Integrating Simulation and Physical Testing Leads to Shorter Design Cycles
Find out from a premier software vendor how to match the meta-physical to the physical successfully in your product design process.
Integrated into and end-to-end design
validation process, simulation and testing together yield shorter
development cycles, fewer late-stage errors, and a higher return on
intellectual property such as design, simulation, and testing data.
Factoring out habit and tradition, there is nothing to stop simulation
and test engineers from working more closely. Cost barriers that
divided them in the past have vanished. Today's simulation and testing
technology is inexpensive, accurate, and easy to use, enabling
front-line engineers to run as many simulations and tests as time
allows. With the tools available, all that remains is getting
simulation and testing in synch at the procedural level.Common Need Creates Opportunity
Just as they perform complementary tasks, simulation and testing
engineers have a common need: more certainty. With no physical object
to provide a baseline, simulation engineers are never certain their
models accurately represent the finished product. This is especially
true of boundary conditions. Engineers can calculate them to a
reasonable certainty, but not precisely. Without a proven baseline, the
simulation analysis results are suspect. On the testing side, engineers
have a physical object to work with, but that presents its own set of
challenges. Limited by the number of gauges and sensors they can place
on a prototype, test engineers are often unsure of whether they're
focusing on all of the potential problem areas. They are also limited
by their applications' representation of data in numerical form.
Testing applications depict a design as a series of abstract numerical
snapshots, rather than as an object, which makes it harder to identify
problem areas. The result of this gulf between simulation (virtual) and
testing (physical) is uncertainty on both ends. The solution is baldly
obvious. Combine simulation's visuals and testing's precision, and you
get a potent solution for perfecting designs in less time and with
fewer prototypes than conventional design processes.
Marcelo DaLuz, manager of the Power of One (Toronto, Canada) solar
car project, uses finite element analysis software to validate designs
before they go to prototyping. He recently used COSMOS-Works to
identify weak points in the front support arm for the project's solar
car and redesigned it without risking a prototype that could have
failed.
"The design was modified and reinforced at the weak points. After some
changes to the design, the analyses indicated it continued to present
the same weak points," DaLuz said. "Once again the design was modified
to withstand anticipated loads, but its mass had to be increased
considerably in order to fulfill its function. Since weight was a
critical issue, a new design was created. The new front support arm
design passed with flying colors. The finite element analysis (FEA) did
its job by being able to highlight potential trouble spots. But was it
the best possible design? The application was not dealing with
boundaries and other conditions based on a physical object. In this
case, had [our] engineers went to a physical prototype equipped with
testing sensors, they could have found their initial design worked.
They could have used data from the physical test to circle back with
the FEA application, refine the conditions, and make more accurate
predictions on the designs long-term performance."Design Cycle Appearances
So what does a design cycle that properly incorporates simulation and
testing look like? Much the same as today's cycles that feature
separate systems, with a few important differences. Engineers develop
designs in 3D computer-aided design (CAD) environments. Simulation
engineers use CAD integrated with FEA applications to validate these
designs. When they're as confident as they can be in their design, they
turn it over to test engineers for prototyping and testing. Up to this
point, the old and new development cycles are identical, but everything
after changes. In most existing design processes, testing engineers
will take the design through production without involving the design
and simulation engineers (who in many cases are the same people). They
take a first pass at the design, identify trouble spots, modify the
design, order another prototype, and test it. They repeat this process
as many times as necessary to get a production-ready design. In this
process, the physical prototype-expensive and time consuming- is the
vehicle for perfecting a design. That's mainly because the test
engineers know they can trust results based on a physical object, as
opposed to a simulation that may or may not accurately represent the
design.Simulations Yield Results
Simulations, however, could yield equally trustworthy results without
prototypes' cost and inefficiency. The key is what happens after a
company produces its initial prototype; adding a simple loop back to
analysis. The first prototype can take the aforementioned guesswork out
of simulation models. Design and simulation engineers can use testing
data from initial prototyping to perfect their models by cross-checking
their boundary conditions, for example, against the prototype's actual
boundary conditions. This yields new models that, when integrated with
testing data, visually display likely trouble spots so engineers can
concentrate on them. Perfecting and optimizing the design now occurs in
the virtual environment instead of the physical environment. In the
virtual environment, engineers can make modifications instantly at no
cost, simulate its behavior to identify potential trouble spots, and
then zero in on and correct them with testing technology. Then, when
it's time to re-enter the physical world with another prototype, that
prototype will be freer of errors and closer to production-ready.
Engineering organizations that integrate simulation and testing this
way give their companies competitive edges in quality and
time-to-market. The initial product will go to market sooner than
competitors who rely on extensive prototyping. There will be fewer
post-production errors because, freed of prototyping's cost, engineers
will simulate and test designs throughout the process instead of just
at select junctures. Modifying and upgrading products is easier because
engineers can work off the virtual "base" of models they created during
initial product development instead of starting all over again. These
gains are within reach of any engineering organization that's ready to
get a little more "real" in some areas and a little less in others.
