Personal Projects

Edited the unofficial End Of Season Recap that was presented at the Sponsor Acknowldgement Night. This video features clips from manufacturing, testing, and Europe competitions. Some clips used are from the official FSG and FSAA livestreams, all right regarding those clips are theirs.

Designed a mock-up of a WEC Prototype as a fun CAD project. This car takes inspiration from the Audi R18 TDI, the car known for the diesel ban its supremacy led to. The front end however is also inspired by Formula designs with a more pointy nose in an effort to allow for better flow management. CFD was never conducted on this prototype but that might change someday if I'm bored and out of ideas.

The first iteration of this design was made on CATIA V5 before switching to Solidworks n order to get used to different CAD softwares. Screenshots were taken with Rhinoceros 3D for its more pleasant user interface and pre-rendering functions. IGES file of the prototype is available for download in the Archives section.

As part of the cornering aerodynamics and lap time simulation projects, it was necessary to create a lateral tyre model that would allow for the calculation of the steering angles and sideslip angle to input into the CFD model. Later, this tyre model was re-worked to include longitudinal forces, accelerations and establish a friction ellipse to use in the lap time simulation.

Caution : all data, values and results thereafter have been offset/tempered to guarantee confidentiality tied to the FSAE Tire Test Consortium. The methodology is however valid and proves good results when used with the data found on the FSAETTC website.

Raw Data

Raw tyre data was extracted from the private files of the FSAE Tire Test Consortium. The FSAE Tire Test Consortium (FSAE TTC) is a volunteer-managed organization of member schools who pool their financial resources to obtain high quality tire force and moment data targeted for Formula SAE and Formula Student competitions. The FSAE TTC is organized by Milliken Research Associates, and the tests are ran at the Calspan Tire Researh Facility , Buffalo, NY.

The particular tyre used by Formule ETS that this model is based on is the Hoosier 16x7.5-10 R20 compound present in the Round 9 of testing. Most of the exact data will remain undisclosed here due to the participative nature of the TTC, and the necessity to fund it to have access to the data.

The first objective was then to extract the data from the run files provided by the TTC, meaning it was necessary to filter out some parameters to obtain the necessary curves.

The second step was then to curve fit this raw data for each Fz load case following the 1996 edit of Pacejka's Magic Formula. This Formula is defined by :

$F_y=D\cdot sin(C\cdot arctan(B\cdot (\alpha +S_h))-E\cdot (B\cdot (\alpha +S_h)-arctan(B\cdot (\alpha +S_h))))+S_v$

$F_x=D\cdot sin(C\cdot arctan(B\cdot (\kappa +S_h))-E\cdot (B\cdot (\kappa +S_h)-arctan(B\cdot (\kappa +S_h))))+S_v$

where D is the peak factor, C is the shape factor, B is the stiffness factor, E is the curvature factor, Sh is the horizontal shift, Sv is the vertical shift, α is the slip angle, κ is the slip ratio.

The curve fitting was achieved through a Python script to a root mean square error (RMSE) below 30N. It was performed on multiple Fz load cases for both lateral and longitudinal Pacejka Magic Formula to then create a general all-Fz approximation for the range of vertical loads experienced by our FSAE prototype. The loads experienced by a tire range from 400N to upwards of 1000N.

The general formulas are created by making the Pacejka coefficient load-dependent, rendering them B(Fz), C(Fz), D(Fz), E(Fz), Sh(Fz) and Sv(Fz). For the lateral magic formula, we assume Sh = Sv = 0. Those generalized formulas allow for the plotting and calculation of any Fz case.

From then on, we are able to compute any lateral/longitudinal load from those two magic formulas assuming we know the vertical load and the slip angle(or ratio).

Knowing the load is rather easy for an FSAE car as we know the aerodynamic coefficients for a wide operation range (ride heights, roll, pitch, yaw) and the weight of the car. Knowing the geometry and properties of the suspension, we can also model the load transfer and calculate exactly the load experienced by a tire for any condition.

The slip angle / ratio is however mainly tied to driver ability, and creates a sense of random in how the car will actually react. When in a context of simulation, we can assume a perfect driver that maximises the tire load at any given moment, or introduce and "random" thresholded variable to account for driver skill. The goal here is to fix the slip angle to other variables so the Fx/Fy computation is straightforward.

From there, whether we assume a perfect or human driver, the formulas give us a theoretical maximum for Fy and Fx that we can use to create a friction ellipse. The prototype being a 4-wheel-drive electrical vehicle, with an almost exclusively regenerative braking, we will assume that forward and backward Fx are equivalent resulting in a way more ellipse-shaped friction ellipse, while more usual friction ellipses present a non-symmetrical behavior. Our friction ellipse is defined by :

$(\frac{F_x}{|F_x{|}_{max}}{)}^2+(\frac{F_y}{|F_y{|}_{max}}{)}^2=1$

Similarly, we can produce optimal slip angle / vertical load correlations, or maximum lateral force / vertical load graphs that will facilitate the computations within a lap time simulation.

The combined slip approach offered by the use of a friction ellipse will give more realistic results for the loads created by the tire. A more precise and thorough tire model will increase the fidelity of a lap time simulation. The key here is for the solver to catch the vertical load experienced by the tire, and the repartition of longitudinal/lateral force for the studied segment. From that we can compute a realistic longitudinal and lateral acceleration on the car.

Tires are very complex, and regarding this complexity this tire model is very simple. For example, it does not yet account for pressure effects, temperature effects, camber angles or surface roughness. All those additions towards a more precise model (such as a Pacejka 5.2 or 6.2) remain to be integrated for a more precise simulation. However, since the lap time simulation focuses more on the impact of aerodynamics on a car rather than a full-car comprehension, we will stick to that simpler model.

For now, this model is used to compute steering and sideslip angles for the Cornering Aerodynamics Project while offering a basic tire modeling for the Aerodynamics-focused Lap Time Simulation Project.

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Paper available in the archive section of the website under the reference FETS-2025-01, under the "Correlation of Aerodynamic Models to Experimental Data". Integration to the website is on-going.

As a side project, I was asked by the ÉTS Student Rocketry team, RockÉTS, to simulate, analyze and optimize their home-made hybrid rocket engine (thereafter HRE) with CFD methodologies. The challenge of this project lied in modeling the combustion phenomenas specific to the function of HREs. Specifically, RockÉTS' HRE works off the chemical reaction between Nitrous Oxide (NOS) and paraffin wax (C31H64) through complex pyrolysis and carbon decomposition mechanisms. The paraffin being a solid, some assumptions had to be made to simplify this problem, and create a RANS 2D-Axisymmetric baseline before scaling the model to DES or LES methods in 3D space.

This section explains the different steps towards modeling a HRE, and serves as introduction for further rocket combustion work.

Pana Hybrid Rocket Engine

Pushed by the desire for innovation and after years of using commercial solid-fuel rocket engines, the RockÉTS student club began the design of a homemade rocket engine to compete in the SRAD (Student Researched and Designed) category of rocketry events.

Started in 2018, the Pana program aimed at designing a hybrid rocket engine that would power future rockets manufactured by the RockÉTS Student team. After 7 years of development, Pana made its debut at the 2025 IREC event but suffered a mechanical failure before its launch. However, multiple static fire test of both the lab-scale and rocket-scale prototypes of the engine have been successful, and allowed to gather sufficient data to validate a CFD model of the combustion. The combustion process involves an industrial-grade paraffin wax (C31H64) and nitrous oxide (N2O) that brings the thrust of this motor upwards of 4kN with a 7 seconds burn-time, that would thoretically bring RockÉTS' latest rockets to speed upwards of Mach 2.

The first objective of this project is to provide a first modelization of Pana, by recreating the principal features of the engine in a simplified setting to prepare for CFD integration. Pana can be decomposed in several parts, each of them being subject to optimization to modify the engine's performance.

1. Pre-injection Chamber : the nitrous oxide is vaporized from the tank into this chamber before being injected in the combustion chamber.

2. Injector Plate : the nitrous oxide is injected through optimized holes to improve the mixability of the nitrous oxide with the parrafin vapors.

3. Engine Starter : a fire starter that ensures the beginning of the pyrolysis process of the paraffin wax and the ignition of the engine.

4. Fuel Grain Cell : a solid piece of paraffin wax with one or more ports for the nitrous oxide based mixture to flow through.

5. Post-combustion Chamber : a chamber to reduce the flow velocity of the gases and reduce the loss of nitrous oxide in the exhaust.

6. Exhaust Nozzle : an optimized shape exhaust to accelerate the gases to supersonic velocities and produce an increased thrust from the combustion while reducing over-expansion and under-expansion shocks.

7. Pressure Sensors : (only on the static fire versions of the engine) allow for the precise measure of pressure differentials within the engine, resulting in precise correlation data.

8. Engine Casing : a thermo-mechanical resistant cover to accomodate the extreme temperatures and pressure within the combustion chamber.

The precise Pana assembly is very complex and follows high standards of manufacturing to ensure the success of every static fire test and rocket launch with this engine. However, this assembly is too detailed for CFD simulation purposes as is. The first step towards the simulation of the combustion process resides in simplifying the geometries and features of this assembly. This step will ensure the quality and simplicity of the mesh, resulting in a more robust and stable solution down the line. Thereafter is a longitudinal section view of the CFD assembly of Pana, color coded to differentiate important parts that will have different boundary conditions or optimization targets.

$F_y=D\cdot sin(C\cdot arctan(B\cdot (\alpha +S_h))-E\cdot (B\cdot (\alpha +S_h)-arctan(B\cdot (\alpha +S_h))))+S_v$

$F_x=D\cdot sin(C\cdot arctan(B\cdot (\kappa +S_h))-E\cdot (B\cdot (\kappa +S_h)-arctan(B\cdot (\kappa +S_h))))+S_v$

where D is the peak factor, C is the shape factor, B is the stiffness factor, E is the curvature factor, Sh is the horizontal shift, Sv is the vertical shift, α is the slip angle, κ is the slip ratio.

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Toolkit to ingest MoTeC logs, align channels, detect events, and generate actionable driver & setup insights. Can also ingest LapSim data and compare them. Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. You thought I wasn't going to put a Lipsum heh ?

CAD was based on an outside resources I can't find again, but honestly the CAD was pretty rough so I repaired and tailored it to usable shape.

ML-based surrogate to predict aerodynamic performance of tubercled airfoils, trained on SU2/Star-CCM+ datasets. Imagine being a Mech/Aero Engineer doing CompSci, couldn't be me

Integrated corner model and aero-map ingestion to produce lap predictions and guide setup choices. For real though, I just fell down the VD rabbit after doing the tyre model and ended up doing that #InspiredByMikeLaw #LinkedIn #Social

All of that because of a blurry video I really wanted to repair that was even repaired by this project... So much sunk hours, it's almost depressing.

Might never do this one to be honest, also a 2021 project, not very interesting above all else, it's just a solver for big matrixes in the end.

Integrated corner model and aero-map ingestion to produce lap predictions and guide setup choices.

Easter eggs NO Hardhat Squirrels YES

The rafale definitely is better than the F-35, say what you will but french engineering can rock when it isn't mutalised by EU.

There's probably a way to create Lipsums in HTML, but I want to finish this show it to my mom and go to sleep.

This whole WIP screen has really come to a full story which will lack sense when I reorganize all the projects.