Race Talk with MV Agusta - How to Improve Moto GP Aerodynamics
For more information, visit https://www.airshaper.com
MV Agusta is one of the teams competing in the MOTO GP - moto2 race class. In this interview, technical director Brian Gillen reveals how they blend physical & virtual wind tunnel testing to improve the aerodynamics of their moto2 race bike.
Wind tunnel testing process
Wind tunnel testing is great: you get a lot of empirical data. But there are a number of steps involved before you get there:
- Test definition: set up the goals & methods of the test plan, which sensors to use, which loads to expect, ...
- CAD design: design the parts for the prototype in a 3D CAD package
- Component prototyping: manufacture the components using 3D printing, carbon fiber, CNC machining, ...
- Mounting on a test bike: mount all the components on a frame (which can be the frame of an existing bike)
- Transportation: move all the required people, equipment and the prototype to the wind tunnel.
- Wind tunnel test: execute the wind tunnel test, monitor the forces, log the data, apply different visualization techniques, set different angles of attack, test at multiple speeds, apply tufts, ...
- Post-processing: process the data into a comprehensive report and draw conclusions
Wind tunnel testing advantages and disadvantages
+ Accurate force & moment data
+ Live feedback
- Indirect visualization
- Challenging to find causality
Wind tunnel simulations
- Traditional CFD programs: the time required to prepare a CAD model (close all gaps & holes) and manually set up a simulation is too long compared to the rapid steps of evolution of the bike in the development process. By the time simulation results were in, they were already 3 steps past with the real prototype.
- AirShaper: take a non-closed model and very quickly put it through the CFD algorithms to get back data. This allowed for very fast iterations to understand where they were in terms of fluid dynamics around the bike.
For them, as engineers, data means nothing if it is not validated.
They were able to validate the AirShaper data versus data from the wind tunnel test and found an almost complete convergence in terms of forces & moments on the vehicle.
That allowed them to build a virtual avatar around the bike as a mirror image of the physical world. From there on, they could dive into the virtual world for an iterative process before going back to the real world, where testing costs a lot of money. Designing, constructing, testing & scrapping prototypes when they do not work is very expensive.
On the track, the drivers can test the real performance of the bike and give qualitative & quantitative feedback to the development team (vibrations at certain speeds, wind buffeting affecting visibility, aerodynamic stability, ...).
Here too, like in the wind tunnel, you need to be able to correlate numbers to sensation. For this to be possible, the rider has to be very lucid & clear in his description of what he's experiencing on the track (what & where is he experiencing an issue) so that this can be matched to data to find a mathematical reason.