Drexler 2010 Differential Always Fully Locked?

[Jonathan Stewart]

Hi, I'm Jonathan from the University of North Carolina Charlotte. I was responsible for the chassis design and manufacture this year, and now, given the size of our team I mostly just do whatever is needed. I am also one of the drivers.

We've been testing our car in preparation for Michigan 2015 for about a month now, but only recently got our data logging wired and working. What stood out to us the most was that both of the rear wheels are always the exact same speed. We could not physically do anything with the car to get any difference at all in rear wheel speeds. After the first test with data logging I opened up the differential to see if everything was okay, and also changed it from the default setting of 1 to setting 2, since it is the most open for both power and coast. In our case, the differential is flipped from the image in the manual so we have more locking under coast than power no matter which setting we choose. After cleaning and assembling the differential with setting 2 chosen, it is behaving exactly the same. There did not appear to be much wear on the clutches in the differential from previous cars running it, so it may have always been working this way.

As a result of basically having a solid axle, we ended up with a balanced skidpad setup that has a very stiff rear in roll, effectively giving us with a kart. We did not expect to have the differential behave this way and it's not what we were planning to test with, or use in our Dymola model. The chassis was designed to be simple and light rather than stiff, but this could now become a concern because we are far from having a 50% distribution of roll stiffness front to rear.

Is it possible that there is somehow a huge amount of preload in the differential, which we never overcome? Is it possible that since we are unloading the inside rear tire so much already that we never even provide enough of a torque difference to unlock the differential? We did try changing to softer rear springs (which created understeer), but nothing extreme.

Has anyone else had this issue with this differential?

[Drew Price]

It could definitely be a pre-load issue, especially if you can see no wear on the clutches. With the car jacked up off the ground, how hard is it to turn one rear wheel and hold the other stationary? You should be able to tell pretty quickly if they are actually locked together.

Does it always spin the inside rear around corners?

Make sure there isn't something stupid going on, like one inner driver is too long and is engaging both the left and right side output gears simultaneously or something like that. You never know.

[Lux5018]

Hi,

If you think there is too much preload in the differential you could measure it (Lock one wheel and apply torque to another, put the engine in neutral there shouldn't be any torque acting on the sprocket). Normally Drexler differential has a low amount of preload, i think preload isn't a problem if you haven't changed anything in the differential.
With our Drexler diff you can turn the inner ramp assembly around so you have more lock on power than on coast.

[Loz]

These Drexler diff's have 6 possible ramp configurations for each orientation of the diff. It is possible to achieve all 6 configurations any each orientation (sprocket left or right).

If there doesn't appear to be wear on the clutches, then it means there is no relative movement between the clutch plates (or very little) and the diff has possibly not been operating as differential at all. (i.e. it is always locked).

There is no preload to "overcome". The preload is an additive load and applies forces in the same direction as the forces which lock-up the diff. Preload only locks the diff up more.

The diff also doesn't "unlock". On the contrary, it locks-up under load (i.e. it should operate as an open diff under low load and then progressively locks-up under increasing load).

It sounds like there is too much static (preload) and the diff is permanently locked.

It might be an idea to do some revision of differentials, it would seem the understanding of how a diff works is quite a bit off.

[mech5496]

What Loz said. We also had some difficulties with Drexler at the past, but it was due a not-so-carefully done modification on the housing we did ourselves. Drexler diffs have two large Bellevile (conical) springs/washers that provide preload. Those rest on some thin shims on each side of the diff and compress the clutch discs. Try removing one of those shims and see if it fixes your problem, then try thinner shims accordingly. Have you modified the diff by any means? Oh, BTW, you indeed can flip all the insides of the diff so to have more "lock" on power than on coast.

[Bemo]

To me it sounds like the diff is just blocked mechanically and has always been. That would explain that there is close to no wear of the clutches. As there is no relative movement there is no wear. We had that problem once and it turned out someone from our team disassembled and reassembled it. While doing so, he messed it up. I don't remember what exactly the mistake he made was.

[TiJei]

Hi Jonathan,

I faced a similar issue with one of our diffs this year.

In our case it was an assembly error because of an missleading drawing in the documentation we got from Drexler.
We installed the preload spring on the wrong side leading to an offset of about 1 mm between the clutch package center and the center of the gears. I found this by checking the CAD-assembly of the diff.

The behaviour then is just like you describe - a fully blocked axle.

With a functional diff in stock condition you should be able to aply the required torque to overcome the preload by hand when turning the tires on a jacked car. Maybe you need two people if you have 10" wheels.

Regards

[Paul Achard]

Hello,

A friendly PSA to those running the Drexler for the first time.
We had the same problem mentioned by Jonathan where the diff had an incredible amount of preload; one guy was barely capable of diffing one of the rear wheels.
This was causing the car to be nearly undriveable (hard bounce/skip on corner exit) so we wanted to investigate the problem.
It turns out that the bolts fastening the stubshafts were slightly too long and were clamping the cross-bars which hold the planetary gears (see Picture: http://imgur.com/iiU90QE).

After fixing this problem, we find that we can now easily differentiate the wheel by hand (which makes sense considering the nominal preload of 40 N-m).

Hope this helps someone!