Much has been said in the past few years about simulation and physical
testing, and there are many theories to subscribe to depending on which
side of the fence one sits.
What is for certain, however, is that the momentum is now with CAD/CAM
and CFD. The pendulum is swinging and the only question that now remains
is to how far it will go in favour of computing power and software codes.
Can the wind tunnel really be eradicated?
“In some areas potentially this trend [of simulation replacing wind
tunnels] can and is already happening,” comments Riaz Sanatian,
automotive industry sector manager, CD adapco Group. “In the automotive
industry, for example, things like the cooling block of an engine have
already gone to virtual prototyping and simulation and this has almost
replaced experiments because the testing of this type is complex, very
difficult and expensive.”
Yet it would seem that the battle for domination amongst the two camps
in the testing world is far from over. The people backing wind tunnels
and real world testing will not lie down.
Sanatian continues: “Only time will tell when it comes to other
things like external aerodynamics. At the moment, many flow simulations
are done [for external aerodynamics] but there are still a lot of wind
tunnel tests conducted as well.
“The major issue for us in those areas is modelling things like
turbulence since we are not simulating absolute reality on the model of
reality. There are models that are coming along which should take into
account the random nature of turbulence, like larger dissimulation, but
these require huge computing power and far more orders of magnitude than
what we have.”
For the immediate future at least, wind tunnels are assured to continue
to play a vital role in the development of the vehicle. It is when one
casts an eye further into the future that one envisages potential problems
for real world testing, something that Sanatian hints at.
He says: “I would say that, at the rate we are going with increased
processing power, yes, simulation will take over. When you say will virtual
prototyping ever replace real world experience, well, of course we will.
It won’t happen in the next five years, and over the next 10 years
I can only say maybe, but when you say will we ever replace wind tunnels,
then I have to say yes.”
Sanatian is quick to look to the past in order to support his theories
on the future role of simulation and wind tunnels in the automotive industry. “The progress and usability of software like ours very much depends
on the computer technology. I don’t think any of us could have guessed
15 years ago that the performance of computers today would be so powerful.”
He continues: “The thing is, we’re not talking about one or
two orders of magnitude, we’re talking about four orders of magnitude.
“Fifteen years ago we were running calculations which involved 30,000
points – and that was considered to be really big and large. Today
we consider for 30 million points to be reasonably large, not very large.
So we’ve gone from 30,000 to 30 million, that’s three orders
of magnitude we’ve jumped.
“I would guess easily that it’s moving more rapidly, so I’d
say that in 10 years we’ll be solving simulations of the order of
hundreds of millions.”
The trust factor
Yet despite computing power fuelling the advances made in CAD/CAM and
CFD, carmakers continue to come back to real-world testing and wind tunnels,
something that Sanatian acknowledges.
“This issue is still around but less so than before,” he says.
“Ten years ago everybody used to question simulation.”
One recent study highlighted that each of the 17 major wind tunnels in
Europe provided different drag data on the same car, with the results
differing by anything up to 8 per cent plus or minus to the next result.
Sanatian continues: “There is the issue that experiments can have
errors as well, but traditionally we’ve been doing experiments for
thousands of years, and so we are confident with this process. CFD have
been around for 15 years and only in the last 10 years has it been used
extensively for industrial applications. So the idea of reliability is
becoming more and more accepted.
“People always feel more confident when they’ve been doing
something for a very long time. The difference now is instead of spending
2,000 hours in the wind tunnel, the engineer now does some simulation
and spends 200 hours in the wind tunnel. You can clear most of the silly
designs in simulation and then go and fine tune one or two designs in
the wind tunnel. There is a place for the two to work together CDF is
not here to replace the wind tunnel.” Try telling a wind tunnel
engineer that.
Staying ahead of the game
The CD adapco Group is now firmly looking ahead into the next decade
to not only keep up to pace with the rapid acceleration of computing
power, but to also build on its role as a key player in the simulation
industry.
Its flagship product – used by the likes of GM, Ford and DaimlerChrysler
– continues to evolve. “What we’ve been doing over
the past few years is looking at more than just the solver engine,”
explains Sanatian.
“We’ve put a great deal of
effort into providing tools and capabilities for the system to be
better integrated into a CAD/CAM and CAD environment with better links
to CAD systems for importing geometry and more automation for setting
up the models and meshing.”
Yet the group refuses to sit on its laurels. The company has, for
the past few years, been secretly working on a new generation of codes
called Computational Continuum Mechanics (CCM+).
Sanatian continues: “That will hopefully take us into the next
decade and keep us ahead of the game. All the CAD integration will
continue and become even more integrated, while on the engine and
solver, there are designs to utilise better the massively parallel
processing capabilities available.”
Already CCM+ has been used in the automotive industry, with one key
carmaker giving Sanatian positive feedback on the new generation of
codes. “The feedback on the user interface, and the way it interacts
with parallel computing, has been positive. It’s only been out
for a month, but all the initial signs are good.”
Despite a global downturn, the CD adapco Group – says Sanatian
– enjoyed a healthy 12 months last year. He explains: “For
us it was a reasonably good year, it was a tough one, but we still
showed growth and did extremely well in Europe and we continue to
do externally well in Japan and the Far East.
“In the
US, it’s been tough, but we’ve kept our own. We didn’t
grow too much but we didn’t shrink either. There are clear signs
now that things are definitely picking up in Japan and the US.” |
|
|
