Correct. Perhaps someone else can do a better job of correcting ShapeGSX misunderstanding of how LSD's work.

He doesn't seem to grasp that by locking the diff. you redistribute torque according to load (ie - traction). On the front wheels this helps fight understeer.

Again, Mitsubishi (Evo), Honda (integra), Subaru (STI), aren't stupid. They don't put LSD's on the front to make the car plow more. They do it to help curb understeer.

 

Haha, thats funny, never heard of that.

 

Yep. If you exceed maximum potential grip of both front tires, then this is an issue of needing more tire, suspension or perhaps take your foot off the throttle a bit. In this situation, an open diff will have plowed you off the road long before you reach this point and certainly isn't going to help in any way.

 

I claim that a LSD front diff could either increase or decrease understeer depending on circumstance, since it's really doing two things. First, it's transfering torque to the outside front tire (an open diff will give equal torque to both tires, neglecting internal friction, contrary to popular belief, because it is mechanically equivalent to a lever). This will cause an oversteer moment because the outside tire is pulling forwards more.

However, at the same time, it's also saturating the outside front tire by trying to use up more of its grip for thrust, reducing it's ability to provide lateral force. (Remember the traction circle) Since the outside front tire has more weight on it and is thus shouldering more of the cornering responsibility, this reduction in front lateral grip is an understeer moment.

So it really depends on which of these factors is stronger what the total effect on the car will be. The effect it definitely WILL have is to let you put more power down coming out of the corner, which is the real point of it. My guess is that in general a front LSD would increase oversteer at light throttle and switch to understeer at heavy throttle. If it also locked on deceleration/coast, it would always increase understeer, because now both factors above will be understeer moments, so such a car might understeer at turn in and oversteer at mid-corner and turn exit. Not to mention what all this locking and unlocking will do to steering feel.

 

Have you been reading the whole thread from the beginning? I thought we all come to an understanding that open diff causes more or all of the torque to transfer to the wheel that has little or no traction hence, the wheel with traction the "outside wheel" will receive little or no torque because there is only 100% of torque to be split between the left and right wheels. If one wheel has 0% the other will have 100%.

if you have no LSD on your car go outside, start up your car and rev the engine up then drop the clutch. Now come out of your car and look behind you to see if you have one or two tire marks. Betty yet have someone else do the clutch drop and look from the outside. You will notice that only one tire is actually spinning the other is just rolling with the car. That one "spinning" tire has 100% of the torque. If you see 2 tire marks behind you then I will stop posting on this thread.

 

Of course one tire will spin with an open diff. The torque is still equally split between the two wheels--just picture how a differential works mechanically.

With hard launch with an open diff, due to any slight asymmetry in the car setup, one tire will reach the limit of adhesion slightly before the other. Under acceleration, the wheel is rotating slightly faster than the road speed due to tire deformation, forming a positive slip ratio. Much like slip angle for lateral friction, the more slip ratio increases up to a point, the more the longitudinal force increases, but at a certain slip ratio the force will start to fall off. Since one tire (let's say the right) has reached this point first, the force will fall off on the right tire first. Since the driving torque is the same to both wheels due to the open diff, but the reaction torque from the tire is less on the right wheel, the right wheel will start to spin faster than the left. As it spins faster, the slip ratio increases even more, and the friction drops off even more, and so it just gets worse and worse. Basically, this property with the open diff causes any tiny difference between the wheels to be quickly magnified into one wheel spinning, one wheel rolling. Of course it's even worse on a car with a live axle due to the torque reaction from the propshaft causing unequal loading on acceleration.

 

I am glad to see this forum can still generate a good thread once in a while.

 

I disagree. You can make an electrical circuit analogy of this. If one wheel is spinning, i.e. has no resistance, then all the energy will go that direction and none will go to the other wheel which is still connected to the road and hence has resistance. The purpose of an LSD is to apply resistance to the spinning wheel, thereby directing more energy to the wheel with traction.

If you are in a hard turn, and the outside wheel loses traction, then there will be little contribution from either front wheels without an LSD of some type. Which means that the car will behave like a rear wheel drive car, in proportion to how much torque actually gets planted on the road from the front tires.
I might add, that applying the brakes would act like a front LSD to some extent and then create more traction from the front.

 

This is getting to be hair-splitting, as we all agree on the ways in which open diffs suck, but here's my last attempt to explain the "equal torque" thing, and then I'll shut up:

When the engine is making torque, the ring gear exerts a force against the centerpoints of the pinion gears, which are allowed to turn freely, enmeshed with the two side gears.
|
v
|-O-|
- here is my lame diagram of the force being applied via the ring gear to the axle of the pinion gear, which acts like a symmetrical lever and applies a symmetrical force to the two side gears, causing a symmetrical torque to the two wheels. Just for clarity, more *power* is being put to the spinning wheel although the *torque* is the same, since
power = torque X rpm / 5252
and the rpm is clearly higher on the spinning wheel.

I agree that you can sometimes use a brake to get some effects of a limited slip (such as the e-brake trick for starting a RWD car with one wheel on ice)

 

Oops, that "diagram" didn't come out right--the downwards arrow was supposed to be over the O in |-O-|

That'll teach me to try to draw a diagram with ASCII...

 

So an open front diff will exhibit "near FR handling" in a corner, just like the topic of the post says.

And that may or may not be ideal.

 

This should help everyone out on how these things work. Great site, answers all my questions whenever something like this comes up.

http://auto.howstuffworks.com/differential.htm

Really in this case the differential isn't the reason for oversteer/understeer characteristics of the car. A great example would be the good old Dodge Neon ACR. FWD with an open front differential... oversteers like an AE86. Or maybe an old Rabbit GTi? It actually oversteers so severely its one of the few cars that can do decreasing radius donuts in parking lots.

The Evo handles the way it does because of suspension setup. Something as simple as wheel alignment can be the biggest factor. Adding a limited slip to the front of an Evo, I believe, would allow it to exhibit more rear wheel drive characteristics while increasing grip. But like I said, alot of the things that makes the USDM Evo such a joy to drive could be hampered by that addition. Plus the same results could be had by introducing more negetive camber in the front alignment to increase front grip on that outside tire.

There are sooooo many factors that determine how a car handles I think the scope of this conversation is missing the big picture. Weight transfer, suspension load, wheel alignment, tire choices... there's so many variables... thats why teams have racing engineers as well as mechanics... its difficult enough that in the WRC there are actually multiple engineers just to study handling dynamics at each event. It gets complicated and depending type of LSD and manufacturer it gets even more complicated.

I think the conclusion could be had that the open diff is not the reason for the mild understeer associated with the US Evo, but its more a function of alignment and suspension setup. Adding a LSD with no other changes will result in less understeer or could possibly introduce mild oversteer depending on type of diff, but you still haven't solved the underlying issues and may not result in faster track times.

So why didn't mitsi go with an LSD and a correct alignment? After watching and building the rumor mill of the US evo for the last two years I think its political. They got a product to the market first and it sold well. Subaru countered with something with much larger numbers, trumping the Evo on paper. Mitsi will return fire with an FQ300 type evo next year with all the goodies but still under the price of the STi. Subaru will then counter by adding the stereo and floor mats in at no charge and by having wierd al do an evo parody song. Mitsi will counter by adding JATO rockets to their evos and showing commercials of naked girls street racing each other in clear body paneled cars to the beat of "Like a Duck" or some other indie dance tune. Subaru will counter by retreiving the plans from the war room, race up the stairs to the communication room and transmit the plans to the allies...

I'm sorry.. I start typing and I get confused.

 

Correct. That is exactly the analogy most often made to explain why/how open diff. leads to runaway wheel spin.

One picky point though, a distinction should be made between power and torque. Open diff the power goes to the wheel that's losing traction. Technically speaking there is also no torque being put down to the road by the wheel, if there's no traction, but still....that is where the power is going.

 

I am giving up now, but here's a quote from the site DistantTea linked above:
"The open differential always applies the same amount of torque to each wheel."

You have to distinguish, though, between the torque applied to the wheel through the driveshaft with the torque reaction caused by the tire contact patch, as well as between torque and power.

I agree with DistantTea that you can never look at one componant in isolation when it comes to car handling--changes to shocks, springs, bump stops, swaybars, alignment, suspension geometry, weight distribution, bushings, diffs, tires, fuel load, etc all interact with each other in a complex way. Adding in AWD with trick differentials just makes it more complex. There are many basic rules of thumb such as stiffening the front springs/shocks/sway will increase understeer, but they are simplifications. Suspension tuning is usually a combination of rules and trial and error, and ideally the suspension has to be tuned to match the driving style of the driver. It's easy to make any car understeer or oversteer by changing any number of things, what's really hard is to make a car that's near-neutral at turn in, mid-corner, and track-out.

If anyone here wants to learn the nitty gritty of whats going on at a physical level, try a fat book called "Race Car Vehicle Dynamics" by Milliken and Milliken. For a more approachable/practical level addressed to racers, not engineers, try Carrol Smith's Tune to Win.

 

Here is a page out of a article that says that the Evo without AYC is better suited to sideways driving.