CFD for Automotive Industry

cfd for automotive


The automotive sector with its very ambitious design targets and tight time deadlines utilizes CFD in many areas on a daily basis. The topics such as chassis aerodynamics or IC combustion modeling are the most eminent examples of flow simulations. However, fluid flow and heat transfer analyses are utilized for many other purposes.

They include:

- engine compartment cooling

- HVAC of the passenger compartment

- fuel sloshing in the tanks

- oil flow in the gearboxes

- modeling of dirt

- simulation of exhaust ducts

As computing power and software gets more available, it makes sense to use CFD even in simple design cases.

On the other hand, the cars turn into very complex machines packed with sensors and lots of secondary machinery. Hence, new areas of engineering arise. Many of them (battery, sensor dirt-interaction, high-current chargers) will surely benefit from physical simulations.

Efficient cooling for electromobility

One of the key progress drivers in the automobile industry is the electro-mobility. The range and usability of electric cars improve on a yearly basis. Still, the thermal management of batteries or electric superchargers is a very demanding task. Battery cell’s performance strongly depends on the external temperature, own voltage, age and charge level. The charger’s performance strongly depends on the cooling efficiency as very large electric currents inevitably lead to heat generation during conduction. In both of these areas, CFD definitely plays one of the major roles.

User Story

cfd for automotive
Valeo is a multinational Tier-1 automotive supplier based in France. The company provides a wide range of products to auto manufacturers and after-markets. In one of their manufacturing facilities, they routinely design frontal radiators. In this process, they have been using the standard CFD package for modeling of the process of heating up of the new designs.  It provided high-fidelity results – the discrepancy between the simulations and the manufacturer’s experiments was less than 0.5%! However, this approach had several drawbacks.  A single simulation involved the work of 100 computing cores for 36-48 hours, which was a huge cost. Preparation of the geometry was also very time-consuming (CAD in 3D + meshing). The entire design cycle, although well established and validated, was 14-days-long.

It was necessary to accelerate the design workflow. But it was impossible unless some parts of the process were simplified.

QuickerSim delivered a dedicated application for simulating heat transfer in the radiator using specifically tailored model (not available in standard CFD packages). Such a solution did not require CAD geometry and mesh generation as the shape was parameterized with a few parameters and the radiator surrogate was build automatically. The calculations also lasted much less – it took only 45 minutes for a single desktop processor to complete the job.  The accuracy of the calculation turned out to be 2%, what is acceptable in most cases.

What were the results? The project cycle has been shortened from 14 to 4 days, and over 80% of cases are currently being analyzed using our solution! The costs of licenses and electricity have also been reduced.