Variable F/R torque split (to 50/50) idea

Lets talk clutches and everything drivetrain related

Variable F/R torque split (to 50/50) idea

Postby MWP » Fri Oct 11, 2013 3:32 am

Ok, this maybe an insane idea but call this brainstorming ;)

There is a company called Haldex (part of BorgWarner) that make compact electronically variable clutch units.

They look/work like this (this is a recent gen 5 unit):
Image

How its used on an Audi TT:
Image

They mainly get used in Audi's and look like this (this is an Audi TT unit):
Image

When combined with a rear diff it looks like this (also from an Audi TT):
Image

I hope you guys can see what im thinking here...
We lock up the viscous center in the transfer case so there is always a 50:50 split of torque to the front axles and to the rear drive-shaft.
We then adapt a haldex unit in front of our rear diff, or use a full diff from something like an Audi TT.

We would then have an electronically controlled variable 100F/0R to 50F/50R torque split.
Obviously moving more power to the rear than 50/50 would be nice, but its not possible without using a different gearbox.

Im unsure yet how the haldex units are controlled, but it wouldn't be too difficult to come up with a controller for it.

Thoughts?
Members don't see the above ad. Register now - it's free!
Last edited by MWP on Fri Oct 11, 2013 5:10 am, edited 1 time in total.
My collection of GT4 documentation: http://gt4.mwp.id.au
Daily: Celica GT4 ST185 (~170kw/atw)
Project: Toyota RA28 '77 Celica (1UZ-FE powered)
Chairman of the SA Classic Celica Club.
MWP
Established Member
 
Posts: 1599
Joined: Mon Jun 01, 2009 11:02 am
Location: Adelaide, South Australia

Re: Variable F/R torque split (to 50/50) idea

Postby MWP » Fri Oct 11, 2013 3:51 am

There is also the possiblity of using a much more advanced unit like:
Image

... but i think that might be taking it too far.
The control system required is probably out of the range of DIY.
My collection of GT4 documentation: http://gt4.mwp.id.au
Daily: Celica GT4 ST185 (~170kw/atw)
Project: Toyota RA28 '77 Celica (1UZ-FE powered)
Chairman of the SA Classic Celica Club.
MWP
Established Member
 
Posts: 1599
Joined: Mon Jun 01, 2009 11:02 am
Location: Adelaide, South Australia

Re: Variable F/R torque split (to 50/50) idea

Postby lumbercis » Fri Oct 11, 2013 4:41 am

Why would you want to do this?
Caldina powered...
Image
lumbercis
Gold Member
 
Posts: 2346
Joined: Fri May 13, 2005 4:52 am
Location: Washington, D.C.

Re: Variable F/R torque split (to 50/50) idea

Postby MWP » Fri Oct 11, 2013 5:05 am

Because the viscous center diff's we have are far from ideal for high performance applications.
They were ok back in the early 90's, but they are VERY outdated today.
If you need to ask why, you should be looking at some older threads like: viewtopic.php?f=43&t=37969

Can we please not start a debate about viscous centers or 4wd vs all-wheel-drive vs rear-assisted-drive in here! There are plenty of old threads on those topics.
My collection of GT4 documentation: http://gt4.mwp.id.au
Daily: Celica GT4 ST185 (~170kw/atw)
Project: Toyota RA28 '77 Celica (1UZ-FE powered)
Chairman of the SA Classic Celica Club.
MWP
Established Member
 
Posts: 1599
Joined: Mon Jun 01, 2009 11:02 am
Location: Adelaide, South Australia

Re: Variable F/R torque split (to 50/50) idea

Postby MWP » Fri Oct 11, 2013 1:50 pm

There is a great tech doc here about how the Haldex unit works in a VW 4MOTION equipt car:
http://admin.audionlinetraining.com/Upl ... n%20IV.pdf
My collection of GT4 documentation: http://gt4.mwp.id.au
Daily: Celica GT4 ST185 (~170kw/atw)
Project: Toyota RA28 '77 Celica (1UZ-FE powered)
Chairman of the SA Classic Celica Club.
MWP
Established Member
 
Posts: 1599
Joined: Mon Jun 01, 2009 11:02 am
Location: Adelaide, South Australia

Re: Variable F/R torque split (to 50/50) idea

Postby underscore » Fri Oct 11, 2013 1:54 pm

So from what I'm gathering here this basically allows slip before the rear diff, reducing the power that goes to the rear end? Wouldn't this make it more 50/0 than 100/0?

If you were going to do this I would say mating it to the front of our rear diff would be easier than trying to cram an Audi rear diff into our subframe and try to make some funky axles. But that begs the question, why would you want to reduce the power going to the rear wheels? It sounds like Audi would be using this to make a normally FWD car up to 50/50 if needed.
★ 1991 GTFour RC ~ "Rebel Scum" ★
It's for sale! http://www.alltrac.net/phpBB2/viewtopic.php?f=64&t=52181
Build thread http://www.alltrac.net/phpBB2/viewtopic.php?f=44&t=44216
underscore
GTFour God
 
Posts: 3822
Images: 0
Joined: Mon Apr 09, 2007 11:29 pm
Location: Kelowna, BC

Re: Variable F/R torque split (to 50/50) idea

Postby CMS-GT4 » Fri Oct 11, 2013 2:32 pm

Unless you add a front LSD, I don't see how this system is an improvement over stock. Unless you can find a way to take power from the front and have a rear bias, you will always need to resolve the open front LSD.
Coldiron
92 st185 (205 swap) | 03 350Z Track (sold) | 13 Pilot | 17 Mazda3
www.noritoy.com | noritoy.deviantart.com | http://www.alltrac.net/phpBB2/viewtopic.php?f=44&t=19891
User avatar
CMS-GT4
GTSilver Member
 
Posts: 10345
Images: 7
Joined: Mon Aug 09, 2004 9:12 pm
Location: St Louis Area (IL)

1992 Toyota Celica

Re: Variable F/R torque split (to 50/50) idea

Postby MWP » Fri Oct 11, 2013 2:39 pm

underscore wrote:So from what I'm gathering here this basically allows slip before the rear diff, reducing the power that goes to the rear end? Wouldn't this make it more 50/0 than 100/0?


No.
When its fully unlocked, youll get 100% FWD and 0% RWD.
When its fully locked you get 50% FWD and 50% RWD.

Before you start thinking this is worse than the viscous center diff... it is not, its far better (more below).

If you were going to do this I would say mating it to the front of our rear diff would be easier than trying to cram an Audi rear diff into our subframe and try to make some funky axles.


Yeah, agreed. Or it could be put on the end of the gearbox transfer case.

But that begs the question, why would you want to reduce the power going to the rear wheels?


Because you have to... you cant have a solid 50/50 torque split.
There has to be some way of the front wheels being able to spin at a different speed to the rear to allow for cornering, slightly different wheel sizes, etc.
This is why the viscous center (and all diff's for that matter) is there.

A haldex unit would take the place (not physically, but operationally) of the viscous center.
It has the advantage over the viscous center of being able to completely lock up when the front wheels start to spin.
The viscous center doesnt do this. It slowly transfers more torque to the rear as there is more slip. There is also a limit to how much torque it can transfer, as they age, or you increase engine power, you may only be getting say 20% of torque to the rear wheels.
My collection of GT4 documentation: http://gt4.mwp.id.au
Daily: Celica GT4 ST185 (~170kw/atw)
Project: Toyota RA28 '77 Celica (1UZ-FE powered)
Chairman of the SA Classic Celica Club.
MWP
Established Member
 
Posts: 1599
Joined: Mon Jun 01, 2009 11:02 am
Location: Adelaide, South Australia

Re: Variable F/R torque split (to 50/50) idea

Postby MWP » Fri Oct 11, 2013 2:42 pm

CMS-GT4 wrote:Unless you add a front LSD, I don't see how this system is an improvement over stock. Unless you can find a way to take power from the front and have a rear bias, you will always need to resolve the open front diff.


If you transfer more torque to the rear wheels, there will be less power to spin up one of the front wheels.

Even with an open front diff, it'll still be far better than the 4WD system we have at the moment.
Last edited by MWP on Fri Oct 11, 2013 3:07 pm, edited 1 time in total.
My collection of GT4 documentation: http://gt4.mwp.id.au
Daily: Celica GT4 ST185 (~170kw/atw)
Project: Toyota RA28 '77 Celica (1UZ-FE powered)
Chairman of the SA Classic Celica Club.
MWP
Established Member
 
Posts: 1599
Joined: Mon Jun 01, 2009 11:02 am
Location: Adelaide, South Australia

Re: Variable F/R torque split (to 50/50) idea

Postby MWP » Fri Oct 11, 2013 2:59 pm

BTW, if you want to do a quick test...
Jack up the front of your Alltrac/GT4, and leave the rear wheels on the ground (chock them!!).
Get in the driver seat, start the car and put it in 1st gear. Very slowly let out the clutch.

Youll probably find youll spin up both front wheels no problem, while the rears dont turn.
This is the viscous center not transferring torque to the rear wheels like it ideally should be.
If the viscous center was giving you a proper 50/50 F/R torque split, you would either stall the engine, or the rear wheels would drive over the chocks.

Using the haldex unit, you would get that true 50/50 split.
It would would either stall the engine, or the rear wheels would drive over the chocks as a good 4WD system should do.
My collection of GT4 documentation: http://gt4.mwp.id.au
Daily: Celica GT4 ST185 (~170kw/atw)
Project: Toyota RA28 '77 Celica (1UZ-FE powered)
Chairman of the SA Classic Celica Club.
MWP
Established Member
 
Posts: 1599
Joined: Mon Jun 01, 2009 11:02 am
Location: Adelaide, South Australia

Re: Variable F/R torque split (to 50/50) idea

Postby CMS-GT4 » Fri Oct 11, 2013 3:59 pm

We have already done this test, and examined this operation in this thread:
viewtopic.php?f=4&t=45503

Unless your VC is shot, the car should operate at 50/50 split all the time. There is some give in there because if it was a locked system all the time it would bind and eventually something would break. The same reason you can't drive the early manual locking version at high speeds or all the time. The locking mechanism is only designed to act when there is a difference in front to rear speed.

If you watch video 3 of the series it will explain to you fully how the power split operation in our cars actually work and how they operate in slip and driving conditions. The first 2 videos covers the original manual locking system.

You can try this if you want, but I really think for the money and effort you would be better off trying to just swap an awd gearbox with a proper LSD system already. Most likely what will happen is if you permanently lock the VC section, you will likely bind something if and when power gets transferred to the rear at high speeds or at a long period of time. From what I have looked at, the EVO boxes are the closest match, and one gtfour uses an aftermarket evo gearbox already.

I was looking at a gearbox and transfer that was pulled out of a 185 this weekend because we thought we might be able to adapt a transfer case from another car to give us better options, but sadly the diffs are in the gearbox and just the VC is in the transfer, so that rules that out.

I think more interesting research could be done into the Porsche that uses VC in its awd system as it has a rear bias. I have talked to a few mechanics and no one really understands how this works since the VC is supposed to prevent difference in speeds, but there must be a way to mimic this mechanical system in that way.
http://www.awdwiki.com/en/porsche/

EDIT: After reading this, I think even trying to change the bias might be far harder than I could want to attempt. http://www.stealth316.com/2-awd3.htm
Last edited by CMS-GT4 on Fri Oct 11, 2013 6:08 pm, edited 1 time in total.
Coldiron
92 st185 (205 swap) | 03 350Z Track (sold) | 13 Pilot | 17 Mazda3
www.noritoy.com | noritoy.deviantart.com | http://www.alltrac.net/phpBB2/viewtopic.php?f=44&t=19891
User avatar
CMS-GT4
GTSilver Member
 
Posts: 10345
Images: 7
Joined: Mon Aug 09, 2004 9:12 pm
Location: St Louis Area (IL)

1992 Toyota Celica

Re: Variable F/R torque split (to 50/50) idea

Postby underscore » Fri Oct 11, 2013 6:02 pm

MWP wrote:There is also a limit to how much torque it can transfer, as they age, or you increase engine power, you may only be getting say 20% of torque to the rear wheels.


If this is your concern, I'd suggest getting one of the upgraded units from Mario. Otherwise as far as I can tell you're doing a lot of work for basically the same system we have now.
★ 1991 GTFour RC ~ "Rebel Scum" ★
It's for sale! http://www.alltrac.net/phpBB2/viewtopic.php?f=64&t=52181
Build thread http://www.alltrac.net/phpBB2/viewtopic.php?f=44&t=44216
underscore
GTFour God
 
Posts: 3822
Images: 0
Joined: Mon Apr 09, 2007 11:29 pm
Location: Kelowna, BC

Re: Variable F/R torque split (to 50/50) idea

Postby MWP » Fri Oct 11, 2013 11:33 pm

CMS-GT4 wrote:Unless your VC is shot, the car should operate at 50/50 split all the time. There is some give in there because if it was a locked system all the time it would bind and eventually something would break. The same reason you can't drive the early manual locking version at high speeds or all the time. The locking mechanism is only designed to act when there is a difference in front to rear speed.


Our viscous centers never "lock".
There is always some slip in between the viscous plates, which increases as they wear (which they do), or as you increase torque loading (they can only transfer so much torque before they slip).
Our viscous centers "try" to produce a 50/50 split, but they never actually do it unless you are on a rather slippery surface.

The VC centers when brand new were only designed to transfer a max of XX ft/lb of torque.
Dont ask me what that figure is, but it wouldn't be much more than 50% of what the factory engine is capable of producing.
Above this value they do slip.
This is why rally teams and the units Mario modify have a higher torque rating, but its still not high enough for even a modestly modified 3SGTE engine.

Its not a problem on loose surfaces as the VC can easily transfer enough torque to spin the rear wheels as well.
On sticky surfaces (tarmac) with a powerful engine, its a very different story.

You can try this if you want, but I really think for the money and effort you would be better off trying to just swap an awd gearbox with a proper LSD system already. Most likely what will happen is if you permanently lock the VC section, you will likely bind something if and when power gets transferred to the rear at high speeds or at a long period of time. From what I have looked at, the EVO boxes are the closest match, and one gtfour uses an aftermarket evo gearbox already.


At a cost of?
Locking up our VC, and using a Haldex unit will be *far* cheaper.

I think more interesting research could be done into the Porsche that uses VC in its awd system as it has a rear bias. I have talked to a few mechanics and no one really understands how this works since the VC is supposed to prevent difference in speeds, but there must be a way to mimic this mechanical system in that way.
http://www.awdwiki.com/en/porsche/


They do it by having a hard link to the rear diff, and using a clutch system to change how much is transferred to the front (the exact reverse of how our gearbox/diffs are setup).
There is no mystical voodoo involved, and many other cars with rear bias do it the same way.

EDIT: After reading this, I think even trying to change the bias might be far harder than I could want to attempt. http://www.stealth316.com/2-awd3.htm


We'll never be able to change to rear bias without using a completely different gearbox/diff. Its just not possible.
We can however make the best of the gearbox we have which is what this thread is about.

underscore wrote:
MWP wrote:There is also a limit to how much torque it can transfer, as they age, or you increase engine power, you may only be getting say 20% of torque to the rear wheels.

If this is your concern, I'd suggest getting one of the upgraded units from Mario. Otherwise as far as I can tell you're doing a lot of work for basically the same system we have now.


As i said, it still wont be as good a Haldex unit.

The other big plus about a Haldex system is its electronically tunable.
At the drag strip for example, you could flick a switch and set the Haldex unit to full lock before you even get to the line.
Youll then get a proper hard 50/50 torque split from the moment you drop the clutch. It'll shave a decent amount off the 60ft time.

There is a very good reason why viscous centers arent used that much anymore, and active clutched systems like the Haldex are popular.
They simply work much better at giving the rear wheels more torque, and quicker.
Why would manufacturers install a more complex, more expensive system otherwise ;)
My collection of GT4 documentation: http://gt4.mwp.id.au
Daily: Celica GT4 ST185 (~170kw/atw)
Project: Toyota RA28 '77 Celica (1UZ-FE powered)
Chairman of the SA Classic Celica Club.
MWP
Established Member
 
Posts: 1599
Joined: Mon Jun 01, 2009 11:02 am
Location: Adelaide, South Australia

Re: Variable F/R torque split (to 50/50) idea

Postby phattyduck » Tue Oct 15, 2013 5:22 pm

I want to make sure we are all on the same page here...

For a standard open differential (Front, Rear, Center, whatever) - there is ALWAYS a 50/50 torque split. When one output side of the differential spins and 'loses' torque, the the other output shaft loses the same amount. Hopefully we can agree on that one.

A fully locked differential (well, that's not a differential now, is it?) - is anywhere from a 100:0 to 0:100 split - whatever output side is going at a slower rate gets 100% of the torque.

Fun video:
http://www.awdwiki.com/en/videoviscous/

As for a viscous center - it is good for safety (getting out of low traction situations), but not the best for performance. A well designed one should be able to transfer a large amount of torque, though not instantaneously or preemptively... Keep in mind that video above though - VC units can easily be ruined by abuse or age/wear.

In a perfect world for my uses though, I'd want a viscous center and F/R torsen diffs. I'm not 100% about performance though - street-ability is high on my list too.

-Charlie
'89 Camry LE Alltrac 3S-GTE - SV25/ST205 hybrid
'99 4Runner SR5 4WD 5VZ-FE
Previous: '88 Camry LE Alltrac 3S-GE BEAMS, '90 Camry 3S-GTE, '90 Camry DX
User avatar
phattyduck
Established Member
 
Posts: 1045
Joined: Wed Dec 07, 2005 8:09 am
Location: Pasadena, CA

Re: Variable F/R torque split (to 50/50) idea

Postby mx6er2587 » Mon Nov 25, 2013 5:21 pm

I have a haldex gen 3 in my S60R. Now, I know they've come a long way in improving them since 2005 but its not a very good AWD system at all. The alltrac stock system is much better.

The selling point for the Haldex system was always its compact size and ease of use. It made it very easy for manufactures to convert FWD into AWD platforms.

Look at the vehicles that employ this system, they are either not high performance cars, or they are factory hotrod variants of FWD vehicles.

This is not something to waste your time or effort on.
mx6er2587
Established Member
 
Posts: 1163
Joined: Sun Jan 27, 2008 12:08 am
Location: Pittsburgh PA

Next

Return to Drivetrain Central

Who is online

Users browsing this forum: No registered users and 3 guests