21 replies to this topic

[Chris Wilson]

I have fitted a Swedish made Tractive sequential dog box to my RWD Skyline track car, and need to now make a custom propshaft for it. the Tractive box has a C/V type rear flange on it. The stock prop is 2 piece, with a cross type front (gearbox flange) joint, and a middle and rear C/V joint. What governs whether I can make the setup a single piece prop design, with a C/V joint at each end? If I have a 2 piece made is there any potential issue with having all 3 joints of a C/V type? Thanks.

[cheapracer]

Length is certainly a factor for a single tailshaft but I don't know any rule of the thumb for what lengths etc. - Oz has many very good tailshaft builders and I always let them decide (gearbox/engine swaps have always been very popular here).

Depending on hp, errr torque, errr force, umm thrust you will reach a point where you may need to use universal joints which are far stronger.

[sblick]

Balancing multiple shafts so you don't get first order vibrations can be a real pain. If it is a street car that is

[gruntguru]

The maximum length is determined by whirling speed (the speed where the shaft turns into a skipping rope). Someone here who does this every day should be able to post a formula, but essentially the tube diameter needs to increase if you increase length or rotational speed. If you are OK with engineering formulas, Google "shaft whirling".

[phantom II]

You might want to plot the loci of the diff pinion. Complex movements can cause vibration under acceleration, hop or asymmetric loss or gain of traction which can be absorbed by a split shaft design. If that checks out, you can get these people to make you a single carbon fiber one. Send length and couplings.



I have one of these on my Vette. My original shaft whipped so much with the extra 150hp that the rubber ring inside the torque tube disintegrated. Now I don't need it. Don't even need couplings. The stock car uses rubber couplings.
Carbon Fiber has three times the torsional strength of steel. The torsional spring rate allows better tire hook-up and smoother shifting.The critical speed of carbon fiber is 2x better than aluminum. Critical speed is the RPM at which the driveshaft is computer projected to bend or whip. Exceeding critical speed can produce vibrations that can result in driveshaft failure.
There is much less shock load to your drivetrain.

My roadster drive shaft is only 12" long and 3" dia, coupled to a 07 BMW M3 diff and a T56 Tremec tranny. A local shop made it for me with my specs.

Originally posted by Chris Wilson
I have fitted a Swedish made Tractive sequential dog box to my RWD Skyline track car, and need to now make a custom propshaft for it. the Tractive box has a C/V type rear flange on it. The stock prop is 2 piece, with a cross type front (gearbox flange) joint, and a middle and rear C/V joint. What governs whether I can make the setup a single piece prop design, with a C/V joint at each end? If I have a 2 piece made is there any potential issue with having all 3 joints of a C/V type? Thanks.

[Greg Locock]

The classic whirl speed equation is indicative only for a propshaft since the mass distribution is non uniform and the ends are compliantly mounted. The errors could go either way.

Yes I did use to be a driveline engineer.

You can find individual cars that operate above or even at critical speed quite happily, others will blow up at 95% of critical. you might be lucky, you might not.

if the stock car is two piece I'd be hesitant to go to a single piece, unless you have room for a large diameter tube.

Graphite wrapped aluminium propshafts are a marketing ploy, the improvement in whirlspeed is minute. True graphite or MMC propshafts do offer a measurable increase in whirlspeed.

There is no problem as such with running 3 CVs in a row, just weight and cost. You will probably need an axial slip joint somewhere in the system. This can be as simple as a splined output shaft at the gearbox, or you could use a plunging type CV joint, or an inline spline. The latter is the cheapest and most compromised. On an IRS with a two piece it is probably good enough.

You could go

CV plunge CV CV

CV plunge UJ CV

or

CV plunge UJ UJ

The latter will give you the biggest problem with 2nd order from the UJs

[Scotracer]

1) Get a scafolding pole

2) Cut it

3) ???

4) Jone Done

[Catalina Park]

My old HQ race car had a good example of a poor tailshaft design. They used a long skinny tailshaft that used to get a whip at high speeds (shortly followed by the destruction of the tailshaft, the gearbox extension housing and the pinion shaft) I never broke one but I did get it to start whipping at about 110mph at Eastern Creek. The track had a bit of a dip on the main straight where the spectator tunnel goes under the track and when the car hit the dip the shaft would flex and stay flexed!

I went to Hardy Spicer in Sydney who made me a new standard shaft (that was not really standard!) and I never had a tailshaft problem again. They used to do a lot of motorsport shafts and would give a very good discount in exchange for a couple of stickers on the car. My HQ shaft was free.

[Greg Locock]

That sounds about the right speed. You could usually get away with a single piece shaft on a 2700 mm wheelbase roadcar if you are running a 2.92 or 3.08 diff, if you've got room for a 4 inch aluminium propshaft, and then drop back to a 3 1/4 inch steel if you can show it is safe. But once we went over 200 hp in the base car that wasn't an option - every car would have been able to drive through whirl. Or not.

[Greg Locock]

Incidentally the critical speed in rpm is proportional to the square root of k/m, for a single piece prop, if you ignore end effects and so on, which is dangerous, but so long as you INCREASE k/m then the critical speed should increase even in a real world installation.

k is proportional to the tube's wall thickness, times its young's modulus, times the OD ^3

m is proportional to the materials density *OD*wall thickness.

So, the critical speed = fudge_factor *(E*D^2/rho)^.5

That is why there is no advantage in increasing the wall thickness, and why aluminium vs steel makes no odds. E/rho for those two are virtually equal.

fudge_factor is hopefully more or less constant for a given car, so to increase the critical speed by 10% the simplest thing to do is to increase the OD by 10% (or a bit more)

Incidentally any junk in the main tube, such as axial sliding joints, or noise absorbing liners, will have a measurable effect on critical speed.

[cheapracer]

fudge_factor is hopefully more or less constant for a given car, so to increase the critical speed by 10% the simplest thing to do is to increase the OD by 10% (or a bit more)

I like rule of the thumbs. Thanks.

[Chris Wilson]

Some great info here, and my abject apologies for the huge time differential getting back. I had a PC crash and failed to log on to the forum as I hadn't kept an accurate record of user name or password.... I am now at the stage of building up the driveline. The stock layout on the 4 wheel drive versions of my car is as below. I am using the 4WD diff assembly, so have a C/V type flange on that and the gearbox has a C/V type flange that takes a GKN Motorsport C/V type MS3N014 I think I ought to suggest the propshaft people build an all C/V jointed shaft (3 C/V joints) ? Although more costly Greg seems to suggest the dynamics may be better. If NVH isn't an issue should I have the centre support solidly mounted? The C/V joint above, as recommended by Tractive, the gearbox makers, is a high plunge type. I was going to use these in all 3 positions, does this sound reasonable? Thanks. I have some photos of the dummied up gearbox and diff installation, plus the stock 2 whhel drive prop and a front 1/2 of a 4 WD prop I have acquired at http://www.chriswils...propshafts.html

[Greg Locock]

Some great info here, and my abject apologies for the huge time differential getting back. I had a PC crash and failed to log on to the forum as I hadn't kept an accurate record of user name or password.... I am now at the stage of building up the driveline. The stock layout on the 4 wheel drive versions of my car is as below. I am using the 4WD diff assembly, so have a C/V type flange on that and the gearbox has a C/V type flange that takes a GKN Motorsport C/V type MS3N014 I think I ought to suggest the propshaft people build an all C/V jointed shaft (3 C/V joints) ? Although more costly Greg seems to suggest the dynamics may be better. If NVH isn't an issue should I have the centre support solidly mounted? The C/V joint above, as recommended by Tractive, the gearbox makers, is a high plunge type. I was going to use these in all 3 positions, does this sound reasonable? Thanks. I have some photos of the dummied up gearbox and diff installation, plus the stock 2 whhel drive prop and a front 1/2 of a 4 WD prop I have acquired at http://www.chriswils...propshafts.html

Hookes cv cv is plenty good enough, I think, because the centre bearing doesn't move much, and the engine doesn't move much, so the hookes joint has a very easy life. The rear diff nods up and down so the rear propshaft's joints see a bit more action. Basically if hookes CV CV works, then why change it?

is 37521K a plunge joint?

Edited by Greg Locock, 16 September 2009 - 23:45.

[carlt]

I have fitted a Swedish made Tractive sequential dog box to my RWD Skyline track car, and need to now make a custom propshaft for it. the Tractive box has a C/V type rear flange on it. The stock prop is 2 piece, with a cross type front (gearbox flange) joint, and a middle and rear C/V joint. What governs whether I can make the setup a single piece prop design, with a C/V joint at each end? If I have a 2 piece made is there any potential issue with having all 3 joints of a C/V type? Thanks.

Send the front prop to a prop specialist ,[ give them the new length required , type of joint required for the new Gbox ]

Pay them some money

Bolt their new prop on to your car

Job jobbed

[Greg Locock]

Send the front prop to a prop specialist ,[ give them the new length required , type of joint required for the new Gbox ]

Pay them some money

Bolt their new prop on to your car

Job jobbed

wallet jobbed as well, for a job that largely consists of fitting accurately machined parts into accurate tubes and then gluing them together permanently. Admittedly it'll still need balancing.

[McGuire]

Send the front prop to a prop specialist ,[ give them the new length required , type of joint required for the new Gbox ]

Pay them some money

Bolt their new prop on to your car

Job jobbed

"Go buy one" is not really a technical response, is it.

[mariner]

Slightly OT but the "wallet" comment reminds me of a story my finance boss in the USA told me long ago.

He had actually qualified as an engineer under the GM technical graduate scheme along with a few other guys from Rochester NY. One of them was big into drag racing in a pretty high power stock class. He was having serious driveline breakages which were jepordising his championship bid. All his fellow GM grads were out in various parts of the empire but knew of his problem. One day a small, neat wooden crate arrived at his desk, inside is a specially fabricated drive shaft with a clever annular polymer sleeve trapped between the inner and outer tubes. Attached to it is a cryptic note "try this" and no names or addresses.

You can guess the result, no more failed driveline and one won championship, it is not what you know but who you know.

Apparently he never dared ask how much it had probably cost some GM budget in R+D money in case he felt guilty.

[mariner]

One thought on re - reading all the posts.

When I went about "designing" my propshaft whirl was a non-issue as it is about 200mm long. However one thing that seemed to be important from my research was the limits on angularity AND the importance of slightly offsetting the propshaft angles to get the bearings to move with each rotation. I presumed this was to avoid any "brinnelling" efect of always banging the bearing rollers in the same spot with each power impluse.

Is this need to make the bearings work really that important these days?

[gruntguru]

one thing that seemed to be important from my research was the limits on angularity AND the importance of slightly offsetting the propshaft angles to get the bearings to move with each rotation. I presumed this was to avoid any "brinnelling" efect of always banging the bearing rollers in the same spot with each power impluse.

Is this need to make the bearings work really that important these days?

Power impulses are usually not the problem, rather torsional vibration modes which will always exist - usually at quite low rpm - Greg will know more about this.
Re the need to make UJ bearings work - I don't think much has changed. Fatigue life will be much greater if the loads are shared around by keeping the needle rollers moving.

[carlt]

"Go buy one" is not really a technical response, is it.

Nope

But the opening post was talking about 'getting it made' ?

Sooo - technically it was ?

Edited by carlt, 18 September 2009 - 18:12.

[Greg Locock]

That's called a rubber up tube propshaft. As used in some Jaguars, and the Lotus Carlton and elsewhere no doubt. Great for damping out driveline torsionals, such as powerhop, lousy for whirl speed. Every driveline engineer should have one just for diagnostic purposes.

[Chris Wilson]

Thanks again for the help guys. I have now sourced a GTR (4 wd) propshaft as in the parts diagram earlier in this thread. By changing the metal support brackets for the centre bearing for GTS-t (2 wd) ones the rear half fits the diffand tunnel perfectly. I also grabbed a second good front section, meaning I now have two bits with a Hardy Spicer joint one end and a C/V joint the other. I have bought a new 108 mm GKN C/V and boot for the gearbox end, and will chop up the 2 front prop sections and turn off the splined end caps. A prop company can then make up a longer tube, weld on the two splined end caps, and balance the whole caboodle. It has, if it works OK, not cost as much as I feared, either. Luckily Nissan have used a GKN Motorsport sized C/V joint themselves, so it's pretty painless so far.

Thanks again. In due course I might treat it to a one piece carbon prop to reduce inertia, but this should be good for now