Torque Bias Ratio Help

[Underthefloor]

Does anyone know how I would go about finding the best torque bias ratio for our car? We are currently running the Torsen University Special and it's not very tunable so that limits testing unless I swap it out with other diffs.

Is there a good reference out there for how torque bias affects performance?

Are there any teams that have run an open diff?

This is my first year doing Formula and I barely knew what a diff was is September. We ended up using the Torsen because that is what was used the last few years. I would like to have a better understanding of TBR so that I can make an informed decision on what diff to use next year.

[Underthefloor]

Does anyone know how I would go about finding the best torque bias ratio for our car? We are currently running the Torsen University Special and it's not very tunable so that limits testing unless I swap it out with other diffs.

Is there a good reference out there for how torque bias affects performance?

Are there any teams that have run an open diff?

This is my first year doing Formula and I barely knew what a diff was is September. We ended up using the Torsen because that is what was used the last few years. I would like to have a better understanding of TBR so that I can make an informed decision on what diff to use next year.

[jdstuff]

John,

Check out http://zhome.com/ZCMnL/tech/Torsen/Torsen.htm link. It's an old paper, but a pretty good description of how the Torsen works.

[Z]

John,

This doesn't answer your question, but it might be something to think about.

From a drivability or vehicle dynamic point of view, it is best to have an open diff during braking/corner entry, and a locked diff (spool) when powering out of the corner. Mid-corner depends on power/speed/grip/etc. - open when coasting around the corner, and locked when using high power.

The simplest "differential" that satisfies these requirements is a "free-wheel" at each wheel - either wheel can rotate faster (forwards) than the final drive sprocket, but not slower. So like the free-wheel on every rear bicycle wheel. For FSAE a friction type one-way-clutch would be better than a bicycle's ratchets.

The disadvantages of this system are;

1. Can't use engine breaking. Not really a problem in FSAE where the brakes are lightly stressed (anybody ever seen FSAE brakes glowing yellow?).

2. Can't use reverse gear (try it on a bicycle). This is the main reason this system isn't used almost everywhere. A "detroit locker" is a bit like two of these (for forward and reverse), but even more complicated because of the mechanism to switch over. However, how many FSAE cars use reverse gear?

Advantages are;

1. Can't hurt ("buzz") engine by putting into low gear at high speed and over-reving. May need faster idle speed if using turbo or prone to stalling.

2. Fairly simple, so possibly also cheap, light, and reliable. (A spool is better here.)

3. Four x free-wheels would give simple, easy 4WD system. In 3 x open-diff 4WDs, if one wheel lets go - no drive from other three! Only option is to start locking up diffs, which gets complicated.

4. Fantastic dynamics! Easy turn-in 'cos no outer-wheel(s) drag. Great power-on mid-corner and exit 'cos acts like a spool. Not clunky mid-corner like a "locker".

There is a "loophole" in FSAE rules - no reverse gear required - that is waiting to be exploited.

Z

[Denny Trimble]

Not trying to shoot this idea down, but has anyone seen it in use in a car?

I've ridden a few trikes with this system... can't coment on understeer/oversteer though http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Z]

Denny,

Again, you certainly won't find it on production cars, because it doesn't work (at all!) in reverse gear. Likewise any racing formula that requires reverse, which is most of them.

In Carroll Smith's "Drive to Win" the Weismann locker is one of his favourites. The Weismann is sort of two friction-free-wheels per wheel (for forwards and backwards) with a "dead-zone" in the middle (sorry, poor description, you'd best check DTW).

Above I said a friction version would be better than ratchets. Ratchets might be easier to make yourself. If the ratchets have too big a gap between them then they get clunky, and balancing the car mid-corner can be harder. A finer gap (<5 degrees?) and they might be ok.

Z

[Underthefloor]

Jason,

That's a great resource for learning about the internals of the Torsen.

On a side note, I think that some of the equations are a little bit off (I could be wrong). I worked through them back in October so I don't remember exactly what was wrong but I think that some of the coefficients of friction were switched around. Anybody else notice this?

Z,

Would the ideal LSD be one which has the maximum torque bias without causing too much understeer?

Why is a spool ideal when powering out of a corner? Does the inside wheel slip while the outside wheel accelerates the car forward?

With an open diff, which wheel breaks free first when accelerating out of a corner?

[James Waltman]

It's not a production car but we have Mechanical Diodes in Viking 23, our BioDiesel/Electric Hybrid car.
http://www.viking23.org/images/pics/wall03sm.jpg
The goal is to improve efficiency by eliminating engine braking (rear drivetrain). The electric motor takes care of reverse (front drivetrain).

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
The disadvantages of this system are;
1. Can't use engine breaking... </div></BLOCKQUOTE>

I can certainly do without anymore engine breaking.

[Z]

John,

Re: "Would the ideal LSD..."

Ideal diff would be a separate CVT to each wheel, with electronic box reading steering-wheel-angle and other sensors, then providing correct amount of thrust to each wheel to steer the car where the driver wants to go. This may mean more Torque and RPM to one wheel, which no normal (passive) diff can do. This would be possible in FSAE with 2 x rubber belt CVTs (like the ones used on agricultural quad-bikes, but 2x), but the development of the controller could take lots of time...

I stick with my original suggestion that the best "torque bias ratios" are 1:1 (=open), or less (=freewheel), at corner entry (=no throttle). And Infinite:1 (=locked) at corner exit (=full throttle).


Re: "Why is spool ideal when powering out..."

During corner entry a spool exerts a forward force on the inside-wheel and backwards force on the outside-wheel - the tighter the corner the bigger the forces. This acts to make the car go straight (instead of turning), so is an understeer (US) influence. This is a BAD thing if the car has any difficulty turning in, which most cars have, to some degree.

This is not so bad at mid-corner because with medium power the inside-wheel provides all the thrust and the outside-wheel rolls with no slip, hence no drag or thrust. Again, this is an US influence, but by now the corner is opening out and US is stabilizing (ie. no sudden tail out) so a GOOD thing.

At corner exit if you power out hard enough then both wheels slip, with the inside-wheelprint slipping backwards, relative to the road, a bit faster than the outside-wheelprint. The outside-wheel is more heavily loaded so which wheel exerts the bigger forwards thrust depends on lateral-load-transfer, slip-ratio curves, etc., etc. But the spool exerts the biggest total forwards thrust compared with any other common diffs, which in racing is a GOOD thing (a very clever 2xCVT might (???) do better). More total forwards thrust = more rear load transfer = more rear grip = less oversteer = more GOOD, since all rear-drive cars can suffer OS when powering hard out of a corner.

A spool might exert an oversteer influence under full power by giving more thrust at the more heavily loaded outer-wheel. This can be a problem if the car has too much understeer at entry and mid-corner, because the front wheels will be turned too sharply into the corner. The solution is to adjust the car for less entry understeer, and get the front wheels pretty straight before fully opening the throttles.


Re: "With an open diff..."

Unless the outer-wheel is on oil/sand, etc. the more lightly loaded inner-wheel will definitely break free first. The amount of total forward thrust available is equal to the Inner-wheel-thrust x (TBR+1). For an open diff TBR is maybe 1.2:1 ('cos of friction in the bevels, etc.), so Total-thrust = Inner-wheel-thrust x 2.2. If using a Torsen with TBR = 4:1, then Total-thrust = Inner-wheel-thrust x 5.

This is ok if the inner-wheel has some reasonable grip. But when the inner-wheel loses most of its grip (say, when lifted off the ground) then Total-thrust -&gt; 0! With a spool or free-wheels (TBR = infinity) the car still has all the thrust capability of the outer-wheel.

Strictly speaking, a Torsen with TBR = 4 should lose all thrust if it lifts its inside-wheel. However, bearing friction and inertial resistance will still give the outer-wheel some thrust capability. I've never used a Tor$en. Can anyone report on its behaviour with one wheel off the ground?



James,

Sexy car!

Yes, as with Viking a 4WD that has free-wheels on its rear "diff" and an open diff at the front would work well.

"Mechanical Diodes" reminds me of the CVT that is a mechanical version of an inverter/transformer/rectifier DC-to-DC power supply. They work great at fractional horsepower but I don't think anyone's been successful at automotive sizes.

Re: "1. Can't use engine breaking... "
Oops! That should be under "Advantages - 1. Can't break engine..."


Z

[rjwoods77]

Being a huge fan of cvt's I will be the proud informant of this very cool bit of race car history.

http://www.ritzsite.demon.nl/DAF/DAF_cars_p17.htm

I always loved the packaging on this thing with the driveshafts being placed between the pulleys.

[threehondas]

Rob you beat me to it with that post of the dual CVT DAF!

We use a camp and pawl diff, torque bias around 2-3, and yes, with a poorly (say stiff rear ARB, and soft front ARB) you can easily lift the inside rear at corner exit and have difficulty getting the power down. Just try a donuts and watch what happens. You'll lift the inside wheel you get no torque on the outside wheel, no more donut.

[Denny Trimble]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
I've never used a Tor$en. Can anyone report on its behaviour with one wheel off the ground?
</div></BLOCKQUOTE>

Torsens work just like open diffs with one wheel off the ground or close to it. One of our worst-handling cars, from '98, had 50:50 weight distribution, but wider tires in the rear (because it's a racecar... http://fsae.com/groupee_common/emoticons/icon_wink.gif ). So, it was set up with a stiff rear ARB and soft front, and with a center rear brake you could see the inside rear freewheeling backwards on corner entry, and spinning on corner exit. Crappy!

That's the car I inheireted when I came along, and the next year we did much better http://fsae.com/groupee_common/emoticons/icon_smile.gif

[osubeaver]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">So like the free-wheel on every rear bicycle wheel. For FSAE a friction type one-way-clutch would be better than a bicycle's ratchets. </div></BLOCKQUOTE>

Are you referring to a "slipper clutch" like those found on newer sportbikes?

[John Bucknell]

Like WWU, I built a mechanical diode rear end in a fit of inspiration around about 1994 for my FSAE car (it had CVT as well). Unfortunately, never got the CDI ignition on the new ZX-6R to work - so c'est la vie. Maybe it woulda worked....

PS I'm in Melbourne right now, visited RMIT today...

[Z]

Quote
---------------------
Are you referring to a "slipper clutch" like those found on newer sportbikes?
---------------------

I guess so, though I've not seen the ones on bikes. They're also known as "Sprague" or "overrun" clutches.

They often have inner and outer races, like roller bearings, except one race has flats ground on it. The spring loaded rollers get wedged between the flat and the other race in one direction, locking the two together, but get pulled out of the wedge in the other direction, for free-wheel.


Z

[BryanH]

Free wheeling overdrive units were fitted to thousands of Rover cars in 60's & 70's, They were also fitted to Jags and MG's.

Rob, Techno is still making high quality karts, in a hi-tech 3 storey factory in his mum's backyard! only in Italy can this happen.

As I grew up 1/2 hour from Phillip Island I'm glad that John Bucknell got the usual experience; Sunburnt & sticky on Sat. Cold,wet & windblown on Sun!
You could go to FPR and con Rotor into giving you a quick guided tour of a V8 Supercar team. www.fpv.com.au/fpr (http://www.fpv.com.au/fpr)
Bryan H

[Garlic]

Like Z said you want a more open diff on corner entry and a more locked diff on corner exit. The Torsen can accomodate a difference but it's difficult to change it.

A differential is very important in the overall car dynamics. The diff you want depe3nds greatly on car setup and vice versa. The best way to properly decide on diff characteristics is to write a 4 wheel simulation. Sounds very complex but you can start simple and add resolution as you are able.

[John Bucknell]

Hmmm, shoulda checked mail earlier - maybe coulda visited Rotor. I might be at the Auckland round after WRC in two weeks http://fsae.com/groupee_common/emoticons/icon_smile.gif If I'm there, I'll look him up.

BTW, the mechanical diodes (brand name) I used were effectively really strong sprague clutches - little spring-loaded struts in one race that grabbed notches in opposite race. Commonly used in helicopter drivelines to allow autorotation - hella strong.

[James Waltman]

It looks like John and I are talking about the same thing. We have a spare one in the shop right now from Epilogics (http://www.epilogics.com/md/index.htm).
http://www.epilogics.com/images/mechd-navi.jpg
Some more information is here. (http://www.epilogics.com/md/tekintro.htm)
They can take a fair amount of torque. (http://www.epilogics.com/md/logp01.htm)

[Underthefloor]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">...write a 4 wheel simulation. </div></BLOCKQUOTE>

How would I go about getting started with this? What do you use to write it?

[Garlic]

Get Milliken or Gillespie and start using the formulas you find there to write a simulation using code of your choice. Start with a bicycle model in a steady state corner, add modules for tire data, engine (longitudinal force), make it four wheels and add weight transfer, then the diff can be involved.

[VFR750R]

Sounds just like a starter clutch on our YZF/FZR. Probably same deal on a CBR/GSX-R ect. They are a neat deal, they are always ingaged; in one direction roller bearing, in the other locked solid. The rollers ride up on their cage and lock.

[Z]

Here's some other things to be aware of regarding different differentials.

1. An open diff always gives approximately equal thrust at each wheel. So if during acceleration one rear wheel hits a low grip surface (oil/sand), then the total thrust drops to about twice that of the low grip wheel, and the total thrust vector remains aligned with the centreline of the car.

If a locked diff (ie. spool, or a free-wheel diff) has one wheel on a low grip surface, then the other wheel still generates full thrust. This moves the total thrust vector sideways, towards the grippy wheel, giving a quite large yawing moment on the car. This is why purpose built drag cars, which always have spools, have a narrow rear track and a long, high yaw inertia, chassis.

An open diff is safer, but you'll never find one on a serious drag car.


2. If you want to use free-wheels in FSAE then the brakes have to be outboard of the free-wheels. So no single central caliper on the final drive sprocket.

Using a central brake with a Torsen tends to lock it up, and thus interfere with turn-in (like a spool). This is, apparently, why some teams with Torsens fit dual rear brakes. Dual brakes can be smaller, so in total weigh about the same as a bigger single brake.


3. Very hard braking, with any type of diff, can lock the rear wheels and stall the engine. This is especially so on low grip surfaces - the road doesn't keep the wheels turning so only light braking can stall the engine. If the car is still rolling, then releasing the brakes will allow the wheels to restart the engine.

This restart won't work on a free-wheel diff. In fact, just locking up one wheel, say the inside-wheel during corner entry, will stall the engine and neither wheel can restart it. So with a free-wheel diff you want either:

3a. A good driver and a well setup chassis. The driver can apply a bit of right foot (assuming left foot braking) and/or declutch at critical moments. Something to try while testing...

3b. A fast enough idle speed to overcome the braking force of an inside-wheel off the ground.

3c. An automatic anti-stall system. The simplest might be a pneumatic clutch release that operates whenever the brakes (or brake light) are on.

3d. A good starter motor.

Z

[Underthefloor]

If you run the free wheel diff, what happens when your engine speed drops below your wheel speed?

When you hit the gas, the engine catches up to the wheels and you go from zero torque to possibly way too much torque, breaking both wheels free before you can react. On the otherhand, since the engine will only engage the the diff when the wheel speed and engine speed sync up, there might be a smooth transition from zero to full torque.(This is just speculation)

It would be interesting to test the free wheel setup but I would definatly have a backup diff in case the free wheel setup is impossible to drive.

[Underthefloor]

Im reviving an old thread here. The following is pure speculation.


Z wrote:

"The amount of total forward thrust available is equal to the Inner-wheel-thrust x (TBR+1). For an open diff TBR is maybe 1.2:1 ('cos of friction in the bevels, etc.), so Total-thrust = Inner-wheel-thrust x 2.2. If using a Torsen with TBR = 4:1, then Total-thrust = Inner-wheel-thrust x 5."

This is only true if the inside wheel is slipping. If the inside wheel is NOT slipping, the maximum thrust for the open diff is Inner-wheel-thrust x (TBR+1)/TBR which comes out to Inner-wheel-thrust x 1.833. For the torsen, the total forward thrust comes out to Inner-wheel-thrust x 1.25.(I used your TBR values, I think that most of the university specials run around 3:1)

An open diff should be able to handle about 46% more forward thrust before the inside wheel starts slipping. Once the inside wheel starts slipping the torsen has a huge advantage but it also makes inside wheel slip much more likely. It would be interesting to compare the thrust for the torsen with the inside wheel slipping to an open diff with both wheels gripping.


Anyone out there ever try an open diff on a FSAE car?

[Denny Trimble]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Underthefloor:

...If the inside wheel is NOT slipping, the maximum thrust for the open diff is Inner-wheel-thrust x (TBR+1)/TBR which comes out to Inner-wheel-thrust x 1.833. For the torsen, the total forward thrust comes out to Inner-wheel-thrust x 1.25.(I used your TBR values, I think that most of the university specials run around 3:1)

An open diff should be able to handle about 46% more forward thrust before the inside wheel starts slipping. Once the inside wheel starts slipping the torsen has a huge advantage but it also makes inside wheel slip much more likely. It would be interesting to compare the thrust for the torsen with the inside wheel slipping to an open diff with both wheels gripping.


Anyone out there ever try an open diff on a FSAE car? </div></BLOCKQUOTE>

Why do you care about the total thrust if you're not slipping the inside wheel? I don't think you'll spin the outside wheel first on corner exit, with whatever diff you run.

I don't think "maximum thrust" and "not slipping a wheel" are simultaneous conditions.

[Underthefloor]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Denny Trimble:

Why do you care about the total thrust if you're not slipping the inside wheel?

I don't think "maximum thrust" and "not slipping a wheel" are simultaneous conditions. </div></BLOCKQUOTE>

I was wondering if you could set up a car with an open diff so that you don't get inside wheel spin around tight turns. When you accelerate out of the turn, you would have two tires gripping. Would this possibly give you better acceleration than if you used a torsen and had one wheel slipping and one gripping?

[caroline]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">...If the inside wheel is NOT slipping, the maximum thrust for the open diff is Inner-wheel-thrust x (TBR+1)/TBR which comes out to Inner-wheel-thrust x 1.833. For the torsen, the total forward thrust comes out to Inner-wheel-thrust x 1.25.(I used your TBR values, I think that most of the university specials run around 3:1)

An open diff should be able to handle about 46% more forward thrust before the inside wheel starts slipping. Once the inside wheel starts slipping the torsen has a huge advantage but it also makes inside wheel slip much more likely. It would be interesting to compare the thrust for the torsen with the inside wheel slipping to an open diff with both wheels gripping. </div></BLOCKQUOTE>

Underthefloor, where did you get these equations?

[mtg]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Underthefloor:
Anyone out there ever try an open diff on a FSAE car? </div></BLOCKQUOTE>

I haven't tried an open diff, but a couple years ago we ran a clutch pack LSD and experimented with very low amounts of preload (which acts very similar to an open diff).

Turn in was fast, but you had to really roll on the throttle super slow coming out of the corner to keep from roasting the inside rear tyre like a marshmallow.

Having an open diff on decel/turn in isn't always the best situation. Some racecars run high TBR on decel (sometimes higher than the accel TBR) to produce a yaw stabilizing effect under straight line braking. That might be handy when you stomp on the "pedal in the middle" going 150mph when there's a 40mph hairpin coming up. Useful for FSAE? Maybe not. Anybody try this before?

You can set up a car to reduce inside rear spinnage with an open diff. Basically, you want to put as much load as possible on the inside rear on corner exit. There are many ways to do this; a couple examples are caster, roll rates, downforce...

The idea behind limited slip diffs is that you want the engine to drive the outside rear wheel faster than the inside on corner exit. Open diffs tend to do the opposite. The best passive mechanical LSD (that I know of) on exit is going to drive both rear wheels at the same speed (ie a spool). To get the ideal condition, you need an active diff that drives the outside faster.

I think F1 cars have little calipers on friction discs in the diff so they can actually brake the inside tire and drive the outside faster- I'm not completely sure though.

[Underthefloor]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by caroline:

Underthefloor, where did you get these equations? </div></BLOCKQUOTE>

I made them up. Regular friction type differentials bias to the "wheel with more grip" which actualy doesn't mean the wheel with more grip but the wheel that is moving slower, or wants to move slower(like the inside wheel when cornering).

mtg,

On our 2005 car, for some reason, we don't seem to be able to spin up the inside tire around turns. With a large amount of TBR, the inside tire should spin up untill it equals the speed of the outside tire.

Without wheel speed sensors its hard to tell what the tires are doing but from looking at the inside wheel around tight turns it seems like we are not breaking the tire loose at all.

[Z]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by mtg:
...To get the ideal condition, you need an active diff that drives the outside faster.

I think F1 cars have little calipers on friction discs in the diff so they can actually brake the inside tire and drive the outside faster- I'm not completely sure though. </div></BLOCKQUOTE>
Definitely forbidden in F1! Along with every new technology that would be useful in passenger cars if someone developed it - eg. regenerative braking, better CVT's, more efficient engines (2-stroke diesels...), etc., etc... F1 diffs can be electronically controlled but must behave like normal passive diffs.

FWIW, I think "active diff" usually refers to a fairly standard diff that can be locked-up (hydraulically/pneumatically/electro-magnetically) via electronic control. A diff that can send more power to the outside wheel - ie. more torque AND more rpm - is often called a "torque-steer differential" (although I am sure there are other names). These are extremely useful, and are common on military tanks, bulldozers, and (not via a diff) the common-as-muck skid-steer-loader that I recently spent a week on (great fun http://fsae.com/groupee_common/emoticons/icon_biggrin.gif, err... except when I crashed into the house http://fsae.com/groupee_common/emoticons/icon_frown.gif).

I believe Honda briefly had a torque-steer diff on a front-drive production car, about 10-20 years ago?

Z

[Denny Trimble]

Z, there was an article in racecar engineering several years ago that mentioned exactly what you just said about the torque-steering differentials (not the part about crashing into houses), and I believe it was the Honda Prelude. Cool stuff.

[mtg]

Apparently I was just plain wrong on the F1 diffs.

It seems odd that a car manufacturer would use the term "torque steer differential". I think most people would say "torque steer sucks, why would you pay more for it?" The idea is awesome, though.

Underthefloor:
You're not spinning the inside rear most likely because:
- The high TBR (very low TBR spins inside rears)
- Not enough power (probably not the case)
- Driver being a "throttle wuss", or stomping the throttle after the car is pointed straight instead of rolling in on starting around the apex.

There's probably a 90% chance its the high TBR that keeps the inside rear in check.

[Schumi_Jr]

Z,

Why don't you mention salisbury differentials? They provide different TBR on corner entry/middle/exit and are tunable. They also have a smoother actuation than a torsen which instantly loads up.

I know this diff isn't as "ideal" as your slipper clutches but they can be purchased off-the-shelf.

On an un-related note, do you know where I can buy some "smart springs" http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Z]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Schumi_Jr:
Z,

Why don't you mention salisbury differentials? </div></BLOCKQUOTE>
Schumi,

Yes, salisbury diff's are probably the best standard-type LSD's for FSAE in terms of their performance. However, I'm not sure of their availibility in FSAE (small) sizes - they may be a bit heavier and more expensive that the Torsen or other options (eg. ATV cam-and-pawls).

I didn't mention them because I assumed everyone is aware of them http://fsae.com/groupee_common/emoticons/icon_smile.gif - they are possibly the most common type of LSD. I mentioned the free-wheel-clutch type "diff" because it is relatively unknown, mainly because it won't work in reverse gear. No reverse gear required in FSAE, so an opportunity!

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">On an un-related note, do you know where I can buy some "smart springs" http://fsae.com/groupee_common/emoticons/icon_wink.gif </div></BLOCKQUOTE>
Yeah, steel store. Buy some high tensile steel and bend your own. http://fsae.com/groupee_common/emoticons/icon_smile.gif


mtg, I guess a smart manufacturer would call their "torque-steer" diff an "Activa-Posi-Drive-TSD", or somesuch...

Z

PS. An active torque-steer diff is feasible in FSAE, though perhaps best done as a parallel program.

Start with a spool, then insert a hydraulically activated multi-plate clutch between the final drive sprocket and each wheel. The simplest control strategy would release (disengage) both clutches when the throttle is closed - corner entry, coasting, etc. This would imply dual rear brakes outboard of diff. Then lock (engage) both clutches whenever the throttle is more than ~10% open, so acting like a spool for accelerating, mid-corner and exit, etc.

Then add more sensors - steering wheel angle, wheel speeds, X&Y G sensors front and rear, etc... - and more sophisticated control algorithms. Eg. drive the right rear wheel harder (release left wheel) whenever steering wheel turned left.

IMO this sort of thing would be more useful than either 4-wheel-drive or 4-wheel-steer, both of which have been tried quite often in FSAE. Especially useful on a car with significant rear weight bias (eg. 30F/70R). Not only would it be great for performance, but it might get you some Design points too!

[Marc Jaxa-Rozen]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Start with a spool, then insert a hydraulically activated multi-plate clutch between the final drive sprocket and each wheel. The simplest control strategy would release (disengage) both clutches when the throttle is closed - corner entry, coasting, etc. This would imply dual rear brakes outboard of diff. Then lock (engage) both clutches whenever the throttle is more than ~10% open, so acting like a spool for accelerating, mid-corner and exit, etc.

Then add more sensors - steering wheel angle, wheel speeds, X&Y G sensors front and rear, etc... - and more sophisticated control algorithms. Eg. drive the right rear wheel harder (release left wheel) whenever steering wheel turned left. </div></BLOCKQUOTE>

Yeah, that's basically how the ATTS system on Preludes works.

I was thinking of something like that when looking at the effect of TBR through a corner, but it's probably the kind of project at which the average team would need to throw half their budget and workpower. The Torsen likely isn't ideal, but we're currently not at a point where it'd be a limiting factor anyway.

Marc Jaxa-Rozen
École Polytechnique de Montréal