Dynamic Simulation of Front Wing

[Diff-user]

Hello everyone!!
I am doing my master's thesis on front wing aerodynamics (CFD). The current literature largely deals with analysis of wing as the cars speeds down a straight. The premise of my thesis is that downforce is needed the most while the car negotiates a turn, and thus, when the wing is in roll, yaw and pitch. I have done the static simulations (i.e. steady state simulations). However my prof. believes that the true nature of the variation of the downforce generated by the front wing can only be captured by a dynamic simulation. So I have to get hold of reliable data on the time history of accelerations (both lateral and longitudinal) and the roll and pitch angles as a race car negotiates a turn. The yaw, i assume, will come from the steering geometry. Any advice/help/data my fellow forum members?

By the way, I shall share whatever numbers i get from such a simulation, in case anyone is interested.

[Lycoming]

It's not terribly difficult to code a simple steady state or quasi-steady state lap time simulator. The basic premise is you break the track down into segments of equal length, calculate the vertical loads on the tire, then use some sort of tire model to get force from that. From there, you can work out acceleration and speed along the track with a level of accuracy that may or may not be adequate for your purposes. If those work for you, then with some basic assumptions about ride and roll rates, you have your yaw/pitch data. Such a simulator is essentially just doing forward Euler's method. Probably the hardest part of this is the tire model and track layout... but that's what google maps is for. If you're gonna do it that way, try a track with little elevation change.

I think that for such a simulator as I have described above, once you have corner speed and radius, then yaw follows from them, as opposed to from steering geometry. More interestingly, how do you plan to model the aerodynamic effect of yaw? Are you just going to model it the same way as you would model a sideslip angle, by having the air flowing at an angle over the car?

[Diff-user]

In a fully dynamic simulation, as we do it in our lab, the whole domain has stationary fluid and it is the wing and its mesh that move through it. The yaw can be calculated from the fact that, ideally, the centre of rotation of the car lies on the line joining the rear wheel centres and the instantaneous radius of turning will be very close to the radius of curvature of the racing line. Thus the yaw, as seen by the front wing, can be calculated.

And thanks for your help. I'll make a model based on what you said and keep you updated on my progress.

edit: And i don't have a car model. It is the front wing alone. I am interested particularly in the flow features around the front wing alone. The details of a complete car, even if I do manage to get some, will only clutter the results with flow physics to complicated to analyse or understand and too sensitive to the geometry used.

[Jersey Tom]

If it were me... as a starting point I'd cut out the yaw and roll stuff to begin with, and look at pure vertical dynamics. As in, have a steady state condition as your baseline - then run some dynamic experiments with varying ride input (1, 2, 5, 10 mm?) and ride frequency (1, 2, 5, 10 Hz?) and see what it shows. Could be an interesting result.

Things definitely move around a bit...

[Diff-user]

That has already been done actually. Very interesting results from there
Here is the paper:
http://arc.aiaa.org/doi/pdf/10.2514/6.2001-863

[Jersey Tom]

That's indeed interesting.

Getting back to your original point, I suppose there are two ways of going about it. One would be to try and figure out what the roll, pitch, yaw, etc. vs time would be for a car. You may be able to get some crude estimates of speeds and accels vs. time if you pick one corner of one track, try and figure out corner radius etc. from a map, then take a look at some race footage where they have the overlay of speed or G meter or whatever. Still, to get the pitch and roll magnitude and rate.. you'll have to make some pretty big assumptions on inertias, suspension stiffness, etc. Incidentally, yaw (or really, body slip) will be almost entirely a function of the rear tires (and any elastokinematic rear steer) and not so much a function of steering geometry.

That's also assuming a flat road. Maybe that at least gets you into some order of magnitude estimate for rates. Maybe your highest dynamic rates happen from bouncing on kerbs. Point is that can be an ugly rabbit hole to jump into.

Alternatively, maybe you could do a study from the other direction - pick some arbitrary roll, pitch, and yaw rates and do a sensitivity study. At what point or rate are the dynamics trivial vs. non-trivial? Is there any particular mode of those 3 which is more sensitive than the others? From a practical standpoint, which of those are easiest to control from a car setup perspective and which are pretty well out of the engineers' hands?