25 replies to this topic

[Chris Wilson]

Car in mind is a Mazda RX-7 FD model of mid nineties. They run what Mazda term a "Power Plant Frame" which is a lightweight pressed steel frame tying the gearbox extension housing to the front of the differential casing The car has independent rear suspension by double wishbones. Why would a maker choose this means of locating the rear of the gearbox and the nose of the diff instead of using a gearbox X member and supporting the diff nose of the rar subframe? Some quite low end Opels used a torque tube, as did some of the bigger Peugeots like the old RWD 504. If one were to fabricate conventional mounts for the diff nose using a weld in multi point roll cage to tie the shell together more rigidly, and did something similar with a different but similarly cited gearbox (thinking engine and box change to something with pistons) what disadvantages can you think of, handling wise? Any? My personal take on this is that it may have been instigated to allow compliant diff mounting to reduce NVH, yet control unwanted diff nose movement to control drive train wind up and wheel hop. I can't find much on the pros and cons of this et up. Thanks.

[Greg Locock]

Note there are two different types, tube to the back of the gearbox, then a CV joint , or a rigid frame connecting the diff to the gearbox.

You are talking about the latter, Peugeot et al used the former (I think).

Its an ideal setup for the driveline engineer, but can be a PITA for noise and vibration people - although it offers intriguing possibilities.

Eliminating the gearbox to cross member mounting is a really good idea, and because the moment arm is so long you can soften off the diff bushes a lot. Net result is that it could be very refined. The downside is that powertrain bending frequencies will be much lower. In practice this doens't seem to be a concern.

So far as the driveline engineer is concerned, he's got total contrrol over the UJ angles of the propshaft, in fact he doesn't have any to worry about. That means the propshaft can be designed to do one job only, transmit torque.

[GreenMachine]

Mazda have been building MX5s this way since 1989 (alloy in this case). Must have something going for it ...

[pugfan]

Note there are two different types, tube to the back of the gearbox, then a CV joint , or a rigid frame connecting the diff to the gearbox.

You are talking about the latter, Peugeot et al used the former (I think).

The 504 (and 505) wagon and a cheapened version of the 504 sedan used the former, a solid axle and tube with triangulation is pivoted off the rear of the gearbox. The sedan used the latter, a thick wall tube connecting the diff and the gearbox.

Side stepping the clutch whilst stationary does result in some part of the driveline thumping against the rear-subframe or possibly chassis at about 3-4Hz. Maybe related, apparently works 504 rally cars had a longitudinal link from the torque tube to the rear sub-frame which is mounted on rubber bushes to the chassis.

[mariner]

Greg , IIRC Lotus came up with a very sophisticated version of the central g/box to diff tube for the rear engined "corvette" they built for GM when owned by GM.

I think it also had some very clever fluid feature in the drive shafts but I never understood how that worked?

[Greg Locock]

Greg , IIRC Lotus came up with a very sophisticated version of the central g/box to diff tube for the rear engined "corvette" they built for GM when owned by GM.

I think it also had some very clever fluid feature in the drive shafts but I never understood how that worked?

The Chev Indy showcars? I worked on one of them (we built three). The job came down to how much sound absorption we could get into the firewall. I can't remember what the driveline looked like. Do you know how f-in BORING showcars are to work on? Very.

[Ray Bell]

Originally posted by pugfan
The 504 (and 505) wagon and a cheapened version of the 504 sedan used the former, a solid axle and tube with triangulation is pivoted off the rear of the gearbox. The sedan used the latter, a thick wall tube connecting the diff and the gearbox.....

The use of the 'former' goes back well before the 504...

The 203, 403 and 404 used them throughout their ranges. Even the 202, so I suspect that the pre-war '02' and probably '01' series cars had them. As, indeed, did a great many cars of the teens, twenties and thirties. Buick were still using them into the early fifties, Chevrolet as well if I'm not mistaken, while Ford dropped the torque tube drive in 1949 after having it virtually since inception.

The 'triangulation' you mention is necessary physically, but not important in the context of this subject. One reason for Peugeot's adherence to the regular torque tube was its inherent ability to enhance traction in poor conditions, while it always fitted in well with the worm drive final drive. Of course, from 1967 onwards they were building wagons and utilities with hypoid versions as well, but these didn't venture into the sedans because they'd have intruded too much into the passenger space at the rear due to the increased height of the final drive input shaft.

The 'latter' type as used in the upper echelon 504, 505 and all 604 sedans had a couple of benefits. Noise suppression was one, if I'm not mistaken, while the cancellation of torque reactions meant that mounts at both ends of the mechanical train had less work to do.

.....Side stepping the clutch whilst stationary does result in some part of the driveline thumping against the rear-subframe or possibly chassis at about 3-4Hz. Maybe related, apparently works 504 rally cars had a longitudinal link from the torque tube to the rear sub-frame which is mounted on rubber bushes to the chassis.

Never seen the rally car addition, and I'm sure I looked underneath one of the 504 London-Sydney cars at one stage. It could be, of course, that this came later with the V6 coupes.

Side-stepping the clutch in a 404 gives a 'graunching' sound that I've never been able to work out. It could be that your thumping is the torque tube (which, in this arrangement, Peugeot refer to as the 'link tube') hitting the rear crossmember.
This unit tends to turn upwards at the front as the suspension squats if the rubbers are in poor shape, while it might well be that other reactions cause the differential to push the tube downwards at the same time. In static positions, there'd be somewhere close to an inch between them... no more.

[Chris Wilson]

The RX-7 FD model range use a conventional propshaft with Hardy Spicer type cross joints each end. Maybe the rigidity of the "power plant frame" is questionable, I know they are commonly prone to cracking. Thanks for the replies so far. Putting NVH aside as irrelevant in this proposed scenario, would there be any handling changes by mounting diff nose and `box tailshaft conventionally? I would probably solid mount the diff, in fact. I can't myself see any potential issues. Thanks.

Edited by Chris Wilson, 02 March 2010 - 17:00.

[gordmac]

The Opel system was a live axle located by a torque tube rubber mounted at the front to the shell, two trailing arms and a panhard rod. There was a short propshaft between the gearbox and the torque tube. I have competed with a few Chevettes with this arrangement (would I be right in thinking it provides a lot of anti squat?). I thought the traction "off the line" was a bit suspect and the back seemed to lift and tramp badly under heavy braking (leaving the spare wheel in helped!), cars with a "5 link" setup seemed a lot better.
With an independent system as you describe I can't see a handling reason for it unless there is a compliance issue being managed. My guess would be cost and/or NVH.

[Tony Matthews]

Do you know how f-in BORING showcars are to work on? Very.

It seems that the more glamorous, interesting or exciting a job seems to those who are not involved, the more tedious and boring those jobs are!

[Greg Locock]

It seems that the more glamorous, interesting or exciting a job seems to those who are not involved, the more tedious and boring those jobs are!

(a) you are always limited in access to the car
(b) whatever you do mustn't disturb what's already been built
© the car is practically undrivable and breaks down once a day or more often
(d) the stylists rip out all your stuff when they fit a different interior trim
(e) you can't use black tape glue guns or screws

I like jobs that involve welding, angle grinders and fires.

The other glam job I regard as basically dull is supercars.

Admittedly if done the right way they are a challenge, but the one I was involved in at Lotus basically came down to decsions like - oh, we can't fit the fuel tank in, let's add another 100 mm to the wheelbase. Oh, we can't fit the power takeoff from the (split) V12 crankshaft, lets add another 150 mm (I kid you not) to the width. There is a fairly direct relationship between that bloated pig (that never got off the drawing board) and the first Bugatti. Whereas working on a real car you have to fight to get every 25mm of package space.

[Ray Bell]

A few of competition cars that have used the fixed torque tube (or 'link tube') principle:

1. The Clisby hillclimb car of the early fifties (if not late forties). It had a Douglas motorcycle engine at the front and the chassis was a tube that attached to the clutch housing and had a differential bolted up at the other end. Everything hung off this backbone that enclosed the drive shaft.

2. Taking a leaf from Clisby's book, his friend Eldred Norman built the 'Norholfordor' with a Ford Zephyr 6 inline engine that had a chunk of BHP's best milled out and bolted in place of the timing cover. To it was welded a shortened crossmember from a pre-'53 model Holden. The Zephyr engine had a shroud cast into the block that surrounded the flywheel, so a flange on a piece of pipe of about 9" diameter fitted up to that neatly, with a modified Tempo Matador (look that up...) transaxle on another flange at the other end. The thing was supercharged and was literally dynamite, when driven with gusto it shook up plenty of Cooper Climaxes five years after its 1955 debut. Now raced by Graeme Snape.

3. Much more modern, at least one of the Alfetta Sports Sedans running in Australia is built with a triangulated tubular arrangement that ties the engine to the rear-mounted Hewland. Power is from a 6-litre Chevrolet, of course.

[cheapracer]

(e) you can't use black tape glue guns or screws

(f) I like jobs that involve welding, angle grinders and fires.

(e) count me out then.

(f) ok, you get a straw bed and 2 bowls of rice a day but supply your own vegetables, when can you start?

[gruntguru]

(e) count me out then.

(f) ok, you get a straw bed and 2 bowls of rice a day but supply your own vegetables, when can you start?

and no welding after bedtime. The straw is highly flammable.

[Engineguy]

With the engine/trans solidly connected to the diff I would think there would be quicker realization of throttle application at the diff... no waiting for all those engine and diff rubber mounts to be compresed. In fact with the engine and diff connected by torque tube the engine will not rock relative to the frame at all, won't try to twist the chassis... it'll only try to lift , rotating about the rear axle axis. Yes?

[gruntguru]

With the engine/trans solidly connected to the diff I would think there would be quicker realization of throttle application at the diff... no waiting for all those engine and diff rubber mounts to be compresed. In fact with the engine and diff connected by torque tube the engine will not rock relative to the frame at all, won't try to twist the chassis... it'll only try to lift , rotating about the rear axle axis. Yes?

For live axle layouts the torque tube eliminates the weight transfer transfer across the axle due to tailshaft torque.

[Ray Bell]

How so?

With a live axle you can't have a solid tube to the engine, so it must freely pivot up at the front. This means there's no torsional limitation on the rear axle, which still has a pinion that's trying to climb up the side of a pinion.

It's countered to some degree by the torque reaction that's trying to lift the car, but not totally.

Worm drive... did I mention it?

[Tony Matthews]

How so?

With a live axle you can't have a solid tube to the engine, so it must freely pivot up at the front. This means there's no torsional limitation on the rear axle, which still has a pinion that's trying to climb up the side of a pinion.

But surely you can't have a torque tube and a live axle. The final drive must be fixed to the chassis. That is my understanding - I have a car that has a torque tube, so I am informed, I always thought of it as a propshaft, but the bevel box is fixed and from there the transmission is by chain.

[Ray Bell]

Yes, you can have a torque tube and a live axle...

A propshaft is not a torque tube. A 'propshaft' can run inside a torque tube, however.

[Tony Matthews]

Yes, you can have a torque tube and a live axle...

A propshaft is not a torque tube. A 'propshaft' can run inside a torque tube, however.

I have struggled a bit with the differences, Ray - I just haven't applied mysely!



Not my chassis, but the same as. I have been told that this is a torque tube, and I assumed that this was because there is no angular movement. As usual, I am here to learn!

Edited by Tony Matthews, 04 March 2010 - 12:50.

[Greg Locock]

You are far better at discerning details from photos than me (that or you make a lot of stuff up!), I can't see how that works. If the torque at the wheel is reacted at the gearbox via the long tube, then it is a torque tube.

[Powersteer]

A 'propshaft' can run inside a torque tube, however.

So how is it mounted to the engine when the axles move vertically? You talking about Dedion? Or it is pivoted?

[Tony Matthews]

You are far better at discerning details from photos than me (that or you make a lot of stuff up!), I can't see how that works. If the torque at the wheel is reacted at the gearbox via the long tube, then it is a torque tube.

I'm sorry, that was not the best way of showing what I meant. The box at the back is purely a bevel box, the solid back axle is mounted on radius arms behind this, and transmission is by chain, one for each gear. So, engine obviously rigid, bevel box rigid, is that a torque tube linking them?

Edited by Tony Matthews, 05 March 2010 - 07:48.

[Ray Bell]

If it's a 'link' tube as defined by Peugeot, and which some would call a torque tube, then there will be a propshaft within an outer case. The outer case will attach firmly to the final drive and the back of the gearbox, making a rigid connection between the two.

The conventional torque tube is a case around a propshaft too, but in this case it's pivoted on the back of the gearbox (the propshaft has a universal joint at this point) and rigidly mounted only to the final drive, which is in the middle of a live axle.

And then you have the Alfa Tipo B... but I digress.

[Engineguy]

While both concepts being discussed have been called torque tubes, they're totally different. The torque tubes used with IRS and frame mounted diff (Porsche 924, 928, current Corvettes, etc.) react against the ring'n'pinion torque back to the trans about the longitudnal axis... and the axle torque about the transverse axis. The Old Model T style (live axle) torque tube only reacts against the axle torque (about transverse axis)... otherwise a one-wheel bump would have to rotate the engine relative to the frame. It's more of a "lever tube" then torque tube.
.

[Chris Wilson]

While both concepts being discussed have been called torque tubes, they're totally different. The torque tubes used with IRS and frame mounted diff (Porsche 924, 928, current Corvettes, etc.) react against the ring'n'pinion torque back to the trans about the longitudnal axis... and the axle torque about the transverse axis. The Old Model T style (live axle) torque tube only reacts against the axle torque (about transverse axis)... otherwise a one-wheel bump would have to rotate the engine relative to the frame. It's more of a "lever tube" then torque tube.
.

Yes, I see this difference, my question is about a "lever arm" type frame, the diff is constrained transversely by two top mounts on rubber bushes about 12 or so inches apart. The frame merely stops the diff trying to rotate about the drive shafts axis. Still wondering if it can have any bearing on handling.... The more I consider this the less I believe it has any bearing Thanks for some interesting replies.