo speed of 3.5GHz, 25MB of
Although the base configuration comes with a single Intel SmartCache and a 160-watt thermal design power (TDP) rat-
Xeon E5-2609 v3 processor, the P900 is a dual-socket sys- ing. The pair added $7,150 to the cost of our evaluation unit.
tem and Lenovo offers 42 different CPU configurations. Our The P900 base configuration comes with 8GB of RAM,
evaluation unit came with a pair of 3.1GHz Intel Xeon E5- installed as a pair of 4GB modules. But this ThinkStation
can accommodate up to 512GB of memory using 32GB
DIMMs (dual in-line memory modules). Our system came
INFO ➜ Lenovo: Lenovo.com
with 128GB, installed using 16 8GB DDR4 ECC 2133MHz
Lenovo ThinkStation P900 RDIMM modules, adding $5,625 to the total cost.
• Price: $16,599 as tested ($1,826 base price) Below the CPUs is a second pair of PCIe x16 Gen 3 slots
• Size: 7.8 x 17.6 x 24.4 in. (W x H x D) tower along with another PCIe x4 Gen 2 slot and a pair of PCIe x1
• Weight: 52 lbs. slots. The basic P900 comes with an NVIDIA Quadro K620
• CPU: Two 3.1GHz Intel Xeon 10-core E5-2687W v3 GPU. Again, Lenovo offers lots of choices, ranging from an
• Memory: 128GB DDR4 ECC at 2133MHz NVIDIA NVS315 board up to a pair of NVIDIA Quadro
• Graphics: NVIDIA Quadro K6000 and NVIDIA Tesla K40 K6000s. Our evaluation unit came with an NVIDIA Quadro
GPU accelerator K5200 graphics board installed in one of the x16 slots. The
• Hard Disk: 500GB SanDisk SSD K5200 has 8GB of dedicated memory and features 2304
• Floppy: None CUDA parallel processing cores. This double-width board
• Optical: 16X DVD+/-RW added another $2,034 to the system cost and blocked access to
• Audio: Integrated HD audio (front panel: headphone; one of the adjacent x4 slots. Again, boards are held in place by
rear-panel: line-in, line-out, microphone) another rear panel bracket with its own red touch point and
• Network: Integrated gigabit Ethernet, two RJ45 ports cooled by yet another easily removable cooling fan.
• Modem: None Other options include NVIDIA Tesla and Intel Xeon Phi
• Other: Eight USB 3.0 (four front/four rear), four USB 2.0 ports coprocessors as well as a Thunderbolt I/O card, optional Eth-
rear, one 9-pin serial, PS/2 mouse and keyboard ports, two ernet server cards, an IEEE 1394 FireWire adapter, and Flex
DVI ports and two DisplayPorts on NVIDIA board, 9-in-1 media Bay storage options.
card reader
• Keyboard: 104-key Lenovo USB keyboard Amazing Performance
• Pointing device: Lenovo USB optical wheel mouse We have often commented on Lenovo’s ability to combine
• Power Supply: 1300 watts, 92% and configure quality components for optimum performance,
• Warranty: Three years parts and labor but the results we obtained from our ThinkStation P900 were

Workstation Review /// Engineering Computing
nothing short of amazing. On the SPECviewperf benchmark few are likely to buy a system with this much potential equipped
(see page 42), the P900 scored at or near the top on every with entry-level components. Even if you were not keeping a
dataset, even when compared to systems equipped with more running tally as we listed the cost of each upgrade in our evalu-
powerful GPUs and faster processors. ation unit, you can already anticipate the bottom line. As config-
On the SPECapc SolidWorks benchmark, the P900 sur- ured, our Lenovo ThinkStation P900 priced out at a whopping
passed every other system we have ever tested except for the $16,599, putting it among the top three most expensive systems
single-socket, over-clocked BOXX APEXX 2 (deskeng.com/ we have ever tested, but still nearly $2,300 less than the Mi-
de/?p=23627). And on the multi-threaded AutoCAD render- croway WhisperStation (deskeng.com/de/?p=18866). And we
ing test, no other system could compete against the equivalent could have easily run the cost up even higher by adding more
of 40 CPU cores. The ThinkStation P900 completed our test storage capacity, memory, and GPUs because our P900 still had
rendering in just over 16 seconds, nearly twice as fast its near- plenty of room — and power — to spare.
est competition. Clearly, the Lenovo ThinkStation P900 is not for every-
We also ran the new SPECwpc workstation performance one. It is aimed at high-end CAD, CAE, research, rendering,
benchmark, and here too the Lenovo ThinkStation P900 turned special effects and the oil and gas industries. But if you need
in some of the highest numbers we have ever recorded. Yet the an extremely powerful system with lots of room to grow, your
P900 was nearly silent, even when under heavy compute loads. search has ended. The Lenovo ThinkStation P900 is a perfor-
Lenovo pre-loaded Windows 7 Professional 64-bit. Win- mance leader. DE
dows 8.1 Pro 64-bit is also available. Like other Lenovo work-
stations, the new ThinkStation P900 comes with a three-year David Cohn is the technical publishing manager at 4D Technolo-
on-site warranty. A 104-key USB keyboard and USB optical gies. He also does consulting and technical writing from his home
wheel mouse come standard. The system is ISV (independent in Bellingham, WA and has been benchmarking PCs since 1984.
software vendor) certified for applications from Adobe, Au- He’s a Contributing Editor to Desktop Engineering and the author
todesk, Dassault, PTC, SolidWorks, Avid and Siemens. of more than a dozen books. You can contact him via e-mail at
Although Lenovo touts a P900 starting price of just $1,826, [email protected] or visit his website at www.dscohn.com.
HigH speed clusters

• Your turnkey HPC cluster is • Support for all CAE/CFD
managed by your dedicated solvers, license management,
TotalCAE IT team in your data engineering applications, and
center, a true private cloud. cluster software included.
• Get your results faster

and increase engineering
productivity.
private
• Simple to use. cae cloud
Register today for our free research report on the benefits of TotalCAE Private Cloud
Http://www.totalcae.com/privatecloud
halfpagead.indd 1 deskeng.com /// August 2015 DESKTOP ENGINEERING 39PM
10/13/2014 5:54:23
Big,
Bright,
Fast &
Loud
Origin PC delivers the massive
ultra high-end Millennium Pro X2
workstation.
BY DAVID COHN
W
e agreed to review a workstation from Origin The Origin PC Millennium Pro X2 is housed in a massive
PC, a Miami-based company (founded by former case with a window in its side panel and interior lighting
employees of Alienware) that builds high-per- controlled by a small remote. Ten front panel drive bays
formance computers for gamers and hardware enthusiasts. are accessible behind a hinged door.
According to the company’s website, its systems are hand- Image courtesy of David Cohn.
built using only the best quality/performing components
and tested for 72 hours before being shipped to customers. sculpted case measured 9.75 x 24.8 x 21.44 in. (W x D x H)
Origin touts the fact that it assembles its systems in the U.S. and weighed 55 lbs. But the steel frame was clad in many
and will build a computer using any chassis available. The places with flimsy plastic and the recesses that appeared to be
company sells small form factor, mid-tower and full-tower handles were anything but — adhesive labels warned: “Not
systems as well as mobile workstations. A Lift Point.” The removable lockable side panel has a large
Imagine our surprise when the system arrived in a wooden window to show off internal components while a hinged
crate weighing 92 lbs. After removing 14 screws to open the front door conceals five 5.25-in. front drive bays plus five
crate, we found the computer itself packed in a thick card- hot-swappable, lockable hard drive bays (the locks add $58
board box. Inside that box, the system was supported by foam to the base price).
inserts and inside the computer was a foam block to protect Our review unit came with a white removable panel on the
components and keep them from moving during shipping. left and the door hinged on the right, but the chassis can be
This is standard for all large Origin systems. oriented with the panel on the right and the door hinged on
Once we got it unpacked, we paused for a moment to ad- the left. Behind that door, two of the drive bays in our system
mire its size. The Origin Millennium Pro X2 is billed as the contained a 40-in-1 media card and a 16x Blu-ray burner, op-
company’s mid-size professional workstation, but it is larger tions that added $21 and $86, respectively. You also have a
than many competitors’ full-tower systems. The beautifully choice of one of three standard colors a custom paint scheme.

A recessed area on the top of the case conceals the power workstation-class GPUs (graphics processing units). Ori-
button, four USB 3.0 ports, headphone and microphone gin offers six different NVIDIA gaming cards and a choice
jacks, a reset button, a manual fan control knob and a button of seven NVIDIA Quadro or three AMD FirePro cards.
to switch the fans to maximum cooling mode. Behind this Our system came with an NVIDIA Quadro K6000 GPU,
recess, a perforated plastic panel covers the liquid cooling featuring 12GB of discrete memory and 2880 CUDA
system’s three large fans. When powered up, the Millen- (compute unified device architecture) parallel process-
nium Pro X2 filled the room with a constant 60dB hum that ing cores. Origin has since replaced this with the Quadro
climbed as high as 75dB under compute loads — the equiva- M6000, a $5,244 option with 12GB of GDDR5 memory
lent of a vacuum cleaner’s noise level. What’s more, the fan and 3072 CUDA cores.
controls did little to alter this constant din. Origin also included an NVIDIA Tesla K40 GPU accel-
The rear panel provides six more USB 3.0 ports, two USB erator, providing an additional 12GB of GDDR5 memory
2.0 ports, a PS/2 keyboard/mouse socket, two RJ-45 network and another 2880 CUDA cores. This is the same GPU accel-
connections, an optical S/PDIF output port and six audio erator provided last year by Microway in the WhisperStation
jacks (separate microphone and line-in jacks as well as line- we reviewed (deskeng.com/de/?p=18866). It added another
out/front, side, rear and bass output speaker channels). $3,863 to what was already becoming an expensive system
The interior is lit by an array of LEDs and a small remote and really didn’t contribute much to our benchmark results.
lets you select from 16 light colors and four lighting effects, So why buy a Tesla? There are hundreds of applications
including flash and strobe. You can also dim these lights or across a wide range of fields that are optimized for GPU
turn them off entirely. Ultimately, we found the lighting ef- computing. In the world of CAD/CAM/CAE, some aspects
fects to be quite distracting. of many popular CAD programs — such as the surface and
mesh modeling and rendering functions of AutoCAD, In-
Configuration Options ventor, CATIA and NX — take advantage of GPU accelera-
Origin gives customers a huge range of options to choose tion. When you move into the realm of computational fluid
from and guides them through the configuration process dynamics (CFD), structural analysis and simulation, most
with a multi-step online wizard. The first step is to select directly support the Tesla GPU.
a single or dual CPU. The Millennium Pro X2 we re- Both NVIDIA cards require auxiliary power connections,
ceived is a dual-CPU system based on an ASUS Z10PE- and when you realize that together they consume 485 watts,
D8 WS motherboard with a base starting price of $3,780. you can understand the need for the large power supply.
A pair of 2.4GHz Intel Xeon E5-2630 eight-core CPUs, Even with both boards installed, we still had access to three
32GB of memory and a 500GB standard hard drive are PCIe slots and more than enough capacity to power addi-
included in that price, but no graphics card. From there, tional add-ons.
the configuration fun begins. You can choose your case The ASUS motherboard provides eight DIMM (dual
orientation and color and then start adding and upgrad- in-line memory module) sockets and supports up to
ing components. 512GB of memory. Origin equipped our review unit with
Our system included a Frostbyte 360 X2 sealed liquid 64GB of RAM installed as four 16GB ECC registered
cooling system with its hoses routed to the radiator mounted 2133MHz modules, adding another $694 to the price. The
below the perforated grill on the top of the case (a $68 op- company currently offers the X2 motherboard with up to
tion). Our evaluation unit was also equipped with a pair of 128GB of RAM.
Intel Xeon E5-2687W v3 10-core Haswell CPUs. These Our system also came with an upgraded primary hard
processors, which added $3,162 to the cost of the system, drive, a 1TB Samsung 850 Evo series SSD (solid-state drive)
have a base frequency of 3.1GHz, a maximum turbo speed — a $434 option — installed in an additional drive bay nes-
of 3.5GHz, 25MB cache, and a thermal design power (TDP) tled below the hot swap bays. That meant that there was still
rating of 160 watts. Origin offers a choice of 11 different room in the massive case for three more drives in the stan-
CPUs including a pair of 18-core E5-2699 v3 CPUs. dard bays plus five in the hot swappable bays. RAID arrays
To ensure adequate heat transfer from the CPU to the are also available.
cooling system, our X2 also included the optional Gelid GC- Other options include an ASUS ThunderboltEX II Ex-
Extreme thermal compound ($15). And because our system pansion card ($74), optional audio boards, upgraded net-
would be completely tricked out, Origin included its top- working cards, and so on. After completing the configura-
of-the-line 1500 watt 94% efficient Corsair power supply, tion, the next page of the wizard offers mouse pads, T-shirts,
which added another $327. You also have your choice of posters and hats. On this page, you will also find a choice of
power supply cable colors. keyboards. Since a keyboard and mouse are not included in
The ASUS motherboard provides seven PCIe Gen3 the base price (and were not included with our evaluation
x16 slots and can support up to four gaming cards or two unit), we added the least expensive options — a $20 Micro-

Dual-Socket Workstations Compared
Origin Millenium Lenovo P900 Microway BOXX HP Z820
Pro X2 WhisperStation 8980XTREME
Two 3.1GHz Intel Two 3.1GHz Intel Two 3.5GHz Intel Two 3.1GHz Intel Two 3.1GHz Intel Xeon
E5-2867Wv3 ten-core E5-2867Wv3 ten-core E5-2643v2 six-core E5-2687W eight-core E5-2687W eight-core
CPUs, NVIDIA Quadro CPUs, NVIDIA Quadro CPUs, NVIDIA Quadro CPUs overclocked CPU, NVIDIA Quadro
K6000, 64GB RAM, K5200, 128GB RAM K6000, 64GB RAM, to 3.82GHz, NVIDIA 5000, 32GB RAM
NVIDIA Tesla K40 NVIDIA Tesla K40 Quadro K5000,
64GB RAM
Price as tested $18,186 $16,599 $18,889 $13,454 $9,984
Date tested 5/25/15 2/20/15 5/10/14 5/9/13 7/16/12
Operating System Windows 7 Windows 7 Windows 7 Windows 7 Windows 7
SPECviewperf 12 higher
catia-04 107.35 94.69 94.60 n/a n/a
creo-01 88.49 82.19 79.40 n/a n/a
energy-01 6.85 3.91 6.22 n/a n/a
maya-04 84.17 70.20 68.75 n/a n/a
medical-01 44.28 31.54 33.44 n/a n/a
showcase-01 69.44 49.20 62.54 n/a n/a
snx-02 104.89 151.60 90.09 n/a n/a
sw-03 122.48 113.92 99.53 n/a n/a
SPECviewperf 11 higher
catia-03 110.07 108.16 104.27 78.01 51.69
ensight-04 181.73 140.25 167.13 80.25 44.13
lightwave-01 82.47 75.51 81.01 77.07 59.02
maya-03 204.02 122.35 150.18 125.16 101.67
proe-05 20.16 19.72 15.45 16.14 11.72
sw-02 50.13 67.89 70.70 67.16 57.48
tcvis-02 84.57 80.68 87.47 71.58 44.52
snx-01 159.47 134.18 147.95 81.35 44.86

SPECapc SolidWorks
higher
2013
Graphics Composite 8.63 9.55 6.89 2.69 2.15
RealView Graphics 10.11 11.23 8.05 2.86 2.37
Composite
Shadows Composite 10.07 11.22 7.96 2.86 2.36
Ambient Occlusion 22.14 24.38 17.83 6.16 5.19
Composite
Shaded Mode 8.68 9.56 7.04 2.62 2.27
Composite
Shaded with Edges 8.57 9.55 6.74 2.77 2.03
Mode Composite
RealView Disabled 4.56 5.01 3.69 2.11 1.45
Composite
CPU Composite 4.75 4.77 4.20 4.84 4.50
Autodesk Render Test lower
Time seconds 36.50 16.06 30.33 38.00 41.00
Numbers in blue indicate best recorded results. Numbers in red indicate worst recorded results.

ing your workstation without an operating system. Ours also
came with a three-year warranty, a $269 option beyond the
one-year that comes standard.
Our Millennium Pro X2 was definitely equipped well be-
yond the needs of most Desktop Engineering readers. Elimi-
nating the Tesla card and opting for a less powerful GPU
would immediately save nearly $10,000. But for those run-
ning high-end engineering analysis applications, a system
this powerful may be warranted.
That said, we had to ask ourselves: Would this computer
be appropriate in an office environment? Our conclusion
was: probably not. No matter what you put inside, the Mil-
lennium Pro X2 looks like a gaming system. Its lighted in-
terior is distracting and the din emanating from the fancy
case was more than just loud, it was annoying. A hardware
enthusiast in a small shop might find the Origin PC Mil-
lennium Pro X2 quite attractive, but we doubt it will find a
home in many corporate environments. DE
Origin PC ships its Millennium Pro X2 in a large wooden David Cohn is the technical publishing manager at 4D
crate. Image courtesy of Origin PC. Technologies. He also does consulting and technical writing
from his home in Bellingham, WA, and has been benchmark-
soft USB keyboard and $15 Microsoft USB optical mouse ing PCs since 1984. He’s a contributing editor to Desktop
to our system — but choices here go as high as $215 for Engineering and the author of more than a dozen books. You
a keyboard and $145 for a mouse. Other selections include can contact him via e-mail at [email protected] or visit his
monitors, speakers, webcams and more. website at www.dscohn.com.
Price, Performance and Verdict

INFO ➜ Origin: OriginPC.com
By the time we added everything up, the total came to
$18,186, which was $703 less than the MicroWay Whisper- Origin PC Millennium Pro X2
Station. With two very fast 10-core CPUs, plenty of RAM, • Price: $18,186 as tested ($3,790 base price)
an ultra-high-end GPU and the Tesla GPU accelerator, we • Size: 9.75 x 24.8 x 21.44 in. (W x H x D) tower
knew it was going to be fast. But would its performance • Weight: 55.5 lbs.
match its price? • CPU: Two 3.1GHz Intel Xeon 10-core E5-2687W v3
On the SPEC Viewperf tests, the Millennium Pro X2 • Memory: 64GB DDR3 ECC at 2133MHz (4X16GB), 128GB max
blew away the competition on almost every data set, thanks • Graphics: NVIDIA Quadro K6000 and NVIDIA Tesla K40 GPU
in large part to its K6000 GPU. With the newer NVIDIA accelerator
M6000, we would expect those results to be even better. But • Hard Disk: 1TB Samsung 850 Evo SSD
on the SPECapc SolidWorks benchmark, the Origin work- • Floppy: None
station lagged a bit behind the Lenovo ThinkStation P900, • Optical: 16x Blu-ray burner
in spite of both utilizing the same CPU. And on the Auto- • Audio: Integrated Realtek ALC1150 8-channel HD audio (front
CAD rendering test, a multi-threaded test on which faster panel: headphone and microphone; rear-panel: microphone,
systems with more CPU cores have a distinct advantage, the line-in, line-out/front, side, rear, and bass)
Origin PC Millennium Pro X2 came up quite short — its • Network: Two integrated Intel I210-AT gigabit LAN controllers
36.50 second average rendering time, while still fast, was with two RJ45 ports
more than double that of the Lenovo P900 — despite iden- • Modem: None
tical CPUs and the equivalent of 40 cores. • Other: Eight USB 3.0 (2 front/6 rear), two USB 2.0 ports rear,
We also ran the SPECwpc workstation benchmark. Here, PS/2 mouse and keyboard ports, two DVI ports and two
the Millennium Pro X2 captured most of the honors, though DisplayPorts on NVIDIA board, 40-in-1 media card reader
it lagged behind both the Lenovo P900 and the MicroWay • Keyboard: Microsoft USB keyboard
WhisperStation on some of the individual tests. • Pointing device: Microsoft USB optical mouse
Origin pre-loaded Windows 7 Professional 64-bit. Win- • Power Supply: 1500 watts, 94%
dows 8.1 Pro is also available, or you can save $129 by order- • Warranty: Three years parts and labor

Internet of Things /// Smart Factories
Manufacturing in the
World of Industrie 4.0
Engineers may one day be designing for a future factory epitomized
by this German initiative supported by the Siemens Digital Enterprise suite.
BY PAMELA J. WATERMAN
I
f Germany has anything to say solutions seem to be keeping pace. mation through Internet of Things
about it, the Factory of the Future Driving this radical change are (IoT) technology, both with other
is really the Factory of the Very worldwide demands for improved equipment and with the parts them-
Near Future. Formal, repetitive assem- quality, lower labor costs, individual- selves. High-speed networks and
bly lines will be obsolete by 2025, re- ized products, and shorter product cloud-based computing resources al-
placed by flexible, smart manufacturing lifecycles. Enabling the changes are ready analyze data to direct the next
systems that act and react according to complementary technologies that af- move. And in the coming years, fixed
a two-way digital data-stream. That’s a fect the entire value chain. Manu- assembly lines will be replaced by flex-
pretty bold vision to accomplish in just facturing equipment can now share ible, modular systems including 3D
10 years, but both the impetus and the highly detailed, two-way status infor- printers that reconfigure themselves
Software products from Siemens let engineers simulate and test the digital twin of a current or proposed production
line to evaluate manufacturing flexibility. Images courtesy of Siemens PLM Software.

At the German Research
Center for Artificial
Intelligence, the RFID chip
(seen close up on the back of
the bottle being filled below)
and Siemens hardware
are used to communicate
instructions to the production
equipment for proper
processing and quality
verification.
as needed. This connected approach

would help companies save production
time while adding more value (espe-
cially customization) to their offerings.
Companies, and countries, that
embrace this scenario will be leading
a global manufacturing revival. This
month, Desktop Engineering takes a look
across the pond at German efforts to
network humans, objects and systems
for tomorrow’s smart factory revolu-
tion, with Siemens PLM Software in a
lead role.
Shaping the Next Wave of Industry

Historians have labeled significant eras
of manufacturing after disruptive tech- subtle nod to the world of software re- neering executives at Siemens PLM
nologies. Steam-powered machinery, visions, reflecting the fusion of manu- Software, a company whose products
electrified equipment and computer- facturing equipment, control systems, are directly relevant to the inter-
controlled systems each changed pro- and data collection made possible in connected world. “Next generation
duction methods forever. Though still large part by the technology behind manufacturing offers a way to meet
in its infancy, a new, fourth industrial the Internet of Things (IoT). Small, customer demands for new, high-
revolution promises to do the same, inexpensive sensor hardware combined quality customized offerings at ever-
and it already has a general name: In- with targeted application software has shorter time intervals. It also has the
dustry 4.0. enabled explosive growth in smart potential to reduce resource utiliza-
In the U.S., support for the tran- devices. Adding this capability to in- tion, which will help manufacturers
sition to Industry 4.0 comes from dustrial equipment, processes and in- cope with growing cost pressures,”
groups such as the Smart Manufac- ventory has created an industrial IoT Zvi Feuer, senior vice president of
turing Leadership Coalition and its (IIoT) sub-domain of cyber-physical Siemens PLM Software, wrote in a
program to create the Open Smart systems that is already helping com- recent corporate blog post.
Manufacturing Platform. In Germany, panies monitor, communicate, analyze Siemens PLM Software is an in-
a country with a long history of man- and apply digital manufacturing infor- dustry partner in the Industrie 4.0
ufacturing expertise, a coordinated mation in close to real time. initiative, which began in 2005 and
government initiated project for intel- formalized in 2011 as a forward
ligent production operates under the New Business Models, New Tactics looking project under the German
variant name, Industrie 4.0. Why is this important? One perspec- Federal Government High-Tech
Both nomenclatures give a not so tive comes from manufacturing engi- Strategy. Dual goals are to maintain

Internet of Things /// Smart Factories
market leadership by integrating know that a machine needs to be
smart technologies into traditional tuned, offers various ways to improve
RFID Tag, You’re production industries such as elec-
tronics and chemicals, and to create
process quality and supports tooling
certification (an increasingly impor-
It: Defining and serve new markets enabled by
cyber-physical systems.
tant aspect of cradle-to-grave track-
ing and certification requirements).
Unique Identities Feuer says that German manu-
facturing is already seeing use of
Setting up such production lines,
particularly as consumers demand
T
equipment that can react to parts more customized, one-of-a-kind
he original Universal Product tagged for RFID (radio frequency products, requires rapid, integrated
Code (UPC) bar code system identification) and assemblies in ad- planning. Feuer says that the Sie-
(begun in 1963) is slowly vanced industries such as automo- mens PLM Software toolset is a
being replaced with the Electronic tive. “When a partially assembled ve- crucial part of the Siemens Digital
Product Code (EPC) consisting of hicle is moving on the assembly line, Enterprise Software Suite. Combin-
a bar code plus numbers, whose it carries a sensor and in there is a ing simulation, automation and data
definition has already gone through code of what needs to be done next,” management, the suite offers a com-
several iterations. he says. “When it comes into a sta- plete solution aligned with all the
The basic technology used to tion, the station can react automati- requirements of Industrie 4.0.
support the EPC as a global, end- cally and show the assembly workers “All of our tools work on top of
to-end supply chain standard is who are going to operate this station the PLM Teamcenter backbone,”
the radio frequency identification what needs to be done for this spe- he says. “You start by designing the
(RFID) tag and reader, based on cific vehicle. product then store the design infor-
the newest EPC Gen 2 definitions. “This is not rocket science, but mation in Teamcenter. You then do
Globally, such systems operate still it requires a lot of pre-planning, the analysis, do the manufacturing
over the 860MHz to 960MHz making sure the robotics, tools and process, and then go and design the
band. North American Gen 2 controllers all work in synch,” Feuer production facility that can put to-
uses 902MHz to 928MHz while continues. “You will also see more of gether this product. Today all of this
European Gen 2 uses 860MHz to this in the process industry, for ex- is done in parallel that used to be se-
868MHz; equipment based on EPC ample, with bottling. The RFID tag quential.”
Global tags work across the full on a bottle will tell the machine what
EPC spectrum. kind of formula to fill, like in sham- Cooperation
In order for products to have poo production or perfume. We’re and Standards
a unique identity from birth and going to see this in combination with Within the Industrie 4.0 effort,
be traceable cradle-to-grave com- the more advanced robotic facilities.” Feuer says there is an unbelievable
ponents, assemblies and final Equipping both the parts and the open market attitude of both coop-
products from smart factories will manufacturing equipment with sen- eration and competition. “Our Sinu-
undoubtedly be labeled with RFID sors supports a number of other steps merik CNC (computer numerically
tags throughout the manufacturing critical for the smart factory evolu- controlled) machine controls are
process. To accommodate differ- tion. First, capturing detailed infor- extensively used to operate robots,”
ent types of products, materials mation lets robotic vision systems in- says Feuer. “Each vendor has its own
and pricing needs, subsets of the spect, measure and compare as-built control system but some customers
EPC bands are assigned to differ- parts to the original CAD-defined prefer to use the Siemens control
ent power levels and capabilities. dimensions for automated go/no-go because of its agility, versatility and
Various types of RFID tag/reader decisions. Second, information can ability to connect with various simu-
systems operate with active, semi- be sent back to the design engineers, lations of line programming and vir-
passive and passive tag technology who can learn from mistakes made tual commissioning (system setup)
as well as different read/write data in part design or assembly processes. software.”
structures and content. And third, real-time or near real- With all the equipment and pro-
For a good discussion of the time data gives feedback on how the cesses in use within any given indus-
possibilities, see skyrfid.com/ equipment operation may be devi- try segment (automotive, electrical,
RFID_Gen_2_What_is_it.php. ating from the perfect virtual plan. chemical, etc.) no single vendor can
-- PJW Such digitally recorded information supply every type of system. It’s no
lets motion-control programmers wonder that equipment based on a

cycle, so problems during manufac-
turing can be quickly identified and
corrected. Step two compiles the data
gathered over the course of building
thousands of parts, or running machin-
ery for thousands of hours, providing
insight into deviations or faults; such
information can then guide improve-
ments to both processes and part de-
sign. A third step will require translat-
ing this data at a higher level into new
product concepts offering capabilities
that weren’t previously possible or
identified.
Siemens’ long-term strategy may
take 15 to 20 years, but Feuer be-
lieves many aspects of Industrie 4.0
factories will come sooner. The ben-
efits will go beyond improving man-
Internet-connected autonomous systems have jump-started a fourth industrial
ufacturing efficiency and satisfying
revolution that includes big data mining, customer-driven designs and
discerning consumers. “Industrie 4.0
modular assembly lines.
is not a slogan to generate more rev-
enues,” he says. “It’s a strategy started
variety of engineering and industrial And Siemens PLM Software’s by the government and adopted by
standards gets used concurrently own JT file format is seeing grow- Siemens and some other big players
and must work together. To support ing use for lightweight visualization in the industry because we want to
widespread implementation of the of 3D product data. Since its begin- make sure we are doing something
Industrie 4.0 vision, these standards nings in 1997 and formal publica- good for the society as well as mak-
must be coordinated, modified or tion in 2007, the JT format has sup- ing a profit. We want to create jobs,
consolidated. ported digital collaboration at many create opportunities. Even though it
Areas with multiple possible stan- levels, making it easier and faster to sounds a bit strange because we’re
dards include Ethernet implementa- move 3D data within a company or talking a lot about automation and
tion, data exchange related to auto- enterprise, or between a company robotics — yes, the types of jobs on
mated manufacturing equipment and and vendors. Accepted as the world’s the shop floor are going to change —
data exchange of part files created in first ISO International Standard for but we still will be creating jobs and
different CAD programs. Profinet is lightweight 3D visualization, JT al- they are well paying jobs.” DE
an industrial Ethernet implementa- lows multiple parties to exchange 3D
tion that adds real-time performance information even if they do not use Contributing Editor Pamela Wa-
and robustness to everyday Ethernet the same CAD programs. terman , DE’s simulation expert, is
use and is becoming a worldwide in- an electrical engineer and freelance
dustry standard. Practical Steps, technical writer based in Arizona. You
Automation Mark-up Language Long-Term Strategy can send her e-mail to DE-Editors@
(AutomationML) is a neutral data Implementing Industrie 4.0 tactics will deskeng.com.
format based on XML that enables not be a linear process, but companies
information exchange between can — and are — taking steps right INFO ➜ Industrie 4.0: LinkedIn.com/
cross-disciplinary automated equip- now to bring their manufacturing pro- company/industrie-4-0
ment. It actually incorporates sev- cesses into this new era. Adding sensors
eral different standards that deal and software to equipment and compo- ➜ Open Smart Manufacturing Platform:
SmartManufacturingCoalition.org
with topology, geometry, kinematics nents forms the first step for factories
and logic control. Eventually, this to become data-driven and responsive. ➜ Siemens PLM Software: Siemens.
approach will let users connect dif- Active monitoring of this information com/plm
ferent devices and display all process in real time allows tracking individual For more information on this topic,
information within one system. parts through their entire production visit deskeng.com.

Engineering Services /// Compliance
Image courtesy
Design for
iStockphoto
Compliance
Service providers help the medical device industry

accelerate time to market and reduce risk.
BY LAUREN GIBBONS PAUL
H
ere’s a cautionary tale for design engineers working in the Premarket Notification vs. Premarket Approval
medical device industry. Excited at the prospect of bring- The FDA’s Center for Devices and Radiological Health
ing its innovative new endoscope to market, a company (CDRH) is responsible for regulating firms that manufac-
rushed to get its prototype in front of the U.S. Food and Drug ture, repackage, re-label or import medical devices sold in the
Administration (FDA) with an eye toward fast-track approval. United States. In addition, states may have individual laws ap-
Unfortunately, the FDA immediately pointed out it would plying to the design and manufacture of medical devices.
be difficult to clean the surgical scope adequately in the Medical devices are classified into Class I, II and III, with
time allotted between patients. As a response, the company the last category reserved for devices that pose the greatest risk
pointed out the device was “just a prototype” and not in- to consumers. Most Class I devices are exempt from filing a
tended to represent the final form factor. But their attempts 501(k) Premarket Notification, whereas most Class II devices
to steer the conversation to the target patient population and require filing a Premarket Notification 510(k) and most Class
expected benefits fell on deaf ears. It was difficult to secure III devices require Premarket Approval (PMA). The PMA
funding after this early failure and the company went out of process is viewed as the most onerous of the three and the
business not long after. subject of some trepidation among device makers. For devices
This true story points out the perils involved in paying that fall into Class II, a 510(k) is done to demonstrate that the
insufficient attention to regulatory compliance in the U.S. device to be marketed is safe and effective by proving substan-
medical device industry. Some medical device manufacturers tial equivalence (SE) to a legally marketed device (predicate
have been guilty of viewing regulations as an afterthought, es- device) that is not subject to PMA.
pecially in past years. But there are stringent requirements for Depending on the type of device, it typically takes one to
medical device design and manufacturing, especially for any three years for a device classed as “moderate risk” to attain
device that presents risk of physical harm to the end user. The FDA approval or eight to 10 years for medical devices that
good news: A host of providers offers a wide range of services are at a high level of risk, says Cynthia Nolte, senior director
to help with every possible aspect of compliance — before, of Regulatory Affairs for ICON Clinical Research, a unit of
during and after. ICON plc and provider of compliance-related services.

“Rushing to market is not the right strategy,” says Nolte.
Preparing Your Case
I
The regulations were created to ensure safe and effective
products are delivered to the marketplace with the right risk- n preparation for review by the FDA, medical device
benefit equations. “If we are talking about a Fitbit that re- manufacturers will need to provide the following
cords things like your heart rate, there will be no regulatory information as a starting point:
control because it does not represent any risk,” says Nolte. • Proposed claims
“If we talk about a corneal implant, that is a significant level • Preliminary instructions for use
of risk and much more regulatory control. Global regulatory • Target patient population
agencies take this risk-based approach to regulation.” • Device classification — Class I, II, or III; exempt; 510(k);
Ideally, designing for compliance is just as much a part of or probable PMA
the process as designing for quality, says Pavan Kumar Gari- • Foreign marketing strategy and additional regulatory
kapati, regulatory domain head, Medical Devices, Wipro Ltd., agency requirements
also a service provider. “It is often misunderstood that compli- • Packaging needs and labeling requirements
ance just requires completing required documentation after a • Test plan to show conformance to existing standards
product is ready to sell,” he says.
There is a continual balancing act between providing gov- the appropriate credentials,” Nolte says.
ernmental authorities with the minimum required data needed When sifting through a long list of compliance services pro-
to be — or remain — compliant, while at the same time run- viders, look for one that has proven experience on compliance
ning efficient operations. for a similar type and class of device. As with any kind of out-
sourcing agreement, cost is always a driver in selection but it
One-Stop Shopping can be shortsighted to make the decision solely based on cost.
There are companies that go it alone when it comes to com- Garikapati suggests that device companies looking for
pliance, says Nolte. Generally, in this case, there are engineer- compliance partners use this shorthand checklist:
ing and product development teams with an understanding of • A strong QM system.
the regulatory landscape and its requirements. • A clearly articulated framework for handling intellectual
Most medical device companies do engage a contract re- property (IP) and sensitive information, such that confiden-
search organization (CRO) such as ICON to help them clear tiality and IP protection are ensured.
compliance hurdles. ICON does everything from helping set • Extensive experience working with medical device firms.
strategy to guiding the client’s design work so that the end • Meticulously drafted agreements, sharing all the information
product complies with the regulatory requirements of the tar- that is needed by the manufacturer to validate the design
get market. ICON consultants also interface with regulatory (including source code, QM documents, result of risk analysis
agencies on behalf of their clients, paving the way for smooth and design evaluations).
interactions down the road. For companies just beginning on the design of a new medi-
Along with other service providers, ICON provides guid- cal device: Think twice before you go it alone. The vast ma-
ance as to which quality management (QM) systems clients jority of companies need compliance services and support in
should use to meet the regulatory and quality expectations of order to be successful.
that market. “Most companies wouldn’t even know where to begin wad-
“We help clients institute a QM system for managing de- ing through all of the regulations,” says Nolte.
sign and manufacturing practices,” says Nolte. Further down- Most of all, you want to avoid ending up like the scope
stream, ICON assists engineers with crafting labels, inspec- maker at the beginning of this article. “If you mess up your
tions for use and operator manuals. It also helps design and one chance to speak to the agency because you didn’t have
execute clinical studies when needed. Consultants can pin- your story straight, you will have a problem — especially if
point the likely regulatory impact of a needed design change. your product is a PMA,” says Nolte. DE
“We get involved in all phases of the product lifecycle. It’s
one-stop shopping,” she says “We like to do it in a partnership Lauren Gibbons Paul is a Boston-based freelance writer. Contact
where we participate as a member of their team.” her via [email protected].
Engaging a Partner
ICON Clinical has been in business more than 30 years INFO «Food and Drug Administration: FDA.gov
and is ISO-9001 certified for quality. “That tells our clients
«ICON: ICONplc.com
that we conduct our regulatory and quality services in a
controlled manner. This is very important to our clients «Wipro: Wipro.com
because they need to show that their service provider has For more information on this topic, visit deskeng.com.

Simulate /// Buckling in FEA
Linear & Nonlinear

Buckling in FEA
Ensure your structures can withstand outside stresses with these methods.
BY TONY ABBEY
Editor’s Note: Tony Abbey teaches live NAFEMS FEA classes in the
U.S., Europe and Asia. He also teaches NAFEMS e-learning classes
globally. Contact him at [email protected] for details.
B
uckling occurs as an instability when a structure can no
longer support the existing compressive load levels. Many
structural components are sufficiently stiff that they will
never suffer any form of instability. These structures are classically
described as “short.” In practice, it is the relationship between ra-
dius of gyration and length that is the deciding factor, and hence
longspan girders of a heavy section could easily be clear of any
instability mode. This type of structure would only fail in com-
pression by local yielding if load levels can reach that extreme.
At the other extreme, structures that are slender could fail
at load levels well below what is required to cause compressive
yielding. The failing mode tends to be toward the classic Euler
buckling mode. For long thin rods and struts, the Euler buck-
ling calculation can be quite accurate. The buckling here is of a
bifurcation type — there is a rapid transition from axial loading
response to a lateral response, which is usually catastrophic.
A lot of structures fall into an intermediate category where
the Euler buckling calculation is not very accurate and tends to FIG. 1: Typical FEA buckling analysis set up, axially
overestimate the critical buckling load. The transition to instabil- loaded cylinder.
ity is more gradual in this category. The structure is able to carry
increasing loads, with perhaps changes in deformed shape and
plasticity, until a maximum (or limit) load is reached. At this point,
instability occurs. This may be catastrophic, or the structure may
transition to a new mode shape that can carry further load. Ex-
amples include the initial buckling of a drink can, initial buckling
of a thin wing spar shear web, or the light frame of a screen door.
Linear Buckling
The most basic form of buckling analysis in FEA is linear buck-
ling. This is directly related to the classic Euler type of calcula-
tion. A small displacement of a perturbed shape is assumed in
each element that induces a stress dependent stiffening effect.
This adds to the linear static stiffness in the element. Imagine a
guitar string tightened — the string’s total stiffness goes up and
results in a higher pitch. If the string is slackened the total stiffness
goes down, and the pitch corresponds. The stress dependent stiff- FIG. 2: First eigenvalue or mode shape.

SPOTLIGHT
ness is now subtracting from the linear static stiffness term. This
latter effect causes buckling.
In an assembly of elements in an FEA model there will be
a subtle interaction between the original linear stiffness matrix
SPOTLIGHT
and the stress dependent stiffness matrix. This is analogous to Directing your search to the
the linear stiffness matrix and the mass matrix in a normal modes
analysis. The same solution method is used — an eigenvalue ex-
companies that have what you need.
traction. For a linear buckling analysis, this will find what scaling
factors applied to the nominal static load will scale the stress stiff- GoBOXX 15 SLM
ening terms to subtract sufficiently from the linear static terms to
- Intel® Core™ i7 processor
give unstable solutions.
Fig. 1 shows a typical analysis setup. An axial load of a nominal - NVIDIA® Quadro™ 2100M
1KN is applied to the top of a thin-walled cylinder. The con- - 15.6” UHD resolution display
straint systems are shown. A linear buckling analysis is carried out. - Optimized for SOLIDWORKS
The stress stiffening matrix and the linear static stiffness matrix & other professional software
are calculated in the first linear static step. In the second step the
unstable roots are found using the two matrices in an eigenvalue
solution. The user doesn’t have to do anything special here be-
cause all buckling solvers are “hard-wired” to do the two steps. USA Call : 888-984-7589
The result is a table of eigenvalues and a set of mode shapes, Click : www.boxxtech.com
or eigenvectors. The first mode shape is shown in Fig. 2. Intel, the Intel logo and Iris are trademarks of Intel Corporation in the U.S. and/or other countries.
The critical load that will cause the first buckling mode is cal-
culated from the nominal load (1KN) multiplied by the eigen-
value (2.575E3 found in the table that resulted from the analysis).
So the critical load is 2.575E3 KN. We can see the mode shape
in Fig. 2. An important question is: Can we use the deformation Innovation Starts Here
values shown in the figure? The answer is a definite no. Just like a
normal modes analysis, all we can get is the shape of the buckled
mode. There is no meaning to the values shown in Fig. 4. The
length of the cylinder is only 1.5 m, so a displacement of 0.8581m
as shown would be well beyond any sensible result. We are as-
suming small displacement perturbations — or shapes. We have
no way of allocating displacement values.
The second important question is: Can we use the stresses
calculated from the mode shape and often shown in a linear buck-
ling analysis? The answer again is a very definite no — for two
Learn More: www.eta.com/innovation
reasons. The displacements are arbitrary and therefore the strains
and stresses are as well. The second reason is that the mode shape
is only a perturbation normal to the loading axis, so in fact does
not couple with the axial load present just before the buckle.
What do we get from the linear buckling analysis? An esti-
mate of the critical buckling load and the likely mode shape that Millions of FREE CAD Models
will result at buckling. We do not know what happens next. Will
The fastest way to model
the cylinder collapse or stiffen? What will the final stresses and
a part is to download it!
displacements be? It is rather like a freeze frame photo just at the
initiation of buckling — we are left in suspense. · Millions of supplier certiﬁed models
· All CAD formats, all FREE
The information we get is very useful in design, but it is more
of an indicator than a hard number. We also have to be aware that ™
if we use linear buckling on a structure that is more like the in-
termediate category, then we are likely to get a non-conservative
over estimate of the buckling load. We may also find the mode For those who create in 3D
shape transitions very quickly into something very different.
The boundary condition assumptions for buckling are also www.3DX-US.com

Simulate /// Buckling in FEA
FIG. 3: Cylinder buckling mode shapes 1 to 10. FIG. 4: Initial attempt at a nonlinear buckling analysis
critical. However, if the structure can be categorized as “slender” ear buckling. In linear buckling the small perturbations the struc-
and we can show a good margin over the critical linear buckling ture may see are “hard wired” into the solution. For nonlinear
load, then in many cases that is sufficient for design. analysis, the perturbations have to develop geometrically as part
Fig. 3 shows the higher order modes shapes associated with of the solution and are not pre-defined in any way. The theoreti-
eigenvalues 1 to 10. Often the default in an FEA solver is to just cal solution in Fig. 4 totally ignores fundamental facts of nature.
have the first eigenvalue and mode shape. In fact, the study of the No component can be perfectly straight, have perfect constraint
higher modes is useful. application or perfect load application. No material content will
We can see that mode 1 and 2 are identical and represent a be absolutely homogeneous. All these factors give rise in prac-
repeated mode — any arbitrary axial orientation of the funda- tice to small eccentricities and variations that attract offset axial
mental shape is possible. Modes 3 and 4, 5 and 6, and 7 and 8 loading. This in turn starts to produce offset moments that cause
are also repeated roots. We can also see a distinction between further eccentricity. For a very stable real structure, no buckling
“dimpled” shapes that have a low number of axial lobes and “Chi- will occur; but for an intermediate category real structure, the
nese lantern” modes, which have a high number. The range of eccentricities will grow until instability occurs. In a real slender
eigenvalues is also low — and actually defines critical loads of category structure, it will happen more quickly — but probably
2.57E3 to 2.85E3 KN. The implication is that any small varia- not as abruptly — as the linear Euler solution predicts.
tion in boundary condition, component detail or load eccentricity How do we overcome this limitation? Some components and
could cause any of the modes to occur. The modes are completely loading will have such a large natural eccentricity that the solu-
independent in the linear analysis; so mode 1 or 2 or 3, etc. could tion will find instability. However, for our stubborn cylinder we
occur. One way to imagine this is if mode 1 and 2 pair were not have to introduce an eccentricity. There are several ways of doing
possible in practice, by snubbing against adjacent components, this. All methods can benefit from our understanding of the linear
etc., then mode 3 and 4 pair could occur. buckling mode. The nonlinear mode may transition through this,
It is important to assess the families of higher mode shapes but it is a good starting point. We can either apply a “dummy”
and eigenvalues to see if any practical response implications occur. loading to induce the mode shape, or actually distort the struc-
However often there may be only one dominant first mode, with ture very slightly in favor of the mode shape. The first method is
the next set of modes completely infeasible and at very high criti- usually easiest, as any sympathetic load will usually work. Pres-
cal loads. These can be ignored. sures are better than point loads as they avoid local singularities.
If possible, a sympathetic pressure can be applied in the same
Nonlinear Buckling Analysis distribution as the normal displaced mode shape from the linear
If the results of a linear buckling solution suggest the calculation analysis. It can be captured as a field function and scaled to suit. It
is not representing the real response, then a nonlinear buckling is difficult to assess what level of load to apply, but it should be a
analysis is needed. This uses a nonlinear geometric analysis to lot smaller than the main axial loading.
progressively evaluate the transition from stable to unstable and In the case of the cylinder, I applied a pressure over one
addresses many limitations in linear buckling analysis. quarter vertical strip to give a net sideways thrust. This dummy
Fig. 4 shows the first attempt at a nonlinear buckling analysis. static load was set to give a deflection of 3.5e-4 m — a small wall
It is disappointing. All we see is axial shortening with no sign of deflection. The best way to do this is to ramp up the dummy
buckling. load as the main loading is applied, and then to ramp it down to
This uncovers another difference between linear and nonlin- zero until 100% main load is achieved. Alternatively, a pre-stress

FIG 6: Key point plot; radial displacement
FIG. 5: Key point plot; stress response at key element. response at key nodes.
load case may be possible, which will not get scaled up as the level achievable before the structure sees a local collapse is about
main loading is increased. The setup will depend on your FEA 82% of the linear critical load.
program. If all else fails, you may have to apply a small dummy The nonlinear method used to track the geometric instabil-
load, which will increase with the main load and be there as an ity was the Arc Length method. This is always recommended as
artifact at the end. it enables the softening response of the structure to be tracked.
The other alternative is to capture the linear buckling mode When a structure buckles it has less resistance to load. A real-
shape and apply this back to the structural mesh as an initial dis- world load controlled test would result in a runaway collapse.
tortion. Again this can use a field function from the mode which Alternatively, a real-world displacement controlled test would
is scaled and used as an offset, or the same thing can be done via allow us to monitor the reducing load resistance. The Arc Length
an Excel export. Your FEA solver may well have automated ways method permits this, even though we are applying load.
of handling this. A word of caution here though: If you make the Important considerations are: What is the effective critical
imperfections too big you will effectively corrugate the structure buckling load level and what is the post-buckling behavior? I like
and stiffen it up. Trial and error is required. to use key point plots to investigate this. Fig. 5 shows a shell ele-
The results of the nonlinear buckling analysis shows that the ment identified as being a good indicator of the changing stress
cylinder configuration and the level of eccentricity assumed re- state as instability occurs. Membrane hoop stress is zero and axial
sult in a stable structure that resists buckling until a mode occurs, stress is steadily increasing up to the start of instability at point
similar in nature to the linear mode. There is then a transition to A. This is a classic linear response. From A to B the hoop stress
a highly localized mode. The initial buckling occurs at around increases as the structure distorts. From B to C a second mode
63% of the linear estimate for critical load. The maximum load occurs with a further instability local to this element, the axial
stress decreases as the structure softens.
Fig. 6 shows the radial displacement response of two key
Summary of Key Points nodes, one at an outward moving lobe, and one at an inward
B
moving lobe. Corresponding points A, B and C correlate well
uckling is a critical failure condition for many with Fig. 5 and confirm the onset of instability.
classes of structure. Accurate estimates of In practice, the model should now be investigated for sen-
critical load and response modes are difficult sitivity to initial induced eccentricity and preferably comparing
unless a structure falls well into the “slender” category. several forms of induced eccentricity. Effects of constraint and
Linear solutions may suit such structures if loads and loading implications can be compared to the real-world condi-
boundary conditions are carefully assessed. However tions by experimenting with DOF (degrees of freedom) and using
for the majority of instability-prone structures, a full boundary spring stiffnesses. The load steps can be adjusted to give
nonlinear analysis is required. This type of analysis finer results closer to initial instability. Plastic behavior could also
is very sensitive to assumptions on eccentricity and be investigated in the transition to the second instability. DE
boundary conditions. A methodology is required
that will deal with structural softening. The key point Tony Abbey is a consultant analyst with his own company, FETrain-
method is recommended to help identify the onset of ing. He also works as training manager for NAFEMS, responsible for
instability and subsequent transitional modes. developing and implementing training classes, including a wide range
of e-learning classes.

Editor’s Picks by Anthony J. Lockwood
Each week, Tony Lockwood combs through dozens of new products to bring you the ones he thinks will
help you do your job better, smarter and faster. Here are Lockwood’s most recent musings about the
products that have really grabbed his attention.
Remcom Launches Circuit Element Optimizer

The module is part of the company’s XFdtd platform.
Remcom has a new Circuit Element (PCB), antenna, filter and other RF
Optimizer add-on module for its XFdtd structures.
electromagnetic (EM) simulation solu- Perhaps the neatest characteristic
tion. They engineered the module to of this module is that circuit elements
help antenna designers and RF (radio remain in the EM layout structure during
frequency) specialists select the opti- optimization. This should smooth anten-
mal component values for a matching na-matching workflows.
network layout, printed circuit board MORE ➜ deskeng.com/de/?p=24754
ZWCAD+ 2015 SP2.1 Released

The software is also available in a tablet version.
The thumbnail portrait is that ZWCAD+ support and more.
2015 is a low-cost yet high-performance ZWCAD+ has a bunch of nifty features,
AutoCAD alternative for mechanical CAD including one called SmartPeek that pro-
designers. It has tools for design and vides a simple keystroke to preview draw-
drafting, 2D/3D conversion, libraries of ings, and SmartMouse, a mouse gesture
standard-based parts, collaboration and that lets you execute frequently used com-
file comparison capabilities, customizabil- mands just by moving your mouse.
ity, enterprise and international standard MORE ➜ deskeng.com/de/?p=24851
CloudDDM Begins Manufacturing Services

The company’s industrial-scale 3D printing offerings have a quick turnaround.
CloudDDM is for those of you accustomed browser, you just upload your file, select
to additive manufacturing technology and your material, settle on business details like
who want to better leverage industrial-scale shipping and submit your order.
3D printing to reduce upfront production Material options include the most com-
costs, manage stock, control inventory and monly used engineering thermoplastics and
accelerate product development. resins such as ABS (acrylonitrile butadiene
You can interact with CloudDDM styrene) and ULTEM 1010.
through its dedicated app. Through a Web MORE ➜ deskeng.com/de/?p=24966
BOXX Technologies Expands APEXX Workstation Line

The new systems leverage eight-core Intel i7 processors.
The BOXX Technologies APEXX 4 7402 sors can reach overclocked speeds
and APEXX 2 3402 both run 64-bit above 4GHz.
Windows, have NVIDIA Quadro K620 With overclocking, these systems are
graphics accelerators and come with suited for multithreaded simulation and
256GB of SSD (solid-state drive) primary rendering jobs, since everything is faster
storage. Both are liquid-cooled and and you get more speed by reducing the
have filtered airflow cooling systems. bottlenecks in many CAD applications.
Additionally, their Intel Core i7 proces- MORE ➜ deskeng.com/de/?p=25094

Engineering Case Studies Fast Apps
Optimizing Medical The Neural Engineering Group, led by Dr. Marom Bikson,
focuses on prototyping and verifying tDCS devices. When
Devices with Simulation developing tDCS methods, they needed to investigate the
relationship between patient scan data and CAD models of
Soterix Medical uses Simpleware and COMSOL tDCS electrodes through visualization, analysis and export
Multiphysics to perfect transcranial direct current of finite element models, with data used by Soterix Medical
simulation technology. for device design.
Another challenge for designing tDCS protocols is to be able
I mage-based simulation is becoming increasingly common

in medical device design, particularly when creating mod-
els from 3D scan data such as magnetic resonance imaging
to easily customize models to different types of electrode place-
ment montages and current distribution, as well as for various
kinds of pathologies and material properties within the brain
(MRI) and computed tomography (CT). Models can be used derived from patient data. Models have to be both high quality
to evaluate device designs using anatomical data, and can enough to reproduce the complex internal and external features
contribute to long-term design decisions and protocols. of the brain, and able to be modified to suit different protocols.
Soterix Medical manufactures devices and develops treat- Simpleware software was chosen as the solution for gener-
ments for transcranial direct current stimulation (tDCS), a ating and modifying finite element models of the human head
promising new method for treating neurological and psychiatric from MRI data, with COMSOL Multiphysics used as the solver
disorders through transmission of a low-level direct current for applying electrostatic volume conductor and other physics.
(DC) current across the scalp. The New York-based company, in This approach has allowed Soterix and CCNY to develop detailed
conjunction with the Neural Engineering Group at City College of head models to optimize device designs to specific conditions
New York (CCNY), is applying results from Simpleware software and clinical targets.
to design protocols for tDCS devices. Image-based modeling
techniques are used to evaluate performance and reliability. MORE ➜ deskeng.com/de/?p=25319
Using Big Data and Analytics Advertising Index

to Design a Successful Future ANSYS ....................................................... C2
Siemens and Teradata partner to effectively leverage data BOXX Technologies Sponsored Report ........ 10-11
in the transportation and supply chain industries. CD-adapco .................................................. 21
COMSOL .................................................. C4
B lending old world hard metal with future-focused data-

driven analytics, Siemens Mobility Data Services division
is capitalizing on Big Data and analytics to ensure transpor-
ESTECO SpA .............................................. 13
ICO Mold ................................................... 27
tation around the globe is fast, reliable and more energy ef- Infolytica Corp. ........................................... 15
ficient. Innovations includes predicting failures, ensuring a Livermore Software Technology Corp. ......... C3
seamless supply chain for parts to reduce or eliminate down- Measurement Computing Corp. .................... 29
time and using weather data to differentiate problems in the National Instruments .................................... 3
same train model in different regions.
Okino Computer Graphics Inc. ..................... 36
In 2014, Siemens launched Vision 2020, defining a path
to a successful future. That vision requires innovation with Origin Lab .................................................. 17
data and a commitment to the mission: “We make real what Proto Labs Inc. ............................................. 1
matters.” Stratasys ..................................................... 22
Now that data storage challenges are a thing of the past, Tekscan Inc. ................................................ 25
Siemens engineers are leveraging tens of thousands of sen- Tormach LLC .............................................. 35
sors and the data they send. Data from the trains and rails,
TotalCAE .................................................... 39
repair process data, weather data and data from the supply
chain all go into Siemens Teradata Unified Data Architecture H SPOTLIGHT H
leveraging Hadoop, Teradata Aster and the Teradata Data
BOXX Technologies ..................................... 51
Warehouse.
Engineering Technology Associates Inc. ........ 51
MORE ➜ deskeng.com/de/?p=25328 3DX ............................................................ 51
➜ For the complete application stories visit deskeng.com/de/fastapps

Commentary by Tony Christian
PLM to Support Supply

Chain Collaboration
R
ecent developments in PLM (product lifecycle Taking the broadest view, users of PLM would include
management) technology have increased sub- not only the members of the design-to-delivery supply chain
stantially the potential for sharing of product but also the post-sale/installation service organizations so as
data throughout the extended enterprise. Access to give them access to the current product information as
to such technology has historically been highly variable, well as the opportunity to report product issues back.
especially among smaller members of the supply chain.
However, exploitation of cloud-based computing infra- Expanding the Reach of PLM
structures and major strides in usability are democratizing The problem has been accessibility. To take full advantage of
PLM, making it much more accessible. As a result, accurate, PLM across the supply chain, it needs to involve all signifi-
up-to-date product information can be made available right cant users of, and contributors to, the product data. In the
across the supply chain. past, PLM technologies have been essential components of
In many industries, the supply chain is multi-tiered and companies’ IT stacks but were rather heavyweight in terms
involves complex relationships. In fact, it is arguably more of IT infrastructure, process definition, data structures, etc.
accurate to use the term supply network, especially if we This is changing substantially.
take the broader lifecycle view of the supply chain that en- A combination of technology infrastructure — not only
cloud, but mobile devices, especially for the field service end
Exploitation of cloud-based of the product lifecycle — coupled with and great strides
in ease of use and configurability are making PLM highly
computing and major strides in accessible. A leading example of cloud-based, readily config-
ured and affordable PLM is Autodesk’s PLM 360 technol-
usability are democratizing PLM. ogy, while NVIDIA’s recent GRID GPU (graphics process-
ing unit) announcements demonstrate the progress that has
compasses product development at the front end and the been made in cloud-based infrastructures that enable a wider
product in-service life through to retirement at the back range of users on a wider range of devices to access complex,
end. We have also seen shifts in responsibility through- compute-intensive product information such as 3D models,
out industrial supply chains, from the organization at the simulations and visualizations.
top of the chain pushing its defined requirements for, say, Taken together, technological advances like these are ex-
components and sub-assemblies down to the supplying panding the reach of PLM dramatically. Things are moving
organizations, to those supplying organizations doing the quickly and we expect to see accelerating adoption of PLM
design and subsequent product definition themselves. The and the resulting benefits across more and more industry
increased breadth of supply chain activities and distribution supply chains.
of responsibilities makes the problem of management of I have never seen a commentary on supply chain is-
product data much more complex. sues that has pointed out that speed and cost pressures are
PLM systems have long been regarded as essential for decreasing. Those pressures, together with what seems to
managing the vast quantity and variety of data associated with be increasing complexity as industry players take a full
complex products. From their early incarnations as solutions lifecycle view of their product/service development and
for providing version control for product data and workflow delivery processes, mean that accurate single-source-of-
management through the product development phase (known truth product information is a necessity. PLM’s role as a
as product data management or PDM), today’s leading PLM supply chain enabler will therefore continue to grow in
technologies do indeed offer full lifecycle product data man- importance. DE
agement. For example, Oracle talks about its PLM technol-
ogy in terms of product value chain management in order to Tony Christian is director of independent industry analyst firm
emphasize the completion of the loop that involves exploiting Cambashi. Send e-mail about this commentary to de-editors@
field experience as an input to front-end innovation. deskeng.com.

Livermore Software
Technology Corporation
Four New Solvers for
Multiphysics Purposes
Discrete Element Sphere (DES)
The DES (Discrete Element Sphere) is a particle-based solver that implements the Discrete Element
Method (DEM), a widely used technique for modeling processes involving large deformations,
granular flow, mixing processes, storage and discharge in silos or transportation on belts. In
LS-DYNA, each DE particle is a FEM node, making it easy to couple with other rigid or deformable
structures by using penalty-based contact algorithms. The DE is highly parallelized and is capable
of simulating systems containing over several hundred-million particles.
Here are some distinct features of the bond model:

1. The stiffness of the bond between particles is determined
automatically from Young’s modulus and Poisson’s ratio. Upcoming Classes
2. The crack criteria are directly computed from
the fracture energy release rate. in Michigan
3. The behavior of bond particles is particle-size
independent. • Sept 22-25 Optimization &
Incompressible CFD Probabilistic Design Using LS-OPT
The incompressible flow solver is based on state of the Upcoming Classes

art finite element technology applied to fluid mechan-
ics. It is fully coupled with the solid mechanics solver. in California
This coupling permits robust FSI analysis via either an
explicit technique when the FSI is weak, or using an • Aug 3 Intro to LS-PrePost
implicit coupling when the FSI coupling is strong. • Aug 4-7 Intro to LS-DYNA
• Aug 11-12 Advanced Impact
Electromagnetics Options in LS-DYNA
The Electromagnetism solver calculates the Maxwell
equations in the Eddy current (induction-diffusion)
• Aug 13-14 Contact LS-DYNA
approximation. This is suitable for cases where the • Aug 17-19 ALE/Eulerian & FSI
propagation of electromagnetic waves in the air in LS-DYNA
(or vacuum) can be considered as instantaneous.
Applications include magnetic metal forming, • Aug 20-21 Smoothed Particle
welding, and induced heating. Hydrodynamics (SPH)
• Sept 23-24 Advanced ALE
CESE/Compressible CFD Applications
The CESE solver is a compressible flow solver • Nov 3-4 NVH & Frequency
based upon the Conservation Element/Solution Domain Analysis in LS-DYNA
Element (CE/SE) method, originally proposed
by Dr. Chang in NASA Glenn Research Center. • Nov 9 Intro to LS-PrePost
This method is a novel numerical framework for • Nov 10-13 Intro to LS-DYNA
conservation laws.
For more information email: [email protected] or visit www.lstc.com

Livermore Software Technology Corporation, 7374 Las Positas Road, Livermore, CA 94551, USA
FROM MODEL
TO APP
Verify and Optimize your Designs The Application Builder provides you with
tools to easily design a custom interface
with COMSOL Multiphysics® for your multiphysics models. Use

COMSOL Server™ to distribute your apps
to colleagues and customers worldwide.
NOW FEATURING APPLICATION BUILDER & COMSOL SERVER™ Visit comsol.com/5.1
Product Suite
› COMSOL Multiphysics®
› COMSOL Server™
ELECTRICAL MECHANICAL CHEMICAL INTERFACING

› AC/DC Module › Heat Transfer Module › Chemical Reaction › LiveLink™ for MATLAB®
› RF Module › Structural Mechanics Module Engineering Module › LiveLink™ for Excel®
› Wave Optics Module › Nonlinear Structural Materials › Batteries & Fuel Cells › CAD Import Module
› Ray Optics Module Module Module › Design Module
› MEMS Module › Geomechanics Module › Electrodeposition Module › ECAD Import Module
› Plasma Module › Fatigue Module › Corrosion Module › LiveLink™ for SOLIDWORKS®
› Semiconductor Module › Multibody Dynamics Module › Electrochemistry Module › LiveLink™ for Inventor®
› Acoustics Module › LiveLink™ for AutoCAD®
MULTIPURPOSE › LiveLink™ for Revit®
FLUID › Optimization Module › LiveLink™ for PTC® Creo® Parametric™
› CFD Module › Material Library › LiveLink™ for PTC® Pro/ENGINEER®
› Mixer Module › Particle Tracing Module › LiveLink™ for Solid Edge®
› Microfluidics Module › File Import for CATIA® V5
› Subsurface Flow Module
› Pipe Flow Module
› Molecular Flow Module
© Copyright 2015 COMSOL. COMSOL, COMSOL Multiphysics, Capture the Concept, COMSOL Desktop, COMSOL Server, and LiveLink are either registered trademarks or trademarks
of COMSOL AB. All other trademarks are the property of their respective owners, and COMSOL AB and its subsidiaries and products are not affiliated with, endorsed by, sponsored by,
or supported by those trademark owners. For a list of such trademark owners, see www.comsol.com/trademarks

Desktop Engineering - 2015-08

Uploaded by

Copyright:

Available Formats

Desktop Engineering - 2015-08

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

Desktop Engineering - 2015-08

Uploaded by

Copyright:

Available Formats

August 2015 / deskeng.

CAD Workstation Q&A P. 32

Design in Virtual Reality P. 23

Linear & Nonlinear FEA

WORKSTATION REVIEWS P.37

FACTORIES OF THE FUTURE P.44

DESIGN FOR COMPLIANCE P.48

Victories by Ferrari. Simulation by ANSYS.

Using ANSYS simulation technology, Ferrari is realizing its product

Sounds like Ferrari is in the driver’s seat.

Proto Labs uses proprietary software and a

ISO 9001:2008 Certified | ITAR Registered CNC MACHINING

Direct Data, Design

2 DESKTOP ENGINEERING August 2015 /// deskeng.com

You and NI will. With an integrated platform that combines flexible,

800 891 8841

37 Extreme Power and Performance

12 Dealing with the

28 VR Speeds Design SIMULATE

4 DESKTOP ENGINEERING August 2015 /// deskeng.com

6 Virtual Desktop CONTRIBUTING EDITORS

2015 NAFEMS World Congress ADVERTISING SALES

55 Advertising Index ART & PRODUCTION

55 Fast Apps A PEERLESS MEDIA, LLC PUBLICATION

51 Spotlight 56 Commentary SUBSCRIBER CUSTOMER SERVICE

Visit Deskeng.com distributed free to qualified U.S. subscribers.

Prototype/Manufacture, Engineering LIST RENTALS: For information on list rentals, contact

• Integration Address all editorial correspondence to the Editor,

Check out deskeng.com and let us know

deskeng.com /// August 2015 DESKTOP ENGINEERING 5

Can Simulation Predict the

6 DESKTOP ENGINEERING August 2015 /// deskeng.com

deskeng.com /// August 2015 DESKTOP ENGINEERING 7

Sustaining Product Design

8 DESKTOP ENGINEERING August 2015 /// deskeng.com

➜ For more coverage of rapid technologies, visit RapidReadyTech.com

The Universal Language

10 DESKTOP ENGINEERING August 2015 /// deskeng.com

deskeng.com /// August 2015 DESKTOP ENGINEERING 11

Dealing with the

Instead of requiring the user to define a formula or

12 DESKTOP ENGINEERING August 2015 /// deskeng.com

deskeng.com /// August 2015 DESKTOP ENGINEERING 13

14 DESKTOP ENGINEERING August 2015 /// deskeng.com

deskeng.com /// August 2015 DESKTOP ENGINEERING 15

Execs, Engineers and IT

16 DESKTOP ENGINEERING August 2015 /// deskeng.com

deskeng.com /// August 2015 DESKTOP ENGINEERING 17

18 DESKTOP ENGINEERING August 2015 /// deskeng.com

deskeng.com /// August 2015 DESKTOP ENGINEERING 19

20 DESKTOP ENGINEERING August 2015 /// deskeng.com

DISCOVER BETTER DESIGNS. FASTER.

de0815_Focus_Simulation_Data_Wong.indd 21 7/16/15 1:35 PM

Kenneth Wong is Desktop Engineering’s resident blogger

INFO «ANSYS: ANSYS.com

22 DESKTOP ENGINEERING August 2015 /// deskeng.com

deskeng.com /// August 2015 DESKTOP ENGINEERING 23

24 DESKTOP ENGINEERING August 2015 /// deskeng.com

deskeng.com /// August 2015 DESKTOP ENGINEERING 25

Researchers may use Dassault