Once more, I posted hastily. Wikipedia to my rescue! (Double Cardan-joint shaft, etc.)
So... let's cut to the chase. The pickup truck I'm putting together would have the rear u-joint running at about four degrees but the front joint at about one degree in the wrong direction. Would this create noticeable vibration?

Can you post a diagram, in plan and side view?

Roughly speaking those angles won't kill the transmission/driveline, but may cause booms. Bear in mind that on a live rear axle setup the rear UJ angle is a compromise for the average torque.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies  http://eng-tips.com/market.cfm?

You are in for trouble.  The U-joint planes need to be parallel or start there at least.  If you do the math with respect to the angular acceleration and deceleration during rotation, you will see why.

Get your trans or your rear end shimmed.

rmw

That rather depends on whether he means the diff is nose down.

Hence the request for a diagram.

 

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies  http://eng-tips.com/market.cfm?

All shaft centerlines lie in a single vertical plane. The angles stated are as seen in a side view. The transmission output shaft angles down five degrees at the rear. The driveshaft angles down four degrees at the rear. The differential pinion shaft sits level.
My understanding is that the joints should be out-of-phase by 90 degrees and their run-angles should be equal and opposite, to effectively cancel velocity fluctuations. I trust that this is true, and I'll probably change the wedges at the rear axle housing mounts to accomplish this. I'm just curious how bad the situation would be the way it sits right now- noticeable vibration? Self-destruction of bearings? Noisy?
Regarding the rear angle being a compromise: it's less of a concern than with most leaf-sprung rears, since I've built a 48" long torque arm to react differential housing torque to the frame.

UJs will happily run at 10 or 15 degrees, it's the other bits that don't like the resulting vibrations, and it does reduce their life.

eg http://www.universal-joint.com/sae.php?length_style=Fixed%20Length

 

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies  http://eng-tips.com/market.cfm?

Nose-up your diff so that the differential input shaft is 5 degrees and thus absolutely parallel with the transmission output shaft in all planes with the suspension at nominal ride height. The way you have it now is going to be trouble, and it will be even worse when the suspension compresses.

Hopefully your torque-arm rear suspension is stiff enough so that the axle as a whole doesn't wind up significantly under acceleration, and hopefully you've placed the side-view instant-centre such that the anti-squat geometry cancels out the weight transfer, thus the rear suspension will remain at or close to nominal ride height under acceleration.

Torque-arm rear suspensions (and all other link-type suspensions with non-parallel upper and lower arms in side view, such that they have some anti-squat in them) will always rotate the axle housing with suspension movement, thus sending the U-joint angles out of alignment, but at least if you have it aligned at nominal ride height, the effect will be minimal during normal driving and the amount it goes out of kilter will be equal and opposite (and kept to a minimum) with up or down movement away from nominal ride height.

BrianPetersen has given you the right recipe.  While he accurately describes the effect of torque with respect to axle wind up, that usually happens at low speed when taking off in lower gears and applying a lot of torque.

A mis-aligned U-joint probably isn't too unhappy under those conditions.  It is at higher speeds - road speeds, cruising speeds - that you will see the effects of U-joints being out of line (out of angle match).  Not too much torque then, but lots of revs and lots of potential for vibes - bad vibes.

rmw

This thread has just explained a problem I had with a 1989 Australian model EA Falcon.

It had weak rear lower control arm bushes and I beat on it a lot spinning wheels towing boats up ramps and tramping the axles with intermittent wheel spin on corrugated roads. It regularly chopped the rubber out of the front bush on the rear longitudinal lower control arms, allowing the back uni to hit the top of the tunnel on hard acceleration. It developed high speed drive line vibrations that no attempts at balance could fix.

At one time the clutch pressure plate bolts backed out enough to lock the engine (I thought it threw a rod by the sound it made).

Shortly after that it destroyed a gear box in spectacular fashion. This was at 13 months old after repeated unresolved complaints to the dealer of these vibrations.

The new gear box fixed everything, but maybe they fitted harder bushes at the same time. hmmmm. I always blamed it on gear box lay shaft bearing clearance/misalignment problems.

Regards
Pat
See FAQ731-376: Eng-Tips.com Forum Policies for tips on use of eng-tips by professional engineers &
http://eng-tips.com/market.cfm
for site rules
 

In the absence of any firm data I'd go for diff nose down by 2 degrees with respect to the trans output shaft, at your intended ride height.

There are two separate effects we need to compensate for - pitch of the axle due to vertical suspension travel, and rotation due to torque.

These are controlled by different attributes of the suspension architecture, and as such any setting is a compromise.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies  http://eng-tips.com/market.cfm?

Thanks for all responses.

Quote:

Regarding the rear angle being a compromise: it's less of a concern than with most leaf-sprung rears, since I've built a 48" long torque arm to react differential housing torque to the frame.

So now you've got two definitions for the side view IC, one that's based on the torque arm plus the leaf springs as links and a different one that's based on the leaf spring by itself.  Your effective SVIC and the actual pinion angle variation will end up being based on something in between, and the stiffnesses of the springs, TA, and the various bushings will induce a bit of ride stiffness due to the geometry conflict being "resolved" via compliances and (hopefully elastic) bending.

I am assuming that your front spring eyes are located considerably less than 48" in front of the axle and that your torque arm is not pin-connected on the chassis side.


Norm

Quote:

My understanding is that the joints should be out-of-phase by 90 degrees and their run-angles should be equal and opposite, to effectively cancel velocity fluctuations.

Every drive-shaft I've played with had the joints in-phase, or the yokes on each end of the drive shaft lined-up. I'm thinking you'd get a double velocity fluctuation at the pinion if you turned the yokes 90 degrees relative to each other, which would be a very bad thing.

 

Norm- the orientation of my axle housing is solely determined by the torque arm and causes no "conflict" anywhere (short pivoted vertical front link to frame).

Lionel- yokes lined-up on driveshaft DOES cause the rear joint to be driven 90 degrees out-of-phase with front joint.

Quote:

the orientation of my axle housing is solely determined by the torque arm and causes no "conflict" anywhere (short pivoted vertical front link to frame).

You've solved the TA 'plunge' issue, but as long as the leaves are clamped to the axle, the effective SVSA of the leaves by themselves is still unlikely to coincide with the TA + leaves-as-links SVSA.


Norm

You're correct, Norm, that the two swing arm views do not coincide. However:

The effective radius resulting from clamping to the springs is approximately infinite (leaf packs are symmetrical ahead-of and behind the clamps).

I tend to think of the 52" radius of the torque arm (front Heim joint to axle centerline) as "almost infinite" in light of the limited travel range of the rear axle during "normal" driving- maybe 5"?

Thus, the lack of coincidence causes merely a small increase in spring rate.

Thanks for listening and for helping.

Just a thought, but the fact that the rear eye of the spring is shackle-supported probably upsets the symmetry significantly.  Millikens' RCVD has a leaf spring side view sketch in it somewhere.


Norm

Hunt around for the SAE 3 link approximation for a leaf spring. It is amazingly accurate. Basically you model the leaf as 3 rigid links connected by torsional bushes. The location of the bushes is the trick.

This means that to a large degree the leaf spring behvaes much more like a trailing arm than anything else.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies  http://eng-tips.com/market.cfm?

I hear what pontiacjack is saying and I cannot disagree more.  Back in my drag racing days I had similar setup in a gasser and, over the months, it managed to crack or break something in the suspension almost every event.

It's been too many years for me to remember exact measurements, but with almost every leaf sprung, live axled race car of the last 40 years,  I have used no tq reaction devices save a short (relatively), stiff front spring section with a solid bushing with a longish rear section on compliant shackles or roller sliders along with a Panhard bar or Watts link.  U joints are set parallel and in phase. No extra links. Axle wind up on acceleration is not the big problem but, axle hop on heavy braking can be.  All in all, I am not a proponent of adding any additional axle control devices to a leaf sprung live axle.  

I know of many who do not like my approach, but I have an awful lot of wins and lap records to make me stick with what works for me.

Rod

You used a panhard or Watts link together with a leaf-sprung axle? That's an additional axle control device that can fight with the lateral location given by the leaf springs ...

http://www.ponysite.de/transam212777.htm
http://www.ponysite.de/212777_wattslinkage.jpg

page 16 here, allegedly
http://www.shelbytransam.com/walt/ShelbyTA/1970_TA_Mustang_Suspension.pdf

allegedly a 1969 trans am car with narrowde leafs and a watts linkage

http://1.photos.ebizautos.com/used-1969-shelby-trans_am_mustang-boss302-9423-5421166-21-640.jpg
http://1.photos.ebizautos.com/used-1969-shelby-trans_am_mustang-boss302-9423-5421166-25-640.jpg

Quote:

You used a panhard or Watts link together with a leaf-sprung axle? That's an additional axle control device that can fight with the lateral location given by the leaf springs ...

I think the trick is setting the PHB or WL at the proper height.


Norm

Brian, in our current Lotus Cortina vintage racer the PHB works out well enough.  If care is not taken in the location (with limited adjustability)  some adverse axle movement can be expected.  In roadracing, suspension movement is rather limited and lateral movement is also quite small.  Yes, I have several competitors that think the way you do (some professional teams also agree with you).  All I can say is that my way works for us.  I have had quite a few drivers in the seat over the years and, so far, NO complaints about the handling and drivability of the suspension setup. PHB aside, the big bugaboo with the leaf spring/live axle was "roll oversteer" which was more difficult to overcome.

Tmoose, I've competed against Bud Moore cars in vintage racing, including one Palm Springs Grand Prix in the late 80's where they were driven by Parnelli Jones and George Follmer. PJ broke and at the checker I could see Follmer's Mustang coming up to lap me. Good stuff!

Rod

• http://files.engineering.com/getfile.aspx?folder=d7d0930a-4adf-4035-8e1e-4d

Well, that did not work out all that well.  I clicked on the wrong file.  This is the one I wanted to show.

Sorry 'bout that.

Rod

• http://files.engineering.com/getfile.aspx?folder=c9ceee76-ef97-4384-bee6-c3

Well... thanks for the help understanding Cardan joints. I trust that my truck will work out just fine.

Internet searches let me down, however, on a description of ball-and-trunnion style joints, such as used in my '55 Plymouth (and I'm too lazy to work out the math myself). Anybody know (or direct me to) the expression for output velocity vs. run-angle for such a joint (at constant input velocity)?

Lokks like a tripod joint to me. For all practical purposes they are true CV joints (they aren't but it is not far off).

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies  http://eng-tips.com/market.cfm?

Ah, I see they only have two rollers. As such you will get a strong 2nd order component.  

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies  http://eng-tips.com/market.cfm?

Thanks, Greg. Is that "strong" component in the same ballpark as a Cardan joint? The output rotational velocity of a Cardan joint at a run-angle of five degrees has a maximum second-order variation from .996 to 1.003 of the constant input velocity. Is a ball-and-trunnion variation even "stronger" than this at five degrees?

Hmm, I think it would be the same strength (that is variation in angular velocity for a given offset). That doesn't feel right.

 

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies  http://eng-tips.com/market.cfm?

What doesn't "feel right"?

The vector diagram in my head says that the ball and trunnion would have the same peak/average speed as a Hookes joint. But I think that may be wrong.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies  http://eng-tips.com/market.cfm?

Be careful with Cardan joints.

See the attached paper. It outlines possible instabilities.
Other than that you should be ok.

Fe

• http://files.engineering.com/getfile.aspx?folder=08be98f6-827e-4d4f-8d79-09

Fex- thanks for the link. I won't pretend to having understood the whole paper. But unless I got it completely wrong, it appears that all those horrible instabilities occur at U-joint operating angles greater than fourteen degrees. My U-joints (as do most automotive applications) don't run at more than about five degrees.

You're welcome.
Yes you should have no problem. It's more of something to look out for, and to gain some technical knowledge of the subject.

 

Fe