Wavetrac Differential

[Tyler Jones]

Tyler,

It seems you are using the standard "because racecar" approach to solving your problem. Namely, chase the most complicated and expensive solution you can find. (And if you think spending yet another ~$600, AND making a new diff-case, is NOT expensive, then please send the money to me! )

As noted by others above, there are many simpler and cheaper fixes to your problem. Please re-read CWA's posts, and also get your suspension guys to read them several times. There are many clues there that can help your Team build a faster car.

Regarding your list of options/goals for selecting a diff, I think you are somewhat missing the point. It is important to realise that you need the "whole package" (= diff + suspension + everything-else) to work well, and that means you cannot just choose a diff based on a simplified list of "diff characteristics" taken in isolation.

Because cost is so important (to me!) I would immediately narrow my diff choice down to these two cheapest options.

1. A "spool" diff.
This is about the simplest machined part you can make for the car (= lathe + drill-press?). (Or you can turn your current Torsen into a spool by squeezing some extra washers in there, or filling with epoxy, or weld...) Your current car seems to be already "set-up" for a spool, so no extra work there. And, most importantly, there are countless cars, both past and present, that have won FSAE comps with a spool. The "spool causes understeer" is one of those lame arguments that might be true of a spool in general, but has turned out to be irrelevant with the car set-up used on many winning FSAE cars.

2. An "open" diff.
Your Torsen is close to this, though cheaper ones are available from any car wrecking yard. The set-up changes required on your car, to suit a fully open-diff, are quite easy. From reading your above posts I suspect your suspension guys are, ... ummm, a little lost. [Well done Z! The politeness lessons are working.] Short list: Fit stiffer front springs, and/or softer rears. Or a VERY stiff front-ARB. Make sure frame is torsionally stiff enough for this to work!!! Make sure front dampers reach droop limit before rears. Reduce excessive "steer-wedging" from steering geometry. Beware of VERY short end-view Rear-Virtual-Swing-Arm-Lengths. And some others ... but the above should cover it. Personally, this is my preference because I like the idea of having all four wheels always planted.

The choice is yours, of course, but throwing money around does not equal good engineering. Besides, I am sure you will eventually meet a girl who will teach you how to really squander your hard-earned dollars!

Z

(PS. "A good engineer is someone who can do with one dollar, what any fool can do with ten!")

Z,

The Wavetrac is $600, compared to $400 for a Torsen (which we gut and use a custom housing with anyways), or ~$300 for an ATV front differential. In my mind the price difference is almost negligible over our current Torsen since we would be spending $400 out of that $600 either way. I personally programmed and ran the CNC turning to make the custom Torsen housing using donated 7075, so other than time invested the housing was free. Yes, time is a very valuable resource for FSAE but in most cases I would rather invest my time over straight cash in the car.

I do agree that we should try locking the Torsen and testing with it as a spool, but I don't know of any methods to achieve that which aren't destructive for the internals (its not cheap to destroy the gears in pursuit of something which may turn out not to be beneficial). To manufacture our own spool would require some method of hobbing or wire EDMing the correct spline for our stub shafts, which we don't have the resources for, meaning it is definitely not cheap. We could buy some stubshafts that have a bolt pattern instead of a spline, but the only ones I can find are $185.00 per stubshaft. Perhaps we could manufacture our own but that's more time invested in something the car isn't tuned for at the moment.
I would also like to see a list of recent results for cars using spools as I haven't seen a single car in the top ten at any North American competition using spools to my knowledge.

My suspension guys are the ones that initially suggested switching to a locking differential because they are afraid that standard options for tuning the chassis to keep that wheel down would cause the car to understeer significantly. They have already tried most configurations for arbs, including stiff arb front and no arb rear. The car currently has stiffer springs front than rear, though there are options to try different springrates there. I can't answer for droop limits, steer-wedging, etc but I will have my suspension guys look into it.

Don't get us wrong Z, we are a very fiscally conservative team. Most of our design is centered around keeping the car cheap (usage of a lot of laser cut steel, aircraft fabric body, lots of 3D printed components). We took 2nd in cost at Formula North with a very accurate cost report (I know this doesn't exactly equate to real world cost but it at least correlates to a cheap car).

More quotes from my suspension guys:

"I think our actual weight dist. was 49%, and one of the aero guys said the cg raises about 1.5 in with wings (i'd like to measure that to see for sure). From what I've been learning, I'm pretty suspicious of our roll centers and how they work with the rest of the car, and also how the travel available is mostly for compression. Also we haven't explored 100% of the bar options, there are some left which could likely fix the problem by adding a lot of roll stiffness, but the car would probably understeer a ton (since we'd be adding most of that to the front)." -Suspension guy 1

"Static weight distribution: 45% front actual
CG height: 10.2 measured without wings, 10.5" with wings (estimated, This is what I think it was, can probably get a better estimate later)
Roll Center: 3.0" front 3.8" rear
Chassis Torsional stiffness: Not sure off my head. (Somewhere around 1800lbs/degree simulated)
Wheel Rate: 20n/mm seems low, maybe a conversion error? Check the tech sheet we turned in for lincoln, that will give you a closer estimate. Springs are 225s.
Ignore the ARBs and do not consider them into wheel rates.
Car weight is 405lbs?

Steering Geometry: Look at the design tech sheet. Those are all "designed" numbers,
Static Front Camber: -2*
Wheels: 6in width

The suspension does not have bushings and does not carry much friction at all. Essentially is negligible for most calculations we do. (Not always negligible, but its low enough that I'm ignoring it for now and focusing elsewhere)

We have scales and the car is on them frequently for testing. I don't have a data sheet with me to record actual corner weights.

I think some of the focus shifted on to figuring out the suspension issue. Which while I appreciate it, is not exactly why we need a different differential. I do not think it is a "band-aid fix". While I see the point that the suspension is causing an issue, I don't see how a differential that will handicap the car in odd situations is acceptable. A different differential isn't the solution to this problem, its a solution to a few different problems. One of which, is the diff is currently untuneable.

I have a few hunches to our wheel lifting issues. One of which is an overall roll gradient vs wheel travel issue. The car was allowed to roll quite a bit in the last cars, and I think it rolls more than calculated. Which is easy to prove in this car, but I haven't had a chance to yet." - Suspension guy 2

I guess I will have to track down the rest of those numbers myself when I find the time.

Thanks for the advice, keep it coming!