37 replies to this topic

[McGuire]

Something I forgot and then suddenly remembered for some reason: Earlier someone here was having some trouble getting a Salisbury-type (clutch plate) limited slip differential to do whatever it was he wanted it to do. (Brian Glover, was that you?) In my travels I have had some dealings with these gadgets and can pass along what I have learned.

First, the manufacturers of the things typically produce a variety of clutch packs and preload springs for different applications which use the same differential carrier. I would first investigate what parts are available for adapting your particular axle to your needs. No need hunting, adapting or fabricating clutch plates, shims and springs if you don't have to.

Next, when "tuning" these LSDs (which is not a recognized procedure but can be done, within limits) here is the secret : the friction properties of both the clutch plates and the axle lubricant are extremely sensitive to temperature. Your changes will follow no pattern and make no sense and you will get lost, unless you constantly track them against the axle's actual working temperature.

So while you make your changes, you need to record the axle lube temperature, both "cold" and at critical working temperature (digital shop thermometer, with fluid probe.) You will also need to obtain or fabricate an adapter which bolts on your wheel lug centers with a means to attach a torque wrench at the exact center of the hub (3/4" nut, 1/2" female drive adapter, whatever.) So you will also need a torque wrench, 0-250 ft-lbs, preferably a dial-type as it is more precise. A beam-type wrench will work, but will be hard to read with precision. (Click-type will not work.)

Before you do anything, record the axle lube temperature at cold rest. Then jack up one drive wheel, and record first the breakaway torque required to turn the wheel, and then the torque required to maintain constant motion. Then drop the car, jack it up on the other side and do the same. Then drive the car to critical operating temperature and repeat the entire process. Now you have your baseline.

Obviously, the more torque friction on each individual drive wheel (some folks call this the "preload;" I have no idea why) the more you are approaching a locked rear axle, making the car push on corner entry, as well chuckle in slow corners and parking lot manuevers, etc.. The less torque friction, the less drive off the corner with both wheels on power down, naturally.

Every time you make any change in clutch pack, clearance or spring preload, follow this procedure, both before and after. When you get the diff to do what you want, immediately STOP and carefully take your torque and temperature measurements. Then you will know what you need to duplicate to get the exact results you seek.

[clSD139]

here you go!




attention: this is MY car!

[desmo]

Originally posted by clSD139
here you go!

http://rijtechniek.b...ages/purple.JPG


attention: this is MY car!

[clSD139]

sorry, try again!

http://rijtechniek.b...ages/purple.JPG

[BRIAN GLOVER]

That rotten swine Desmo stopped the Magnus Effect thread so I am unable to sock it to you, Bill.
Consider yourself fortunate, you communist you ;) .
The hurricanes chased me out of state and was only able to lurk a few times.
I don't know how the manufacturers set up their differentials such as Corvette and BMW. It is a huge compromise and can only be effectively set up for one corner only. I chose the BMW M3 differential and I am happy with the result. I don't have to bother my brain with all that stuff. Thanks all the same. One of the members passed on some classified information to me about this matter and unless he contacts you, I'm afraid I can't share it with the members. Its too complicated for me to get into anyway.
The new Ferrari 420 has an electronic differential as does the P996 and 7.
I know that if you drive a BMW or a Vette really hard, you should change the lubricant every race. It makes a huge difference. It would be interesting to change the grades and plot performance curves for each circuit. I had some really interesting experiences with the breakdown of the lubricants in one specific turn which prompted me to delve into this. It is the most complicated part of the car. The electronics is mind boggling.
Consult your Ferrari F1 Peter Wright book to give you some idea. No driver can give any feedback either because of the non linear nature of the inputs.
I can assure you that next to tires, the differential is the most important item in race car set up.

Originally posted by McGuire
Something I forgot and then suddenly remembered for some reason: Earlier someone here was having some trouble getting a Salisbury-type (clutch plate) limited slip differential to do whatever it was he wanted it to do. (Brian Glover, was that you?) In my travels I have had some dealings with these gadgets and can pass along what I have learned.

First, the manufacturers of the things typically produce a variety of clutch packs and preload springs for different applications which use the same differential carrier. I would first investigate what parts are available for adapting your particular axle to your needs. No need hunting, adapting or fabricating clutch plates, shims and springs if you don't have to.

Next, when "tuning" these LSDs (which is not a recognized procedure but can be done, within limits) here is the secret : the friction properties of both the clutch plates and the axle lubricant are extremely sensitive to temperature. Your changes will follow no pattern and make no sense and you will get lost, unless you constantly track them against the axle's actual working temperature.

So while you make your changes, you need to record the axle lube temperature, both "cold" and at critical working temperature (digital shop thermometer, with fluid probe.) You will also need to obtain or fabricate an adapter which bolts on your wheel lug centers with a means to attach a torque wrench at the exact center of the hub (3/4" nut, 1/2" female drive adapter, whatever.) So you will also need a torque wrench, 0-250 ft-lbs, preferably a dial-type as it is more precise. A beam-type wrench will work, but will be hard to read with precision. (Click-type will not work.)

Before you do anything, record the axle lube temperature at cold rest. Then jack up one drive wheel, and record first the breakaway torque required to turn the wheel, and then the torque required to maintain constant motion. Then drop the car, jack it up on the other side and do the same. Then drive the car to critical operating temperature and repeat the entire process. Now you have your baseline.

Obviously, the more torque friction on each individual drive wheel (some folks call this the "preload;" I have no idea why) the more you are approaching a locked rear axle, making the car push on corner entry, as well chuckle in slow corners and parking lot manuevers, etc.. The less torque friction, the less drive off the corner with both wheels on power down, naturally.

Every time you make any change in clutch pack, clearance or spring preload, follow this procedure, both before and after. When you get the diff to do what you want, immediately STOP and carefully take your torque and temperature measurements. Then you will know what you need to duplicate to get the exact results you seek.

[McGuire]

Originally posted by BRIAN GLOVER
That rotten swine Desmo stopped the Magnus Effect thread so I am unable to sock it to you, Bill.
Consider yourself fortunate, you communist you ;) .
The hurricanes chased me out of state and was only able to lurk a few times.
I don't know how the manufacturers set up their differentials such as Corvette and BMW. It is a huge compromise and can only be effectively set up for one corner only. I chose the BMW M3 differential and I am happy with the result. I don't have to bother my brain with all that stuff. Thanks all the same. One of the members passed on some classified information to me about this matter and unless he contacts you, I'm afraid I can't share it with the members. Its too complicated for me to get into anyway.
The new Ferrari 420 has an electronic differential as does the P996 and 7.
I know that if you drive a BMW or a Vette really hard, you should change the lubricant every race. It makes a huge difference. It would be interesting to change the grades and plot performance curves for each circuit. I had some really interesting experiences with the breakdown of the lubricants in one specific turn which prompted me to delve into this. It is the most complicated part of the car. The electronics is mind boggling.
Consult your Ferrari F1 Peter Wright book to give you some idea. No driver can give any feedback either because of the non linear nature of the inputs.
I can assure you that next to tires, the differential is the most important item in race car set up.

That was one of our communist hurricanes which chased you out. Some Soviet weather control technology left over from the cold war....we were running some tests to be ready for November. We are not losing Florida again, I can tell you that.

I agree, the differential may be the most overlooked and underutilized aspect of production race car development...and especially frustrating when one is stuck with the Salisbury. What you say about optimization for only corner is exactly right. You try to tune the thing for every corner, and end up with something that is not ideal anywhere. But even then it was amazing what could be done with one when tweaked properly. Electronic LSD is the way of the future. I'm surprised there aren't more out there already.

[BRIAN GLOVER]

I think I like the tightening turn one at Shanghai. It may present some interesting tactics in the future. I wonder what is happening in the differential during that turn? The diff is not always set up for max forward traction but optimizes or adjust rear slip also. I have always wondered why Alonso yanks on the steering and understeers thru low downforce turns. The diff seems to optimize forward traction. causing more push for the fastest way round. Turn one presents a new challenge as he has to slow down. Could it be that understeer is less costly than oversteer in certain circumstances?. For instance, kodandaram made some interesting observations about driving techniques on another thread which has caused me to pay more attention to the actual driving techniques. Shanghai seems to offer multiple lines thru some of the turns and it would have been nice to see Raikenen actually catch up and try to pass Barachello and use the unusual track design to his advantage.. I really am happy to see Villenueve back in F1 because of his scientific approach to his driving. There where radio transmissions from him asking his engineers where he could go faster. They told him that he could go faster in turn one. He started trying different lines and gained half a second. He has asked engineers to create scenarios for passing offline in the past. I think if he has a fast and good handling car, he would be able to exploit the new track better than anybody else.
Alanso had no understeer into the turn and even after it tightened up, the car was neutral, but the Renault understeered like hell accelerating out of turn 2 . Nobody else does this. Is this the way he prefers the differential to be set up or is this what the engineers have decided? If the marker set up the diff for more torque to the outside wheel, it would have reduced the understeer, but the acceleration may have been less. What do you think? I understand the Renault TC was developed for the wide V engine which was fragile and ignition shut off was kept to a minimum. Maybe the TC prefers a more even split of torque to each wheel or maybe it is simply the best way to set it up. The Renaults sure accelerate faster than the other cars at launch and out of slow turns. His team mate doesn't understeer though. Beats me.

[McGuire]

Originally posted by BRIAN GLOVER
What do you think?

I think the whole issue of limited slip operation will tend to be totally interwoven with individual driving style. Power-down point and corner exit speed are so dependent on corner entry. A late braker will not collect near as much benefit in improving power-down off the corner as one who rolls in on entry and hard-powers out.

Power-down is also very sensitive to the rest of the setup. When you were describing the problem with your sports car a few months ago, my mental wheels were turning in the direction of hmmm, maybe I would decrease the rear roll resistance or raise the rear RC just a touch and see what that did.

What beats me to this day: drivers who truly mastered locked rear axles, namely Mark Donohue. I saw them them do it but I still don't really understand it. I get the do-able-ness of it, but I can't believe it would ever be faster than a halfway proper diff. But they proved it was.

[Freebird]

BRIAN:

have always wondered why Alonso yanks on the steering and understeers thru low downforce turns.

I've noted and posted on this on a past thd. Later I started hearing about how Renault is reputed to be running more static weight to the back then the other teams as a way of gaining the launch advantage that we see on the starts. Apart from the diff performance, I wonder how much of a part this (rear weight bias) plays in the low speed understeer? As the speed goes up the aero comes on, the understeer tends to go away?

msw

[Fat Boy]

Originally posted by McGuire


I think the whole issue of limited slip operation will tend to be totally interwoven with individual driving style. Power-down point and corner exit speed are so dependent on corner entry. A late braker will not collect near as much benefit in improving power-down off the corner as one who rolls in on entry and hard-powers out.

Power-down is also very sensitive to the rest of the setup. When you were describing the problem with your sports car a few months ago, my mental wheels were turning in the direction of hmmm, maybe I would decrease the rear roll resistance or raise the rear RC just a touch and see what that did.

What beats me to this day: drivers who truly mastered locked rear axles, namely Mark Donohue. I saw them them do it but I still don't really understand it. I get the do-able-ness of it, but I can't believe it would ever be faster than a halfway proper diff. But they proved it was.

Well, you said it yourself, you should understand it!

It's corner entry. A spool is damn stable on corner entry. That means that the driver can be on the brakes late, but _much_ more importantly, off the brakes early. When you carry a bunch of speed into a corner, most cars will tend to oversteer. A car with a locked diff will oversteer less than one with a limited slip and much less than one with an open.

This is something that many, many people don't understand. F-1 engineers are some of the worst at getting this point. Oversteer at corner entry makes understeer in the center of the corner. The understeer in the center of the corner makes for oversteer at exit. Fix the entry, and the rest of the corner will fall into line. Furthermore, when you do get on the gas with a locked rear, you better mean it, because it's either down the track right now or hello, wall!

The F-1 guys blather on about, "The driver doesn't have 'confidence' to go deeper in the corner". It's not that the driver isn't confident. It's that the car simply won't do it. The car will get sideways, he'll correct, be off-line a bit across the center which creates understeer and then the mid-corner understeer makes for a snap loose at exit. The driver then complains about understeer, which they go about trying to fix. They make the car oversteer worse on corner entry when trying to fix the understeer which makes the car further off-line and understeering across the center of the corner. Repeat until throughly confused........and then you'll have a Jordan. Read Gary Anderson's interview in Racetech a couple months ago. If it wasn't so bloody obvious, it would be comedy.

Just to point out that even the best have trouble with this problem and to re-inforce my point, try this link.

http://www.farzadsf1...97_ms_brawn.mp3

OK, done with my rant!

[BRIAN GLOVER]

Very interesting,
There is a simple formula to determine how many clutches are used(for torque) and what ramp angles (for pressure)is required for torque transfer.

[(48/tangent (ramp angle)+17]*no of plate surfaces/max. number of working surfaces=%slip

The above formula gives a good approximation of the locking force as a percentage of a diff that has been set up with a full complement of working plate surfaces, and a set of 30 degree ramps.
It can be seen from the formula that 17 percent of the locking action is not provided by the ramps. This small component is due to the reaction forces of the side bevel gears.
The formula is simple, but that's where it ends. Roll centers, camber gain, roll, tow, caster, etc. can make your job difficult. This is a black art and the diff settings is the poison in the mix.
In the case of the Porsche 996/7 and the Ferrari 430, this mechanical mechanism won't give desired results. The Porsche locks the diff 100% on shutdown and is set for push with power. I don't know what percentage slip that would be, maybe 40%. The European Porsche have less push and none on the GT2 and 3. I doubt that they mess with anti-roll settings between the two. Lawyers have input into design here. In fact our lawyers are so smart, they run everything.
If the Porsche system fails, why, you are back to a 993.
The F430 is a lot more complicated and the slip is determined by throttle position, lateral force, traction, torque, shaft speed and steering input. There are also stability control settings that can be set on the steering wheel which effect the diff settings. Very simple compared to F1 systems.
F1 car diffs are not active and are set for each corner and segments of that corner. It is the most complicated part of the car to set up, and much of the testing and practice is focused on the diff. This and TC settings determine final performance and driver preferences counts little. According to Wright's F1 book, the driver is at a loss to give any feedback.
Ferrari 550,575 and 430 have 3 settings for TC. 1 is for wet and 3 is for dry. I would assume that the diff settings will change with each TC change on the 430. What system they use or how it works, I have no idea. electro-hydraulic?
A Corvette has one setting for TC and stability control and is optimized for very wet conditions and the Good-Year F1 rain tire tread. It slows you up in the dry considerably. Lawyers? The diff setting seems perfect. You can hit the gas, lift off, coast or brake hard in any radius turn without going off line with no altering of steering input. How do they do it? Just what do they teach at law school these days?
The 550 weighs a 1000 lb more than the Vette and is a pig. Strictly a cruiser. My boss sold me his 550 real cheap, like a parting gift almost.
Anybody know what a new M3 diff % slip is? It is difficult to get any specs from manufactures. According to the above formula and 35% ramp it is 53%.

Originally posted by Fat Boy


Well, you said it yourself, you should understand it!

It's corner entry. A spool is damn stable on corner entry. That means that the driver can be on the brakes late, but _much_ more importantly, off the brakes early. When you carry a bunch of speed into a corner, most cars will tend to oversteer. A car with a locked diff will oversteer less than one with a limited slip and much less than one with an open.

This is something that many, many people don't understand. F-1 engineers are some of the worst at getting this point. Oversteer at corner entry makes understeer in the center of the corner. The understeer in the center of the corner makes for oversteer at exit. Fix the entry, and the rest of the corner will fall into line. Furthermore, when you do get on the gas with a locked rear, you better mean it, because it's either down the track right now or hello, wall!

The F-1 guys blather on about, "The driver doesn't have 'confidence' to go deeper in the corner". It's not that the driver isn't confident. It's that the car simply won't do it. The car will get sideways, he'll correct, be off-line a bit across the center which creates understeer and then the mid-corner understeer makes for a snap loose at exit. The driver then complains about understeer, which they go about trying to fix. They make the car oversteer worse on corner entry when trying to fix the understeer which makes the car further off-line and understeering across the center of the corner. Repeat until throughly confused........and then you'll have a Jordan. Read Gary Anderson's interview in Racetech a couple months ago. If it wasn't so bloody obvious, it would be comedy.

Just to point out that even the best have trouble with this problem and to re-inforce my point, try this link.

http://www.farzadsf1...97_ms_brawn.mp3

OK, done with my rant!

[Fat Boy]

By the way, this is the best name for a thread in a long time.

[Ross Stonefeld]

About Alonsowheel, wasnt Barrichello doing this the last few years in the Ferrari? Though it seems less pronounced. I spoke to Peter Collins about what I had been observing of Alonso's technique through Ascari in testing at Monza, and he mentioned he thought it was the car; however my recollection of Trulli's driving is much more 'normal'. I was really lookign forward to JV vs Alonso because JV always ran that almost bobsled like steering in the BAR, and Alonso runs enough steering lock to parallel park; and wondered how they'd compromise.

Still, nothing beats the Schumacher-float.

[Freebird]

Don't know about Barrichello, but I did compare the steering technique of Alonso and Trulli earlier this summer. I believe it was Canada during practice. For every 3 "hard over" throws of the wheel that Alonso did, Truilli maybe did one. Much more flowing with steering input. Same track, same session.

Didn't get to see much of JV this weekend. What I did see, he seemed to be driving to the limit of the front end and not thru it. He looked to be loosing time in sector 1 to Webber late in the race.

msw

[Ross Stonefeld]

iirc the Rubens wheel movement was more visible in the wet. I think we had (maybe even I started) a thread on it.

[McGuire]

Originally posted by Fat Boy
A spool is damn stable on corner entry.

I suppose that's one way to describe it.

[mini1275]

What about the torsen diff? Some say that they function as an open diff during braking, so that the car is very unstable during braking and it is impossible to trail-brake(No "yaw-damping"). Is this true, you´ve got torque sent in to the diff by engine breaking?
I´m not sure but i believe that the Radical cars use Quaife-diffs and they seem to work very well for them?

[Fat Boy]

If we are talking about a Quaiffe torque sensing diff, then, yes, it does act a lot like an open differential when not loaded. As soon as there is engine braking the Quaiffe does tighten up a little bit and will provide a small amount of 'lock'. I've spent some time on the inside of a Quaiffe, and there is a stack of belleville washers in there that can be set to vary the preload on the whole thing. You can change the bias ratio by using different internal gears, but I don't think anyone actually does.

One thing that I don't like about the Quaiffe is that as soon as you drop a tire off or get the inside in the air a bit (say, over a curb) then all lock goes away and it acts just like an open. As long as you aren't having any problems, it acts nicely, but as soon as you throw difficult situations into the mix, it tends to give up. Some of the good parts about it is that a Quaiffe takes very little maintanance and since you can't adjust it much, it's hard to screw it up!

I don't know if I would go so far as to say that an open differential is very unstable on corner entry. I think if the car is very unstable, then you probably have a series of issues to deal with. Having a differential with adjustable levels of lock that is independent from drive to coast allows for a lot of tuning. Having the ability to calm transient behavior without messing up roll couple can come in very handy.

It doesn't suprise me that a one make series uses the Quaiffe and is successful with it. After all, what are you racing against? Other Quaiffe's. There's no competitive advantage, so it seems to work well.

[wegmann]

Originally posted by Fat Boy
A spool is damn stable on corner entry.

Just from a theoretical standpoint, I don't think a spool would be appropriate in all circumstances for corner entry. If there's any possibility of locking your rears before the fronts, then both rears are going to lock at the same time. That, of course, could be a problem.

I also don't think it would be appropriate in oval track applications where you might be doing a lot of braking and turning at the same time - it would probably produce quite a bit of understeer.

[Locai]

I saw "Experimenting with LSD", and my first thought was that this was a thread concerning the proposed rules changes and possible drug use by Max and the FIA.

...imagine my disappointment.

[mini1275]

Originally posted by Fat Boy
If we are talking about a Quaiffe torque sensing diff, then, yes, it does act a lot like an open differential when not loaded. As soon as there is engine braking the Quaiffe does tighten up a little bit and will provide a small amount of 'lock'. I've spent some time on the inside of a Quaiffe, and there is a stack of belleville washers in there that can be set to vary the preload on the whole thing. You can change the bias ratio by using different internal gears, but I don't think anyone actually does.

One thing that I don't like about the Quaiffe is that as soon as you drop a tire off or get the inside in the air a bit (say, over a curb) then all lock goes away and it acts just like an open. As long as you aren't having any problems, it acts nicely, but as soon as you throw difficult situations into the mix, it tends to give up. Some of the good parts about it is that a Quaiffe takes very little maintanance and since you can't adjust it much, it's hard to screw it up!

I don't know if I would go so far as to say that an open differential is very unstable on corner entry. I think if the car is very unstable, then you probably have a series of issues to deal with. Having a differential with adjustable levels of lock that is independent from drive to coast allows for a lot of tuning. Having the ability to calm transient behavior without messing up roll couple can come in very handy.

It doesn't suprise me that a one make series uses the Quaiffe and is successful with it. After all, what are you racing against? Other Quaiffe's. There's no competitive advantage, so it seems to work well.

I´m not racing a quaife, yet It´s just that i´ve read about salisbury diffs where you want to have some locking in "coast mode" for braking stability and to be able to trail-brake (between 65-85?). And i´ve heard that people say that torsens are good in fwd cars where it´s really just helping that the diff functions as an open one during braking, but that it´s bad i rwd application since it doesn´t have the locking effect.
And what about the yaw-moment that is created in a turn when you brake and more torque is sent to the outer wheel, if the inner wheel has less traction due to weight transfer. Wouldn´t this cause understeer!? i´m really confused and i don´t know much about vehicle dynamics..

[Fat Boy]

Mini,

First, there is no magic bullet. A limited slip diff is a tool, whether it be a salisbury or Quaiffe. If the problems that you're having with your particular car means that more lock on coast helps it, then it can be a big help. If you don't need more lock on coast, then it will hurt you. A diff can be a relatively effective tool to use because it works very early in the transient portion of the corner and it doesn't effect roll couple. Just don't go looking for that magic bullet on me, there isn't one.

A differential that has some limited slip ability will create understeer when compared to an open differential. You've created an understeer by helping the rear of the car rather than hurting the front. It's not a bad deal. It's very rare that you get a perfect car, and I'll take a little understeer over a little oversteer any day.

I don't know what your particular application is for, but since you say that you don't know much about vehicle dynamics. I'd say that it's much more important that your spend time on that rather than money on a differential.

Wegmann,

I'm not a real proponent of running a spool. It just came up in the conversation, so I commented on it.

While you're right that locking 1 tire means locking both, I think that you'll find that on most cars if you have enough rear bias that you're actually locking the rear tires before the fronts, then you are _way_ off on your bias. A lot of times, you can have too much rear bias and still have the front actually locking first. Let's say that the front and rears lock at the same time if the front brake pressure % is 60%. It could be that the driver runs the front brake at 62% because that's what works best on the track. In theory, you could be at 61%, be locking the fronts first, but still have too much rear bias. Am I making any sense?

You are also right in the statement that it could create entry understeer. In the gospel according to me, I feel that nearly every racecar needs _more_ entry understeer than it has. Interestingly enough, my experience is that the handling balance at maximum cornering is hardly effected at all by having the wheels hooked together in anything but the sharpest of corners. The understeer created is a function of the speed delta across the axle, and in 95% of all racetrack corners the speeds are fast enough that it isn't a big deal. Also, if the increased stability on corner entry has allowed the driver to be more aggressive early in the corner, then it is likely that mid-corner understeer will actually be reduced. Most people don't believe me, but try it, you might be suprized. Ovals almost always have large enough corner radius that a spool won't create excess understeer.

Just my take, your mileage may vary.

[mini1275]

Ok, just one more question

I´ve seen that some people brakes through the differential, how will this affect corner entry on a torque-biasing diff? It feels like you get lots of understeer if you transfer that amount of torque. At least compared to when you´re just biasing the engine torque.
And there´s also a big difference between engine brake torque and brake torque so i guess that the diff can´t handle that amount of torque, if it´s not a really, really light car.


Here´s a funny project where a guy also brakes on both front and rear diff:

http://dpcars.randomresearch.com/dp1/


Thank you for your replies Fat Boy!

[Fat Boy]

It's just an opinion, but I'm not a real fan of the old single brake on the diff housing trick. Under ideal conditions, it works fine. The problem is that we rarely run under ideal conditions.

Under a combination of braking and turning when the inside rear is light, using a single brake has the unfortunate tendency of spinning the inside rear tire backwards. Locking a tire under braking is bad, spinning is backwards under braking is worse. I've only driven one car that had the tendency to do it a lot. It's hard to explain, but it felt really odd. You'd feel the braking go away in the rear, like you'd locked up, and then as you came off the brake you'd feel the thing jerk and spool back up to proper speed.

There are a couple of other things about that particular approach that I don't like. The first is that it puts brake heat into the diff housing. This is rarely a good thing. A Quaiffe makes enough heat on it's own, no need to put a 1200F brake disk on it.

The second thing is that it put large load reversals through the drive shafts. These things live a hard enough life as it is, we don't need to make it any worse for them! All in all, I think that you can probably make individual brake discs weigh the same as the single big one, have a better all around braking system, and have a more robust drivetrain to boot.

So that's my vote.

Don't be so worried about inducing understeer. Traditional tuning methods at the track can take care of that.

I think the guy with that homebuilt car is a bit glossy-eyed concerning his brake system. Imagine the fun of having both front and rear inside tires spinning backwards on corner entry!

[wegmann]

I appreciate your insight, Fat Boy, and you are making sense. However, I didn't understand this:

Originally posted by Fat Boy
Under a combination of braking and turning when the inside rear is light, using a single brake has the unfortunate tendency of spinning the inside rear tire backwards.

I don't know anything about the single brake on the diff housing trick, but I'm having trouble figuring out how a brake could make anything turn backwards. ?

[Fat Boy]

The Quaiffe and the Zexel (used to be Gleason) operate in the same basic manner. They basically use worm gears to make an open differential. Since worm gears create a lot of friction, they make a very poor open differential (i.e. lots of friction). Well, if we really didn't want an open differential in the first place, we want something that will transfer torque side to side, then we got what we wanted!

The amount of torque that you can put to the tire with good traction is a function of the amount of traction you can put to the tire with poor traction. That ratio, the bias ratio, is usually around 3:1.

When you are braking and turning, the inside rear can get pretty light. Front drive cars often pull the inside rear all the way off the ground. Anyway, when you brake the diff tries to get as much torque out of the unladen tire as it can get. Remember, the 2 wheels are physically connected by gears, not clutches. When the tire gets light enough and you are trying to put a large enough torque to it, it spins backwards.

It sounds weird, it looks weird, it feels weird, and it does nothing positive to the handling of the car. I hope I explained it well enough. If not, then you're just going to have to trust me.....or find out for yourself!

Cheers.

[nicholasc]

To understand how a simple diff works go purchase a lego technic set that has one - there are several current ones.

[wegmann]

Originally posted by nicholasc
To understand how a simple diff works go purchase a lego technic set that has one - there are several current ones.

Those things are pretty cool. The one I saw even had the crankshaft and cylinders pumping away, IIRC.

But being an open diff, it doesn't really help explain Fat Boy's phenomenon, so I'm going to have to trust him for now. I understand the theory of his explanation, just not the exact mechanics. But that's fine, I'll go home tonight and figure it out.

[BRIAN GLOVER]

Jack your drivewheels off the ground-engine off-place the car in gear or drive-release handbrake-rotate one wheel by hand-the other wheel will go in the opposite direction. Some old cars had a handbrake that clamped the drive shaft. Yank on that brake while moving forward in a turn and the unloaded wheel will turn backwards. Go to your local hobby shop and check out the RCMs.

Originally posted by wegmann


Those things are pretty cool. The one I saw even had the crankshaft and cylinders pumping away, IIRC.

But being an open diff, it doesn't really help explain Fat Boy's phenomenon, so I'm going to have to trust him for now. I understand the theory of his explanation, just not the exact mechanics. But that's fine, I'll go home tonight and figure it out.

[McGuire]

Originally posted by Fat Boy
I'm not a real proponent of running a spool. It just came up in the conversation, so I commented on it.

Whew, you had me wondering there for a minute.

As I'm sure you know but speaking at large, the spool works on ovals because the corner radii are large, not to mention all left -- you can use tire stagger and other assymmetries (caster and camber split etc) to make the car turn into the monster push generated by the rear tires trying to push the car out to the wall. You are so right that spools are "stable" ...going straight is all they want to do.

And with a smallbore car on a roadcourse spools are not a humongous problem because the inside rear corner is so lightly loaded anyway, and there are things you can do to minimize dragging the inside rear (with droop etc) and get away with it. Kinda reminds me of a famous Colin Chapman quote which resurfaced recently. "Any suspension will work if you don't let it," was his droll observation.

Even in its heyday in the big cars the spool was slower from corner entry to apex, but could make it up on corner out with proper setup and a real racing driver. Against a modern limited slip, no hope. In order to get any "rotation," at some point the driver had to pitch the car with the steering wheel and/or heavy brake bias to get the thing pointed approximately at corner exit. I think the last big-bore cars I saw use a spool to any effect were some varieties of 962.

[McGuire]

Originally posted by wegmann
But being an open diff, it doesn't really help explain Fat Boy's phenomenon, so I'm going to have to trust him for now. I understand the theory of his explanation, just not the exact mechanics. But that's fine, I'll go home tonight and figure it out.

A differential is just a set of four bevel gears which reside in the case to which the ring gear is bolted. There are two crown bevel gears, one on the inside end of each drive axle, and two pinion bevel gears, which run on a cross shaft in the case. All four bevel gears (the "spider gears") are in constant mesh 90 degrees from each other while their case ("carrier") rotates with the ring gear.

If the torque applied to the ring gear is greater than the resistance on the axles, the case and ring gear will turn and so will both axles. If the resistance (traction) is equal on both wheels, the pinion bevels have no reason to rotate, both crown bevels will rotate at equal speed and both axles will recieve the same torque. If the resistance on one wheel is greater, the pinions will rotate and torque will be transferred in proportion to the opposite axle.

The differential allows the two drive wheels to rotate at different rates to compensate for the difference in corner radius from inside to outside. So a conventional open differential is a perfect torque-splitting device, for better or worse. If you jack one wheel off the ground and then rotate the ring gear, the elevated wheel will recieve all the torque while the grounded wheel recieves none.

But if the torque on one axle becomes greater than the torque on the ring gear and case, the other axle will rotate in the opposite direction. One axle and its crown bevel are driving the other through the two pinion bevels. So as Brian says, if you elevate the drive axle on your own car, make the ring gear stationary and rotate one wheel by hand, the other wheel will indeed rotate in the opposite direction. The crown bevel gear on your end is simply turning the bevel gear on the other axle in the opposite direction, through the pair of pinion gears between them.

A spool has no torque-splitting function. The torque delivered to each rear tire is solely a function of their respective friction with the ground. If both rear wheels have equal traction, they will each recieve 50% of the torque. But if one rear tire loses 100% of its traction, the other suddenly sees a 100% increase in torque. That's why dragsters (which invariably use spools) have a very narrow rear track. If one wheel loses traction, the "drive wheel steering couple" is smaller with the rear wheels closer together. If they were spread out six feet apart, a small difference in traction between the two wheels would have a much greater tendency to steer the car.

One very cheap and time-honored method for making a "spool" is simply to weld all four spider gears together so the pinions cannot rotate. Then both axles must always follow in rotation with the ring gear.

Damn, sure wish we had a blackboard.

[McGuire]

Originally posted by nicholasc
To understand how a simple diff works go purchase a lego technic set that has one - there are several current ones.

With all due respect, Screw that. We are not children. At some point it comes time to put away childish things. Buy a $100 set of tools and a $50 car, take it apart to its last nut and bolt and study how it works. When you are done you will have one very large pile of parts, and a better education than you can get in four years in any engineering college on earth. If you can then put it back together and make it run properly, that's a doctorate.

[JwS]

Hmmm.
So what do you get if you buy 2 different $50 cars, take them apart and then assemble the parts into one car that runs properly?
JwS

[McGuire]

Originally posted by JwS
Hmmm.
So what do you get if you buy 2 different $50 cars, take them apart and then assemble the parts into one car that runs properly?
JwS

A photograph of your creation in Hot Rod magazine.

[Greg Locock]

The neat thing about the Technics diff is you can change the friction in the spider gears and make it behave like an LSD.

Also, when I had to build a full ADAMS model of a diff I needed to keep all the coordinate systems straight, a small Lego diff on my desk was a much better option than wandering out to the workshop!

I don't think dismantling things teaches you much. I've been dismantling things since I was 3, but only started putting them back together, better, since I was 15.

[clSD139]

Originally posted by mini1275
Ok, just one more question

I´ve seen that some people brakes through the differential,

Hummer drivers also use this technique if a wheel is spinning in the air instead!

BTW. I pulled lego-cars for years, in turns, over jumps, pavement... cool!!!

[Big Block 8]

Originally posted by McGuire

With all due respect, Screw that. We are not children. At some point it comes time to put away childish things. Buy a $100 set of tools and a $50 car, take it apart to its last nut and bolt and study how it works. When you are done you will have one very large pile of parts, and a better education than you can get in four years in any engineering college on earth. If you can then put it back together and make it run properly, that's a doctorate.

That's a statement I've heard before - but it very much depends of how deep that studying is. Frankly, I'd feel very uncomfortable in a car taken apart and put together by an average doctor from the woods, but I'd feel equally uncomfortable, if I had to rely on a complex engineering thesis to be made by an average mechanic from the woods, even though if he/she could take apart a car and then put it together again!

[Chickenman]

Just a comment guys...well done!! One of the more intriging threads that I have read in some time. Hmmm...want to make F1 cheaper? Maybe they should all be required to run spools LOL