Streamlining car aerodynamic simulation with Aerocloud

Introduction

Automotive manufacturers typically launch a new generation of vehicle model every 5 to 8 years, with regular facelifts every 3 to 4 years [1][2]. Each redesign consists of thousands of engineering changes, with some OEM’s logging 4,000 changes per year [3]. In fact, Ford, GM and DaimlerChrysler recorded a staggering 350,000 engineering changes within their supply chain in 2005 alone [4].

A large proportion of these changes require effective collaboration between designers and aerodynamicists. However, current development workflows are often inefficient, time consuming and costly.

In this article we will explore the current challenges facing automotive designers and how a fully integrated aerodynamic design platform, such as Aerocloud, can help OEM’s streamline aerodynamic development.

The problem with current automotive design and simulation

Whenever a designer makes a design change, it must be validated in Computational Fluid Dynamics (CFD) by an aerodynamicist to assess its effect on vehicle performance. These results are then passed back to the designers, who will often make further tweaks which then need to be re-evaluated. This triggers a series of revisions, resulting in the vehicle design cycling between the design and aerodynamic departments – sometimes for months.

‘This process is tedious and inefficient and at the end of it, neither side is happy,’ explains Luca Oggiano, CEO of Nablaflow. ‘The designers are frustrated as their original concept is typically overruled because they don’t have the access or expertise to use complex proprietary CFD tools. While hardware and licensing costs limit the number of CFD simulations the aerodynamicists can run, preventing them from accurately analysing each update.’

‘Furthermore, keeping track of all the revisions can be near impossible as data is often shared through emails, meetings and presentations,’ continues Oggiano. ‘So, in most automotive companies there is not a workflow which allows both the design and engineering departments to be on the same page right from the start.’

A monolithic approach to CFD

One industry that has achieved a more streamlined aerodynamic development process is Formula 1. Resource restricting regulations, along with demanding deadlines has forced teams to shift to more of a monolithic approach. This utilises a CFD platform that automates and combines set-up, meshing, simulation and post processing within a single tool, that both designers and engineers use simultaneously.

‘The days of spending hours setting up CFD simulations are long gone in high performance motorsport,’ reveals Oggiano. ‘Instead, they use highly developed CFD tools that automate set up, giving designers and aerodynamicists the freedom to spend their time on analysis.’

Accelerating CFD in automotive with Aerocloud

This monolithic approach forms the foundations of Aerocloud CFD software, which has been developed to unite designers and aerodynamicists in one optimized aerodynamic development workflow. It’s intuitive and easy to use interface allows designers to get fast and understandable insights of initial designs, helping them meet most aerodynamic requirements before being passed to the aerodynamic department.

‘It’s extremely valuable to understand how your design performs aerodynamically in the early stages,’ highlights an automotive industrial designer. ‘Aerocloud is the perfect tool for this because it is simple to use, quick and robust. You simply upload your model, spend around 5 minutes setting it up and receive results within hours.’

‘These results are automatically post processed, unlike other software’s,’ he continues. ‘So very quickly you can start assessing the aerodynamic deltas between different designs. You can even download the raw data which is useful for marketing renders.’

Aerocloud also takes care of post processing to deliver a variety of intuitive visuals

Once this initial aerodynamic evaluation is complete, aerodynamicists can then access these results and run additional simulations simultaneously thanks to Aerocloud’s cloud-based architecture. This ensures each design change achieves maximum aerodynamic performance, whilst shortcutting the design process - saving automotive companies resources, time and money.

Aerocloud’s proven accuracy and robustness

Aerocloud is based on an open source C++ code called OpenFOAM, which was originally developed back in 1989 [5]. ‘There are several reasons for this,’ explains Oggiano. ‘Firstly, OpenFOAM is an established code and so there is a lot of trust in it, not just in automotive, but within the entire fluid dynamics industry as well.’

‘Secondly, because it is open source, it is continuously updated by the community so it is already a reliable and robust platform, which we have then tuned to enhance performance,’ adds Oggiano. ‘Finally, it is the most widespread CFD code so aerodynamicists can immediately jump into using our software without a long onboarding process.’

Alongside this trust in OpenFOAM, Aerocloud has also been regularly benchmarked against other automotive CFD codes and fitted well within this data, validating its accuracy.

Aerocloud compares well with data from other CFD codes in benchmark studies.

‘The robustness of Aerocloud is unmatched compared to the other cloud-based CFD simulation software packages I’ve used,’ says an automotive industrial designer. ‘It’s the small features like dealing with non-watertight models that makes the difference. In other programs, this triggers error codes, forcing you to spend time tweaking the mesh. But in Aerocloud I’ve never experienced any errors, it just runs and gives me results every time.’

The benefits of being cloud-based

Even with the latest technology, the hardware required to run legacy CFD codes restricts development. Simulations must run sequentially, leaving aerodynamicists waiting for workstations or clusters to become available before starting the next simulation. Whilst the constant need to maintain the physical hardware and update software incurs expensive costs. 

Whereas, running CFD on the cloud avoids these issues. ‘On the cloud, aerodynamicists have access to hundreds of cores which allows them to instantly scale and run as many simulations as they want, simultaneously,’ highlights Oggiano. ‘Aerocloud can run up to 1,600 simulations on 96 cores, so any peaks within the design phase can be easily accommodated.’

In Aerocloud you can simulate different yaw angles at the same time.

Furthermore, Aerocloud is hosted on the renowned AWS cloud infrastructure and so benefits from stringent security protocols and firewalls that protects data stored on the cloud, guaranteeing security for customers.

‘The ability to run different iterations at the same time is one of the greatest features of using cloud solutions,’ comments an automotive industrial designer. ‘For some vehicle development projects, I need to run 10 simulations at once and I would never have this capacity with HPC’s [High Performance Computing].’

Flexible licensing structure

To accommodate the different needs of each customer, Aerocloud offers a range of monthly subscriptions where companies can choose the number of simulation credits they want. These can easily be scaled up or down depending on workflows.

‘If I want to quickly test a design, I can just pay for a month’s subscription to get some credits, without having to commit to an expensive yearly license,’ reveals an automotive industrial designer. ‘This flexibility makes Aerocloud a nice tool to bridge the gap for designers to complete car aerodynamic simulations without requiring unlimited budgets.’

How to run an Aerocloud simulation

First, the 3D model is uploaded as either an STL or OBJ file. Then, physical parameters such as wind speed, ground effect and any rotating parts are selected from a list of predefined options. Finally, the simulation runs on the cloud and the results are automatically post-processed, displayed through a variety of colourmaps, graphs and tables. Users can then analyse the lift and drag coefficients of different vehicle designs, interact with pressure and velocity plots and benefit from a comprehensive PDF report.

Alongside detailed visualisations, plots and graphs, users also get a comprehensive PDF report of their results.

‘In the past, CFD was run on local hardware with expensive licenses, and you had to be an engineer to launch simulations,’ concludes Oggiano. ‘It would then take days or weeks to receive results which required further post processing using third party tools. We have designed the Aerocloud platform to be fully automated and integrated, so users can get accurate results in as few steps as possible.’

‘This simplicity does not come at the cost of robustness or accuracy, both of which we continuously validate,’ adds Oggiano. ‘Overall, Aerocloud gives designers the ability to easily run CFD on initial designs and aerodynamicists the freedom to scale simulations, uniting these departments and streamlining automotive design and simulation.’

Do you want to run an Aerocloud simulation for free? Start your free trial of Aerocloud by clicking the button below.

References

[1] J.R., 2024. What Is The Difference Between A ‘Facelift’, An ‘Update’ And A ‘Next Generation’ Car? [Online]. MotaClarity

[2] What Is a Facelift in Cars? A Complete Guide. [Online]. Whatiso.

[3] K.N., B.E.N., A.D.B., 2015. Engineering change management: State of the art, a case study and proposition of a detailed model for effective management in automobile industries [Online]. Veermata Jijabai Technological Institute.

[4] A.W., G.S., R.W.V., 2011. An industry approach to shared, cross-organisational engineering change handling - The road towards standards for product data processing [Online]. ScienceDirect.

[5] About OpenFOAM [Online]. CFD Direct.