Simulation drives new cars' performance

In his second report from the European Altair Technology Conference (EATC), Robert Roe discovers the value of simulation-driven design to manufacturers in the automotive industry.

Car designers are turning to computers to create vehicles that deliver high-performance within strict emissions guidelines. But, at the European Altair Technology Conference held in Munich at the end of June, speakers pointed out that to move from a virtual design to a real product is a complicated process that involves many steps.

As CO2 regulations tighten, designers are increasingly trying to shed weight from their cars in order to improve performance as well as fuel efficiency. Maximillian Szwaj, Head of Innovation and Body-In-White Development at Ferrari, said: ‘Competition is quite tough and we really need to build and further develop our competence in lightweight design.’ There was a similar direction to the Jaguar Land Rover (JLR) keynote. Michael Buckley, Technical Specialist at Jaguar Land Rover UK said: ‘Weight is a key enabler for the lightweight vehicles of the future, but it is not the only one. You need to take a holistic approach to the design of your vehicle. You need to look at propulsion, whether using engine recovery systems or electric power systems. We need to look at weight, but we also need to look at the aerodynamics, and rolling resistance. At JLR, we are trying to address all of these things to meet the CO2 challenges that we are facing.’

Designers are turning to complex simulation tools to make this process possible. And as composite materials are now being used in manufacturing processes, the toolkit includes composite design software, as well as finite element analysis, and multi-scale modelling -- increasing the fidelity of models with higher element counts and more complex meshes.

Designs can also be optimised using software such as OptiStruct, a finite-element based technology for conceptual design synthesis and structural optimisation. By predicting optimal shapes of structures during the design process, using topology optimisation methods, OptiStruct can be used to analyse the design as it evolves rather than trying to improve a product once it has been prototyped, resulting in more efficient designs and shorter design cycles.

Weight-reduction is a key task for automotive manufacturers because it not only influences the direct performance of the vehicle – acceleration, for example -- but also aerodynamic performance, braking, handling and fuel efficiency.

Ferrari’s Szwaj said: ‘Performance for us is a top priority and one part of arriving at top performance is to be as efficient as possible. Efficiency from the structural point of view, such as weight reduction, is directly proportional to the performance of our cars.’

Buckley said: ‘It’s a virtuous circle – every gram or kilogram that we can get out of that car is going to really help on the road to reducing CO2.’ Buckley gave an example of a new vehicle designed at JLR, through optimisation of castings for body panels and other weight reductions using topology optimisation software. JLR reduced weight to the point that it could replace parts of the power train and achieve the equal performance figures with the new design. Reduced weight means that less power is needed – this in turn leads to more weight savings in the power train which has a real impact on fuel efficiency and helps to reduce CO2 emissions

However, Szwaj explained that in a real manufacturing environment things are not always as perfect as a model may predict. Szwaj said: ‘what a lot of people do is they arrive at a certain design with the FEA and they say “ok, well this is a perfect solution” because in FEA it is always perfect. But then you start to introduce the errors, with manufacturing, porosity and maybe some adhesion problems, and what people do is they add extra layers of carbon and the weight is not becoming very efficient.’

Both speakers talked of simulation-driven product design and using these kinds of software early in the design process to achieve the maximum reduction in cost of design and shortening time to market, but there have to be strategies in place to deal with these kind of issues when they arise. To stay as true as possible to the FEA design takes a lot of innovative tweaking of models and manufacturing so that the predicted performance can be achieved.

Szwaj said: ‘Of course, we have to arrive at a safe product but there is a limit, where your FEA predicts a certain level of safety. Normally what people do in aerospace is they double the thickness, but we didn’t want to do this because the weight is very important to us. We wanted to arrive at a solution that was as close as possible to the safety margin, to that FEA calculated prediction.’

To reduce the difference between predicted and actual performance while still in the concept phase, Ferrari have been working on a new process that they call C123. Szwaj said: ‘We would like to arrive at the project-phase level of maturity much earlier and we have introduced, with Altair, new development studies in the concept phase.’

 ‘Concept C1 is basically a topology concept, where we have different variants,’ Szwaj explained. The designers will run simultaneous variants through topology optimisation under ‘70 load cases, both static-dynamic and crash’ because this is the ‘only way to analyse these models simultaneously.’

In the C2 phase, the development team are ‘still using quite low-fidelity models,’ said Szwaj. ‘We are trying to understand the importance of the sections, the joint development and of course we are still looking at the target development, and we can adjust the target. In the C3 phase, the models are becoming more complex and we are still doing some linear analysis.’ At this point, he explained, the designers do dynamic analysis by modelling the behaviour of the design in a crash, adding to the fidelity of the model but also to the complexity of the simulation.

However this is not a linear design phase, as Szwaj stressed: ‘It’s not a push button design. We need to follow this process very carefully and understand the importance.’ This can mean moving from C3 back to C2 or C1 if that is required, to fully understand the changes being made to the design.

Szwaj said: ‘In the middle phase, we are analysing our own ideas, we are constantly reviewing the ideas that are generated by the team that is working with us.’ This constant review process helps to drive innovation in the design process at Ferrari. This is something that Buckley and JLR are also encouraging. Buckley said: ‘In the CAE community, we need to continually strive up the innovation chain to drive cost and weight out of our vehicles and now I can see some of this multi-scale modelling that is out there today as maybe being a key enabler for that.’

Twitter icon
Google icon icon
Digg icon
LinkedIn icon
Reddit icon
e-mail icon

For functionality and security for externalised research, software providers have turned to the cloud, writes Sophia Ktori


Robert Roe investigates the growth in cloud technology which is being driven by scientific, engineering and HPC workflows through application specific hardware


Robert Roe learns that the NASA advanced supercomputing division (NAS) is optimising energy efficiency and water usage to maximise the facility’s potential to deliver computing services to its user community


Robert Roe investigates the use of technologies in HPC that could help shape the design of future supercomputers