BorgWarner Goes Full Speed with ANSYS HPC Parametric Packs

It doesn’t matter what car you drive — it could be a snazzy Ferrari or a humble FIAT Punto — ultimately what we’re all looking for is a car that performs well and maybe saves us a little money at the pump.

The upcoming joint ANSYS-ESTECO webinar on September 15th will discuss just how important a single component, in this case, a tensioner arm, can be. Chain tensioner arms may not be as well known as pistons and gearboxes, but, by maintaining the correct amount of tension on the chain at all times throughout its duty cycle, they are important for reliable operation of the accessory chain drive system. The chain tensioner also helps protect other components, such as the alternator and water pump, from undue stress and premature failure. A well-designed chain tensioner can also help boost engine performance and efficiency.

Baseline and Optimized Design for the Ferrari V8 engine tensioner arm

Baseline and Optimized Design for the Ferrari V8 engine tensioner arm

The webinar shows how Borg Warner, the automotive component and parts giant, used modeFRONTIER optimization and integration capabilities with ANSYS Workbench and ANSYS Mechanical to improve the tensioner arm for a Ferrari V8 engine (California and 458 Italia) and a FIAT Punto engine. ANSYS HPC Parametric Packs were also used to reduce the time to insight.

Before we begin, a little backstory may be in order. In automotive design, the process of enhancing performance while improving fuel economy and limiting production costs involves heavy calculations and a variety of analysis tools.

Optimal solutions for the FIAT Punto engine tensioner arm

Optimal solutions for the FIAT Punto engine tensioner arm

Solving complex, multidisciplinary design problems such as the optimization of a tensioner arm, represents a challenge in terms of hardware/software setup, especially when you consider how difficult it can be to integrate the different tools required in the process. Recent advances in hardware solutions such as powerful workstations and HPC systems, have increased the feasibility of design processes that involve high-fidelity simulations. The fact that this kind of hardware can execute computationally intensive simulations in parallel means that process times have significantly been reduced.

Let’s get back to the tensioner arm, the Ferrari and the Punto. The tensioner arm is a typical example of a multiobjective problem, that BorgWarner engineers effectively solved by integrating ANSYS Workbench and ANSYS Mechanical software within modeFRONTIER optimization and integration platform. To improve performance, engineers needed to work on reducing the weight, and minimizing the maximum stress and deflection of the arm. The optimization of the component for the Ferrari was geared towards increasing overall performance; for the Punto the focus was on choosing the proper material and cutting production costs. Coupling the ESTECO modeFRONTIER multidisciplinary optimization platform with the ANSYS HPC Parametric Pack allowed BorgWarner engineers to run multiple designs simultaneously and amplify the usage of available licenses – solving the pesky problem of hardware we talked about in our little backstory. In fact, by running the entire process over HPC, the optimization run time went from 36 days to 10 days.

Both designs achieved their goals with some tradeoffs. For the Ferrari tensioner arm, engineers sacrificing weight reduction for minimizing the maximum stress and deflection in a delicate area of the component where it contacts the piston. This achieved the overall goal of increasing performance for the Ferrari. For the Punto, they prioritized cost and weight reduction over reduced stress and deflection.

Leveraging advanced optimization software with HPC brought better solutions for different scenarios in a small fraction of the time.


One thought on “BorgWarner Goes Full Speed with ANSYS HPC Parametric Packs

  1. It must first time that the words ‘Ferrari’ and ‘saving a little at the pump’ appeared in the same sentence! So I was a bit disappointed to see that this wasn’t an optimisation criteria for the Ferrari model after all 😉

Comments are closed.