|
Springback simulation tools make aluminium more attractive
as a body material, but there are other factors that need to be taken
into account. Tristan Honeywill reports
Up until about ten years ago, spot-welded mild steel cars were the mainstay
of the car industry. Engineers were able to draw on more than 70 years
of experience in pressing and joining the metal and things were relatively
straightforward. Things have changed. In the late 1990s came a massive
focus on weight. OEMs wanted to give customers better safety and more
features at the same or lower weight and so started looking seriously
at high-strength steels and aluminium.
It triggered a proliferation of steel development and radical ultralight
vehicle concepts. But many were just laboratory steels and when the steelmakers
started to perfect their metals, it became clear that the supply base
was not using a single set of mechanical properties and specifications.
TRIP steel from ThyssenKrupp could be quite different to the metal supplied
from Arcelor. With material properties changing over the space of two
or three years – roughly the length of a programme – you could
end up with different crash performance or NVH in production than you
achieved in development.
 Different
steel manufacturers’ tolerance ranges can vary significantly too.
If you are trying to form a part reliably in a press all this can make
it very difficult to achieve your target shapes. Unlike with mild steels,
it’s best not to take anything for granted with high-strength steels
and aluminiums.
At smaller OEMs such as Jaguar-Land Rover, that makes development more
complex. Mark White, chief technical specialist for body structures, looks
after their entire lifecycle, working closely with the stamping and product
engineering teams. “Prior to Job One any technology has to pass
through a series of gateways and each phase requires the manufacturing
people’s buy-in,” says White.
At GM, the resources are larger, but the issue then is setting the right
global standards. GM stamping architect Terry Lanker works mainly on side
closures, but cannot make any commitments to product engineering unless
he knows it’s good globally: “We’re organised so that
we can design a car in Europe, do the dies in the US and make the cars
in China – you have to know that a practice or tolerance is the
right thing to do in every region,” says Lanker.
Naturally, the more homework you do, the easier the design and manufacturing
gets and the better placed you are to tackle springback or formability
issues on new materials.
White makes no secret of the fact that springback has been difficult on
Jaguar’s aluminium vehicles. When work began in 1998, CAE tools
for springback were rarer than today and so Jaguar had to do a lot of
prototype work and developed a simulation package with Ford to cope with
the issues.
Steve Pilz, product manager at Ansys, says most packages now have springback
simulation: “It’s a question of entering the correct material
laws and then taking the springback into account in the die design,”
he says.
Springback prediction for aluminium isn’t too difficult, depending
on what you mean by that. Counter measures in the die are possible –
simple things like swages, “bird beaks”, limits on the depth
of the side wall and secondary flanging operations can help. The way the
part is trimmed must also be different.
“The danger of putting your springback compensation on the part’s
design is that you can over-compensate or find you don’t get the
same amount of springback as before because the geometry of the part has
acted as a counter measure,” says White. “We put springback
compensation into some of the XJ parts only to find that a 3mm of compensation
in the flange produced a 2.7mm error.”
Springback changes depending on the geometry and gauge, he says. Heavy
gauge parts with lots of form can have springback issues, so can light
gauge parts. Medium gauge are more predictable. Jaguar’s final countermeasure
is a CAE tool that tells it how to recut the die to correct any springback
in the physical parts.
Jaguar allows three to four die tryouts, occasionally requiring only one.
It is hard to say whether comparisons with GM’s side closures engineering
are justified, but Lanker says one tryout is typical.
“The physical die is simply a confirmation,” says Lanker.
“If the math-based system runs well, you can get it right first
time. We get some simpler panels like roofs and hood outers right on the
very first hit. Body side outers can have problems, but these aren’t
in the draw; they’re in getting the dies the fit correctly –
they’re in the trim and the flange.”
Lanker believes accuracy in springback analysis comes down to experience
and skills – everyone has more or less the same computers and software.
Having the right structures to keep an organisation’s information
current and work standardised is essential in order to keep improving
is essential too, however.
“The most common problem we encounter is the confidence of our own
people,” says Lanker. “Many GM people aren’t familiar
with the capabilities of the math prediction and they tend to throw cold
water on new aluminium applications.”
But springback prediction is not just about the material – it’s
about the process and the variables in the press shop too. Jaguar tests
its operations by increasing the blank holder force by plus or minus ten
per cent to look at what the optimum setting is. Similarly GM moves the
blank around in the die in the simulation to see its effect on the forming.
Lubrication is also critical. This used to be a plant by plant decision
at GM. “We’ve standardised this globally now,” says
Lanker. “This controls the coefficient of friction in the forming
codes. The maths tools assume the worst and then you verify the part can
be formed successfully so you’re always robust.”
It’s important to understand in detail the effectiveness of the
lubricant in the stamping operation, says White: “A good lubricant
such as the dry wax films used for deep-drawn parts sticks to the part
well and also builds up on the dies.”
Future simulation techniques will have to include more of these environmental
variables before die tryouts become obsolete. “We need to understand
manufacturing parameters like tool temperature, ram speed, press forces,
press shop temperature even,” says White. “They all have an
effect on the dimensional accuracy of the part.”
© Automotive Engineer, 2007 |
|
