RR98ITR

Have you read the latest? The 2013 Renault Clio Cup car is "a big improvement"!

Here:

"Give it more throttle, and start winding the lock off"

"It's unbelieveable. The diff is a lot tighter and it drags you thru the corner
on the throttle. It works for you."

And this from someone presumed to not leave alot on the table: 2009 series champ
Phil Glew. Now I've read this sort of thing before. Many times before. I know it wearily as
magic diff syndrome. I think getting traction down is one of those areas of vehicle dynamics that isn't
widely understood correctly. I know - hard to believe that not everybody glances at a
cutaway pic of Torsen/Quaife and instantly perceives the obviousity of it's function.

You know how an open diff works right? And especially how it works when we're unhappy with how
it works?

The case is driven by the attached ring gear, and the case drives the pinion, and the pinion
gears drive the side-gears which are splined to the axles. And the torque split is 50/50 (and all
of the torque is applied if there's no wheel spin). And the casing doesn't care if one wheel wants
to spin a little faster or slower than the other because it's on the inside or outside goin round
a corner.

We racers get unhappy with the open diff when one wheel is unloaded enough that it offers little
resistance and the power follows the path of least resistance and spins that wheel. There's still a 50/50
split of the torque APPLIED - generally an unsatisfactorily small amount determined by the
coefficient of dynamic traction between the spinning wheel and the track surface. It can be helpful
to picture the side-gear on the laden side as stationary, and the pinion gears and hence
pinion and case spinning up around it.

The Cammed Clutch Pack Limited Slip Differential is functionally an open diff (ie transmitting power
from the pinion gears to the side-gears) with an added clutch mechanism to bind the laden side-gear
to the casing under design operating conditions. The pressure rings are driven directly by the case, and
the clutch clamping force is dependent on the resistance from the side-gears.

In the case of the open diff - we can put thru the laden wheel only as much power as we can put thru the
unladen wheel. That value is less when the inside wheel is spinning.

In the case of the clutch pack - we can put thru the laden wheel a multiple of what we can put thru the
unladen wheel (once it starts wanting to spin).

Do you see a problem? Let's say you've got a front wheel drive race car, and you've got a lot of grip,
and you've got almost no weight on the inside front wheel. What's the difference between one times zero
and three or four times zero?

Unload the front of a front wheel drive car enough, and the reward is Understeer. Doesn't
sound terribly controversial does it? But what I'm saying is that if you make a front wheel drive car
go around corners too fast, no conventional limited slip diff is gonna have enough magic to fix it.
Take those Clio's for example - check the specs - not alot of tire. They're multiplying a larger number.
And the driver reports it feels good. Squares with my experience - my car pushed with 225's on the front
and was a dream with 205's. Not that it didn't go faster on 225's - it did. But the faster I made it go,
the worse it drove.

Now for the rest - what if there was a mechanical diff that worked differently? That didn't work on
a multiple of the unladen wheels capacity. That apportioned power to the laden wheel. Thinking of the
Torsen? Wrong. Not wrong like "Wrong, that's not the diff I'm thinking of", but "Wrong, they don't
actually work like that". I'm talking about the Weismann Locker. Go re-read the 4 Smith books wherein he
discusses this diff, and study the original patent. Clever, simple, a few things to be wary of, some
unknowns. Note also that there is/was surely one across the front axle on the K-Pax Volvos.

What do you think? And is the Weismann Locker The Answer? The question? Fine - 300+ HP NA K24 Powered CRX
with Moby Dick front fenders and lots of sticky tire for eating Porsche GT3 Cup Cars in the dry, or
something like that.

Scott, who is Not working on such a thing...but is actually working on the same old thing for a change...

Pic of typical clutch pack diff:

rice_classic

Patent:
http://www.freepatentsonline.com/3283611.pdf

Closest image of it I can find on the internet (yes, Weismann's own website doesn't have any images of its guts) was this: The Kaiser Locker.

It's the closest thing I could find that resembles what I see in the PDF.

rice_classic

Egezzy

iTrader: (2)

Cool thread. Food for thought. It seems as if on a "club type" level, a "traditional" LSD is a "known" must have.

Xian

You may be closer to the Real Question than you realize... looks like the gearbox/diff is made by SADEV with diff ramp ratios of 23/90. It also appears that SADEV makes a tranny designed to work with K-series engines. Unfortunately, you'd have to make-do with 23/85 ramp ratios and that obviously just wouldn't work. Guess it's back to the drawing board...

ST75-14 Gearbox


Bellhousing adapter:

RR98ITR

Hi Christian,

You don't necessarily have to go That far. Quaife's kit for converting a standard K-series case into a 5-speed sequential is probably a more practical option. And there's no dead end diff-wise. Weismann can build a one-off to fit any reasonable dimensional constraints - they need no more room than a bunch of invex gears or clutch mechanism.

Scott, who thinks it would however defeat the purpose if such a project approached the cost of a Porsche...

Xian

Nah... a dancing elephant is, at the absolute least, a curiosity. Much the way Porsche has continued to tweak, revise, rework, reengineer, their beloved 911 into something that not only dances but dances well. There's no reason to think that the Hype aaaarrrRggghh can't shirk it's factory provided B-series albatross and attain greatness with little K-Power and Quaife engineering.

Think of it as The Singer Approach applied to Honda's. Go make it so, Super-Genius.

Christian, who thinks... sometimes too much.

Black R

I'd buy that for a dollar!

Now make kits and sell them.

How does the wavetrac design not live up?

For such a low price, it's better than nothing.

914Racer

The Xtrac VCP (Viscouse Clutch Pack) used back in the Supertouring days was probably the best design for a FWD.


http://link.springer.com/chapter/10....-642-33744-4_8

I can't find a good article on it online. I have a couple in older issues of Race Car Engineering.

miamirice

it's times like this I feel like a dumb ****.

RR98ITR

Me too.

Scott, who still can't quite see how the Wavetrac's cam/clutch is actuated - I gave up the other night as my brain was expressing a strong preference for unconciousness..."but I've got it all figured out now"...I think...I think it would work even better when the inside front lifts if you had really heavy wheels..."unconciousness - Now!"...

914Racer

The problem with a locker is the same as a spool. The abrupt engagement will show up in the steering and braking. I have met a couple of folks who ran a welded diff in some Production and GT Cars. They claim it was good once moving, but terrible everywhere else. I drove a rally car with a welded diff around my shop parking lot and thought I would break a wrist, so not so sure its a good idea on asphalt.

This is the OS Giken diff:


This is the Wavetrac, it is a Torsen with a small spring plate in the middle of the "spider gears".
wavetrac.net/technical.htm

It basically allows it to operate when one wheel is unloaded. It seems to be a good change to the gear type limited slip.

Xian

And that's basically how Eckerich's diffs work... and they work well but likely not as well as Texa$ $tuff like the Giken. Still, I've wondered how much positive/negative impact the Eckerich diffs make on turn in as they're essentially a 2 way with static lockup rates.

914Racer

I am going to try one of Steve's diffs. He and I talked on the phone one day about it. It should work well in my CG gearbox. Since there aren't a lot of options out there for the first gen Integra gearbox. I suppose it could be measured with a torque wrench. but I don't know how accurate that would be.

Since we are feeding power, steering and braking through the front wheels, we have to be careful how harsh the transitions are between those operations.

Xian

You can definitely measure breakaway torque of Steve's diff's... I did thtat just after install and then about once a season afterwards. There was close to zero drop off even after several seasons.

I really, really like his diff's from a bang for the buck standpoint and would take one over a Quaife every time. I remember the first time I ran it after install... Was in the short chute between T6 and T8 at Roebling. Hit the gas on my outlap and almost had the wheel jerked out of my hands. LOL. Definitely would buy again.

powerneedy

Did I miss something? Who is Steve? lol

rice_classic

Steve Eckerich's LSD's I believe are best known as OPM LSD's.

Xian

Yup ^^^. OPM was the primary (only?) seller of them for ages. Steve makes a ton of different race/aftermarket items as a side business and then typically sells them to the retailers. Everything from diffs to hubs to who-the-hell-knows.

The way his work is by exerting pressure onto the spider gears. You end up with something akin to a 2 way clutch diff in that it works under decal, accel, and with a wheel in the air but without "ramps" since the pressure is determined by the spring rate used on the installed diff. Bonus points are that it doesn't require any funky fluid choices, doesn't wear quickly, and is lightweight.

Robin6

My brain is hurting...lol

Great thread!

Xian

LOL... ok. Imagine if you took a "Phantom Grip" LSD and used higher rate springs. And then you machined the inner faces of the spider gears to provide additional contact area. And then imagine if the outside of the spring "blocks" had a wet clutch material on it's facing. The result is a 2-way diff that's small, light, and very effective.

If I'm remembering my history correctly, PG "stole" his design but missed the critical details that made it... well, work. :-P

edit:
Here's a pic:




PS
I know RJ had a helluva time with his working "right"... that was traced back (years after the fact) to running synthetic tranny fluid which reduced the amount of friction the clutch material could provide. Run conventional tranny fluids and you're fine.

ReDemoN

I like the wavetrac one.
Partly because I dont have OS or OPM money. Lol
Gotta be a hell of alot better than this OBX trash I have

Xian

d00d. OPM/Eckerich Diff's are cheap... seriously cheap. No idea how well it would work for autoX. In my head, I'd think that you'd be better with a 1 way or 1.5 way given the nature of autoX vs. track.

ReDemoN

Yeah that also.
I do know I could tell a difference in my car and when I co drove a H2B swapped teg that had a type r LSD. His car felt so much different in and out of the turns.
So I wouldnt mind a better LSD than the OBX

RR98ITR

I say...these differentials are givin me a headache...

I've been looking at and thinking about the Wavetrac diff, and about what diff's do in general, and how they're marketed, and how they're discussed...I mentioned...yeah...headache...

And I dug out all my Diff clippings from the archives of Racecarengineering and Racetech (they were so much better years ago - now they're like the typical trade magazine)...and read them again...and...headache...

What do we hate? Spinning the inside wheel whilst turning. Why? Because if not for that we Know we could be putting down some more power to the outside wheel and not falling even further behind schedule.

How much power? I guess it depends. It can't be huge amounts otherwise how'd the inside wheel get unloaded and what's the outside wheel doing with all that grip? It Should have about all it can handle with lateral load. So we can usually get happy enough with a torque bias ratio of somewhere between 3 and 5 to 1 - especially if we only unload and spin and NOT lift the inside wheel.

If we lift the inside wheel, maybe say crashing a curb, and we keep our foot in it, then just about every diff out there is going to spin it up while putting very little down thru the outside, and when the inside comes back down it's going to give us a push off center. It really isn't that bad - that's why drivers have brains and should cultivate an ability to adapt if only out of a sense of self preservation.

What are we really talking about? And how do these things actually work? I can really see why so much smoke and hand waving. Few understand, and fewer yet can explain (Craig Taylor does one of the better jobs). And it doesn't matter anyway because only sales matter and almost all of the comparable products perform comparably for all but the most demanding or demented users.

Almost every one of these things - Including the Torque Biasing ones - splits the input torque 50/50 till the start of a wheel spinning. Then and only then with the inside wheel spinning a little bit, and Because it's spinning a little bit and wedging an internal mechanism, we get to put 3 to 5 times the thrust that spinning wheel is providing down thru the outside wheel. And they refer to this as "Applied" torque. That 80/20 split you read about - that's not the percentage of what your foot might ask for. It's the split of what you can put down, Dictated by that barely spinning inside wheel and your particular mechanism. If you've got a few hundred more horsepower to throw at it than it can handle, and you throw it, you ain't going any faster, though you may get points for style. The more throttle you apply, and the faster you spin that inside wheel, the smaller the pie you're sending 80% of to the outside wheel. You just can't take the stupid driver out of these equations.

So we're talking about subtle stuff, happening within a short sector of any particular corner, for a driver who wants what he wants, or wants to try it anyway. How would you know if you could put any more down if you couldn't try. And if you couldn't it should be pretty obvious...in those microseconds before your brain saves the day or lap anyway.

Maybe not as obvious as to all those guys driving 8 tenths who throw alot of throttle at a still pretty highly loaded inside front and rave how it pulls them thru the corner like they just found two more inches (of tire width but what's the difference).

So about the Wavetrac...looks pretty clever. That cam/clutch mechanism has to be set up to not engage on mere differentiation, but then lock pretty hard on wheelspin (lock pretty hard 1 because it's a cam clutch and 2 because it's a really small clutch for the loads it'd see in the lower gears). It really might be an answer worth trying to the question at hand. I looked back thru old posts and found where KIWI said that they'd worked well and met his expectations.

Scott, who thinks his headache is clearing...now I can get back to work...ah, hmm, this looks like a headache...

ReDemoN

Scott, I love reading your posts, very informative and then funny at the ends, well most of the time.
Lol