Formula 1: Multiple Connected Components and Long-range Aerodynamic Correlations

1 Introduction

Formula 1 engineering requires a very high degree of accuracy since the slightest improvement can have a decisive impact of the aerodynamics and hence, on performance. Racing teams’ Engineers use high-fidelity numerical simulations to assess the effect of design variations, without actually building the real prototype.

2 Method Overview

2.1 Numerical optimization using Neural Concept Shape

Computational Fluid Dynamics simulations are at the heart of the car development process of all Formula 1 teams, and one of the key performance factors for the race. Even though teams can generate thousands of simulations every year, and explore as many design variations, they are limited both by the computational cost and the reglementation from the FIA. Among this large number of designs, it is critical to be able to converge quickly towards designs that correspond to a reasonable compromise. Hence, running approximate simulation results without re-running expensive simulations every time can be critical in this phase. In the case of Formula 1, low fidelity simulations are not an option since the very complex nature are the phenomena can only be captured by high-end CFD resolutions.

A classical approach to reduce the number of simulations and quickly map the search space is to use surrogate modelling via Gaussian Process (GP) regressors. This interpolator is then used as a proxy for the true objective to speed-up the computation, which is referred to as Kriging in the literature. However, those regressors are only effective for shape deformations that can be param- eterized using relatively few parameters and their performance therefore hinges on a well-designed parameterization. Furthermore, Formula 1 teams generate hundreds of simulations per day and have therefore sometimes accumulated a very large dataset of existing simulation, which do not share a common paramterization and therefore cannot be directly exploited by such parametric approaches. Finally, all previous available approaches fail to accurately reconstruct 3D flow or fields on a surface or a plane. They all tend to over-smooth the solutions and produce outputs for which the structure looks nowhere near the one of the initial flow. This may be explained by the linear nature of those approximation methods, which goes against the strong non-linearities of the underlying physics.

A recently developed software, Neural Concept Shape (NCS), opens the door to a solution to this problem through Geometric Convolutional Neural Networks, a new way to build surrogates of numerical simulator. It is agnostic to the shape parameters as it processes directly the mesh representation of the design. Hence, design parameters are decoupled from the learning problem and a single predictor can be trained with a large amount of data and used for many optimization tasks. Unlike Kriging methods, the engineer does not have to choose and stick to a specific parametrization from the beginning to the end of experiments. Furthermore, it can leverage on transfer learning abilities of Deep Learning models to blend simulations from multiple sources and different parametrizations. To date, NCS is the only CAE-oriented Deep-Learning code that is able to process raw unstructured 3D data directly withing the scale of what is used in Formula 1 simulations. It uses multi-scale geometric neural networks, through a combination of surface — such as geodesic — and euclidean network architectures. It is able to learn to emulate simulators and reproduce multiple outputs such as integrated scalar values — e.g. drag forces or energy consumption — or surface and volumetric field values —e.g. pressure or temperature.

2.2 Objectives

In this project, we evaluate the performances of the NCS’s ProjectiveRegressor Neural Network model and provide a first element of answer with respect to the feasibility of the use of NCS in F1 racing. We focused our attention on the front side of a Formula 1, from the front wing to the bargeboard, including the nose of the car.

Due to confidentiality and accessibility reasons, note that the data and simulations used within this project are a simplified version of the more complex geometries and simulations found in operational conditions in Formula 1. The CAD design does not offer all the details and subtleties found on nowaday’s cars, and the CFD was resolved at a lower resolution than what is done in the most competitive teams. However, the problem presented here exhibits most of the practical challenges found in operations, including long range interactions, complex vortex structures, multiple connected components and importance of geometrically fine details.

2.3 CAD design

OnShape, a CAD software hosted on the cloud was used to design the Formula 1. This geometry was generated by varying up to 56 different variables, detailed below:

Figure 2: Parameters description
Figure 3: Parameters description
Figure 4: Parameters description
Figure 5: Parameters description
Figure 6: Parameters description
Figure 7: Parameters description

This parametric space allows to perform very fine design variations on the car, and ultimately to assess the influence of each of these parameters on the aerodynamic performances of the car. To generate the training dataset, we automated the geometries generation by using the OnShape API. We defined lower and higher values for each of the parameters, and a random value was chosen at each iteration for the parameters. We were able to build a database of 1644 geometries using this method. It ensured a good distribution of the geometries, and each of them is unique in the dataset. An example of the foils variability within the dataset is shown with figure 8 where it is possible to see a sub-group of 5 shapes representing the variation of the shapes of foils that were generated.

From Onshape, we can export the shape as an STL file. The .stl is then remeshed using Openfoam remesher blockmesh. Similarly, the volumetric mesh is created and following its discretization the mesh of the vertical plane just behind the tyre. Note that due to its irregular discretization we are allowed to have a more fine-grained refinement close to the car shape and a coarser one far away from the car.

(a) Foils seen from the front.
(b) Foils seen from the top.
Figure 8: Overlapping front foils taken from 5 randomly sampled shapes.

2.4 Numerical Simulation with Openfoam

OpenFoam was used to perform the numerical simulations. It is a free, open source software for Computational Fluid Dynamics simulations. The car is run at atmospheric conditions, at 250 kph (70m/s) in straight ahead condition (no crosswind). We therefore use half of the car to spare compute power. The wheel is defined as rotating according to the car speed. The turbulence model used is SST k-omega transport. More details on this model can be found here.

Simulations were performed on the surface of the car, as shown with figure 9, but also on a plane slice taken at the last iteration of flow field at 1.7m, i.e just behind the tyre. Indeed, usually, taking the whole volumetric prediction is too expensive for high-fidelity simulations. Hence using plane slices is a good trade-off and allows to assess the influence of slight geometry variations more precisely. This vertical plane slice also allows to observe some very important flow phenomenon that are happening alongside the Formula 1, and that aerodynamicists want to control as good as possible. A visual example is given with figure 10.

Figure 9: OpenFoam simulation of the pressure field on the surface of the Formula 1
Figure 10: Pressure field on a plane slice from OpenFoam simulation
(a) With hidden front part of the car.
(b) With visible front part of the car.
Figure 11: Streamlines behind the tyre of the car.
(a) With hidden front part of the car.
(b) With visible front part of the car.
Figure 12: Pressure behind the tyre of the car.

3 Setup in Neural Concept Shape

3.1 Geometric Neural Network

We use the recent ”ProjectiveRegressor” defined by the NCS framework as our Geometric Neural Net- work regressor. This model allows for a query mesh different than the original shape to be specified. In this case it is the vertical plane just behind the tyre, in which vortexes are present. Note that the network is unaware of the underlying parametrization and shapes created with other parametrizations can easily be predicted or added to the dataset. Only the similarity to the shape present on the train- ing set will have an effect on the accuracy.

The part using convolution on the shape is illustrated in figure 13. The first part of the model pre-processes the input and constructs a set of features by means of the previously introduced geodesic convolution operations. These features are then used to predict the global scalars via average pooling and two dense layers. The second branch of the network generates pressure fields relying on an additional set of geodesic convolutions and point-wise operations. The new network takes advantage of a GPU efficient implementation of geodesic convolutions.

Figure 13: Network architecture

4 Results

4.1 Predictions

As a first step, we predict on the surface of the car and evaluate the performances. The solution on the surface of the car is quite stable with respect to the deformations of the shape. The all-point-mixed R2 coefficient on the pressure of the surface car is: 0.916. The distribution of the R2-score over the samples is given with figure 14 and an visual comparison is given with figure 15

Figure 14: Distribution of the R2-score over the test samples for the pressure prediction on the surfaceof the F1 car

As a second step, we can predict the pressure on the vertical plane just behind the tyre. The all-points-mixed R2 coefficient on the plane is: 0.9674995.The distribution of the R2-score over the samples is given with figure 16 in Appendix A. Some visual comparisons between the ground truth (simulation done with OpenFoam) and instantaneous predictions of the neural network are provided with the figures 17,18,19 and 20 located in Appendix A. 1:

Figure 15: Comparison of the pressure distribution on the surface between the ground truth (top)and prediction of the neural network (bottom) for a test sample

5 Conclusion

In this report, we analyse the capabilities of the neural network to perform instantaneous predictions on the surface and on a vertical plane located in the volume behind the wheel of the car. This end- to-end workflow demonstrates the abilities of the NCS’s ProjectiveRegressor module to predict flow fields (pressure and velocity in the three-directions) with a good accuracy and by using as only input the geometry of the car. A sensitivity study was briefly performed, to unveil the potential of such a method in terms of design optimization.

Although the simulations considered here are a simplified version of the higher fidelity and more detailed models used by competitive teams, it opens the door to the perspective of re-using all the legacy simulations produced and stored by Formula 1 teams over the years. It will make it possible to explore more geometries and converge faster towards really promising designs, without re-running an undue number of simulations.

6 Appendix A

Figure 16: Distribution of the R2-score over the test samples for the pressure prediction on the verticalplane
Figure 17: Comparison between the ground truth and the prediction for the pressure distribution andthe velocity magnitude on a test sample
Figure 18: Comparison between the ground truth and the prediction for the pressure distribution andthe velocity magnitude on a test sample
Figure 19: Comparison between the ground truth and the prediction for the pressure distribution andthe velocity magnitude on a test sample
Figure 20: Comparison between the ground truth and the prediction for the pressure distribution andthe velocity magnitude on a test sample

Contact

Name

Company name

Company email

Message

Thank you!

Our manager will contact You shortly.

Homepage
Oops! Something went wrong while submitting the form.

Subscribe

Name

Your email

Thank you!

We will update you with events and news about NCS.

Homepage
Oops! Something went wrong while submitting the form.

Demo request

Name

Company name

Company email

Message

Thank you!

Our manager will contact You shortly.

Homepage
Oops! Something went wrong while submitting the form.

Download

Name

Company name

Company email

Thank you!

Our manager will contact You shortly.

Homepage
Oops! Something went wrong while submitting the form.

Download

Email

Oops! Something went wrong while submitting the form.