CFD - PERRINN F1 Car (w/ spec mods)

[Vanja #66]

Please open another thread for your simulations.

[pwlucas]

First off, nice work! As a user of Fluent daily, which solver did you use? I have found that the coupled is slower per time step [obviously] but is actually quicker to convergence. I switched to that about a year ago and found my process has sped up a decent amount and I can run more cases a week.

[Col_Antonio]

3.8GHz Intel i7 quad-core with 16Gb RAM and a 1060 NVidia 3Gb GPU... Unfortunately hyper-threading doesn't really benefit CFD and so I was kinda restricted on running to only 4 cores. The GPU maxed out its VRAM whilst I was running the script to produce all the images since they were in 4k, and theres over 100 produced each run... I mean, what I have is no "potato", but compared to the beast I have at work, it leaves quite a lot to be desired........

As for the CFD app on your phone? Well... There is already one that was produced by Numeca (a meshing, CFD and FEA software company): https://www.algorizk.com/ (the bottom one is the one I have). It isn't "strict" CFD, but it is 2D simplified and solved in real time if i remember correctly.... It allows you to draw shapes with your meat hooks and let it run... You can vary Re and a few other parameters or load up some of their pre-drawn shapes like a motorbike or something. Its more a sort of thing I use when people ask what I do and just showing them a picture or two like above doesn't quite cut it.

I got a 16 thread Ryzen CPU in my rig and I'd love to help render some stuff if you need. Send me a DM and I'll gladly set it up to help with simulations.

[Shakeman]

Does Fluent use the GPU for calculations or are the datasets too big?

As a 3D artist I'm used to massive performance increase for fire, smoke and fluid sims when using GPUs.

Just interested not wanting to derail the thread.

[pwlucas]

Does Fluent use the GPU for calculations or are the datasets too big?

As a 3D artist I'm used to massive performance increase for fire, smoke and fluid sims when using GPUs.

Just interested not wanting to derail the thread.

It can, but doesn't by default. It doesn't help on all analysis as it depends on the physics and solver used. Fluent mainly solves with CPU cores and it is licensed that way. When you purchase a year license, you are allotted a number of cores (based on what you want/need). I think it normally comes with 4 cores for the license. I am unsure on that as we purchase more than that.

[slimjim8201]

Vyssion, do you have a link to the latest Perrinn model. Mine is lacking a few updated aero bits and the halo.

[Vyssion]

Vyssion, do you have a link to the latest Perrinn model. Mine is lacking a few updated aero bits and the halo.

There is no Perrinn with a Halo and other updates -- they are all custom modifications and created assemblies of myself and jjn9128.

[Vyssion]

Okie doke.... another update!!

This week, I didn't get much opportunity to do more than generate a few more post-processing scripts for image generation. One taking slices through the front half of the vehicle tracking FW, FWheels and Halo wakes, as well as another one which took some slices through the car horizontally. Some of these are created sort of "in advance" of doing some other simulations in the future. However, there is also another image which is quite interesting...

It is of the Front Left Wheel wake region and depicts a velocity deficit contour as well as surface streamlines on the planes showing the path of the flow. It is interesting because, in general, this region can be used in order to see how well the Y250, Bargeboard and Turning Vane vortices are managing the airflow between the wheels and chassis, as well how much downwash & outwash is being created by the design of the aerodynamic devices in the local vicinity.

From my understanding, it is generally a better indication of good aerodynamic design if the wake at the base of the wheel is pushed outboard more so than the central upper wake. At the bottom of the tyre, it is important to keep the low energy flow in the wake region away from the underfloor area which prevents disruption to the floor performance. Decreasing the velocity deficit of the wheel as a whole will also help reduce the drag force on the wheel itself. As such, the wake profile kind of ends up being "twisted" as the outwash management increases in effectiveness. There is still quite a bit more improvement to this twisting which could be done when comparing my contour to ones in literature, but nonetheless, I think it will give you an idea of what I am referring to.

Anyways, I have just completed a run with a side-slip "gust" wind of about 4° to the freestream (the limit before which we felt a change in the steering angle would be required to keep it somewhat realistic), and so images of those will follow soon.

(I should probably look into creating a .gif of the slices soon... I'll have a play around and see what I can do with it)

[Dipesh1995]

Awesome work Vyssion and jjn9128.

Would be even more awesome if you could do some “true” cornering simulations using curved flow at some point. I think it would be really insightful regarding the car’s aerodynamic profile differences between straight-line and cornering cases.

[Vyssion]

Just a quick one for now...

Played around with converting the 4k images to a... not-4k .gif..... They are of Total Pressure Coefficient travelling on a YZ plane sliced through the domain travelling from negative to positive X.

The top .gif is of a 0° angle case (i.e. straight forward) @ 50m/s, whilst the bottom .gif is of a 4° angle side-slip "gust" case. More to come on the 4° case soon!!

[Jolle]

Just a quick one for now...

Played around with converting the 4k images to a... not-4k .gif..... They are of Total Pressure Coefficient travelling on a YZ plane sliced through the domain travelling from negative to positive X.

The top .gif is of a 0° angle case (i.e. straight forward) @ 50m/s, whilst the bottom .gif is of a 4° angle side-slip "gust" case. More to come on the 4° case soon!!

https://imgur.com/oZJVIRh.gif
https://imgur.com/NrF5pvv.gif

With this model you can see how important the insides of the FW are, those tips with the "gap" where the vortex lives along the car is quite large and works all the way until the back of the car.

[PhillipM]

You can also see why the overtaking working group advocated for bigger rear wings to loft the wake and pull cleaner air in faster behind the car.

[jjn9128]

While technically not the Perrin I thought I'd share some unsteady CFD of the halo model we added to the car. This simulation was done a while ago so I can't recall all the specifics, but it was completed in Openfoam using the Pisofoam solver with meshing performed in snappyhexmesh for around 10 million cells. I haven't got y+ or anything else noted, but I seem to recall it took about a week to simulate 0.3s of time, what's shown is only the last 0.1s of that simulation.

The vortex shedding from the halo was ~200Hz at 50m/s.

[Vyssion]

Ok,

As mentioned earlier, I ran a case which looked at a very slight yawning condition. Technically, it isn't strictly "yaw", as it has only an approximate 4°of "sidewind" applied to the domain, so I guess you could say that it is more of a "gust" whilst driving forward... But regardless, the angle was sort of picked based on some quick searching through mine and jjn's archive of papers as being one which would kind of be at the limit of what could exist, without the need for a change in steering angle. Beyond this, I would look to delete the suspension on the model, and re-create a parameterized version of it such that I could vary ride height, camber, toe and steering angle as required... buuuuuuuuuut maybe I'll hold off on doing that, as I am sure that my OCD would suck me in to re-creating a lot of the baseline surfaces given how uhhhh "not-good" they are...

I digress... From now on, I think it best if, rather than drowning you all in images, I create a few "standard GIFs" to show, and then perhaps add images which kind of "highlight" any flow features of note etc. If there are any ideas that you guys have for such images, please do let me know; currently, my post-processing rendering script kicks out about 170x 4k images, but hey... you can never have enough CFD porn... uhhhh, I mean, "very carefully chosen images to showcase very specific features of the flow"...

In the slices, you are able to track the vortices as they move around the car, and you can see (for example) the wake of the front left tyre being blown over inboard and then hitting the rear wing (and looking at the CP images, you can see the green region from the top view on the left of the main element implying a loss in performance relative to the rest of the orange wing section). Also, I have included a switching GIF of the wheel wake w/ streamlines so you can see how the velocity deficit region changes its "twist" as I mentioned in the last post due to the slide-slip velocity. Finally, I have included some high res still images of CP which showcase the asymmetry of the flow field now due to the side-slip from top, bottom, front and rear.

Enjoy!

(note: for reference, the sideslip wind is travelling from negative "y" to positive "y"... i.e. in the direction that the green arrow in the coordinate frame is pointing)

[Shakeman]

Do you have any ability to export these data in a format that 3D animation software could import i.e. the iso surfaces as an Alembic cache or the wake as particle data with vertex attributes or even as volumetric OpenVDBs?

Is there any Free CFD software that works on WIn10 by any chance? It would be nice to be able to create these data for myself and maybe use it creatively.

Cheers.