Rear Half Shafts, CV's or U-Joints

[J.R.]

Okay, this question has been asked before, but it turned into a tripod vs. CV discussion. I am designing a rear independent suspension for comparison to our Solid Axle. We are looking at ~14 in half shafts in the rear, approximately the same length as our rear a-arm links.

Now, the benefits that I have heard for using CV's over U-joints
CV 1. Allow for large degree of plunge.
CV 2. Lighter.
CV 3. Easier to install.

And good/ bad about U-joints
UJ 1. Go out of phase.
UJ 2. Cheaper

CV1. Now, if we are running half shafts that are approximately the same length as the A-arms, and lead into the instant center, we should theoretically get ~0 plunge? Is this correct?

CV2. I havn't been able to find any figures, can anyone comment on this?

CV3. Theoretically, shouldn't I be able to attach the HS to the U-joints just as easily as with the CV joints, providing that I have very little plunge? To take up what will inevitably be a small amount of plunge, I was thinking of running a rubber insert that can flex, similar to the carbon shafts that ?Delft? was running a few years back.

UJ1. Can be taken care of by running the second one at 90 deg to the first. Does this actually completely get rid of the discontinuous rotation?

UJ2. From what I have seen it will be more economical to use U-joints and make our own HS's.


Now, this is the first actual system that I am designing for anything, and I have read a ton and learned a lot of theory, but I may have some things wrong, feel free to correct me if I have something even slightly wrong!

Danke schoen!

[Dennis Seichter]

I don't know too much about U-joints, but I have seen them on only very few cars, mainly because of the issues you already stated. I wouldn't use them if I didn't have to.

Be advised that using CV or tripod joints is not as easy at it seems. Especially take into account material wear (required hardness, material coupling etc) and the movement during suspension travel. It is also quite important how you manage to fix the joint axially.

I guess the drivetrain guys from Zürich and Stuttgart can add some information here, since they had issues similar to the ones our team faced with our tripods this year.

A lot of weight can be safed if you design the interfaces right - that is, wheel hubs and diff adapters with integrated housings or with inserts instead of bolt-on adapters.

As far as driveshaft angle goes, let me just tell you that even though relatively high angles are possible and sometimes become inevitable due to packaging problems (had that problem in our first car and ran ~6? static angle), the design judges really don't like it and it leads to increased wear. Just try to design your halfshafts to have 0? static angle from the beginning on, it will help you avoid problems. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[J.R.]

Originally posted by Dennis Seichter:
I don't know too much about U-joints, but I have seen them on only very few cars, mainly because of the issues you already stated. I wouldn't use them if I didn't have to.

I have also state proposed solutions to all of the problems, I really wanted some actual advice for WHY they might not be good, no one seems to actually know.

CV joints were invented for FWD cars, to allow for the large steering angles. Now, on a RWD car, with max +/- 1 deg toe setting, aligning the u-joints with the wheel centers, which is not interfering with packaging, why would this be a bad idea? According to some back of the envelope calc's we won't see any more than 8 deg of deflection maximum. My team is on a tight budget, so I'm looking to cut costs everywhere I can.

If the plunge problem can be solved with rubber bushings, and the out of phase can be solved by placing two joints out of phase, is there anything else to worry about?

[barba_p]

Good day. could you tell me why some teams have their half shafts placed not directly from differential to the wheel, but on some angle. Do they do it only because lack of space, or there is other reason.

[The AFX Master]

Forward inclined halfshafts have a reason, you can increase your weight distribution sligtly on the rear,also making the wheelbase shorter. We use Taylor race engineering shafts and tripods and all i can say that's well made and easy to install stuff, great support from them also.. run the shaft no more than 8~10 degrees of absolute inclination and you'll be fine.

Some judges don't like "superman arms" shafts but, we like a shorter wheelbase and a decent weight distribution, If you manage to run an entire season with that setup, then what's the problem?. I can argue that this is a race car, not a SUV with an estimated 200k mile life.

[Geoffct]

Don't forget flex discs.

I think Colorado did some dyno research on the high angle halfshafts and found <0.5% efficiency loss.

Also didn't UB's Rob Woods have a quote about live axles being the only "sensible" idea for formula? I always got a kick out of that comment.

[rjwoods77]

I never said they were the only sensible idea. I just said if you are going to run a spool then why bother with a double a-arm. I really hope that UB isnt looking away from furthur development on the live axle setup since it does have many merits but with a new crew you always get the "I can do better than they did" kids. I was one myself at one point because I saw trying to copy RIT,RMIT,UWA was the wrong direction because they will always outresource a small underfunded, undersupplied team and I think the complete right turn our car did worked out well with a 18th place finish. Actually best finish UB had had in about 11 years with something everyone said wouldnt work, included my own team members. Car doesnt handle that great but it is nothing development wouldnt solve since the car was designed on hunches and not actual engineering. Last I was told UB was researching the weight of a double a-arm setup versus a solid axle so they had a point to compare to. It seems that may have changed to a "we are doing this" since questions like this are being posted.

JR,

I could have answer all of your questions in about 15 minutes. A drive axle is much like a toe/control link in the sense of front view positioning in terms of plunge that would let you figure out yourself how much you need and why along with why a plunging u-joint half shafts suck. I also have had many hours of discussions, fsae realted and not, with Scottie at Taylor racing so I am keenly aware of the ins and outs of half shaft design and more importantly how you effectively connect them to a diff. If you are worried about economic realities of an IRS then you need to not look at it sense it is an automatic 1000-1500 dollar upgrade. Then again if it is a theoretical look it doesnt really matter since it is all fake money. But I have a feeling it isnthttp://fsae.com/groupee_common/emoticons/icon_smile.gif Mike and I cry at the possibilityhttp://fsae.com/groupee_common/emoticons/icon_frown.gif You do have my cell and I would hope you arent afraid to use it.

[Sam M]

Another point that I have heard brought up about U-joints before is that even if you can get the velocity changes to "cancel out" by having input and output of the halfshaft be parallel, the drive axle would still see 4 accelerations during one cycle.

I can't vouch for this myself, but if this is true clearly it has fatigue life implications.

[dazz]

These are pics of the drive shafts out of a Datsun (Nissan) 180B & are used in heaps of the club level autocross & rally cars here in Aus without too many issues.
http://homepages.paradise.net.nz/lghrnrm/axleass.jpeg
http://homepages.paradise.net.nz/lghrnrm/axles.jpeg

[overdrive535]

In response to Sam M:

If I remember my Ag. Power and Machines, there is an acceleration "hump" every 180 degrees with U-Joints. And this "hump" becomes larger as Z, the angle of misalignment, becomes larger.

In common speak, you get large torque spikes in your shaft, which are canceled out at the second joint if everything is configured properly.

In certain Ag. PTO applications (not that ya'll care) great care has to be given that a vehicle turn in the field doesn't blow something up further down the line.

That all said, because our change in Z with the setups we run is so small, it shouldn't be an issue. I would still do some testing and analysis to see what the rise in torque is for your maximum deflection. Then I would design with an appropriate safety factor.

One final thing to keep in mind with U-Joints is that they need to have some deflection in them during static conditions, otherwise the needle bearings will burn out.

Sorry this is going so long. I said all that to say this: U of B, stick with the solid rear axle. If you are running a four link, with a little development you can get your car to turn well. Don't give up on it just because race cars run IRS...

[VFR750R]

the great thing about CV joints is production ones can be sourced really cheap. We used to use volkswagon CV's, and although i don't know how much we worked on them (our axles and stub shafts were custom), i can't remember ever breaking them.

As far as solid axles, nascar is quite successful with them. You can even bend the axle housing to run toe and camber. And they're nearly as quick as a double a-arm car on a road course http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif

[J.R.]

Originally posted by VFR750R:
You can even bend the axle housing to run toe and camber. And they're nearly as quick as a double a-arm car on a road course http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif

I might just not be thinking clearly, but how would you bend the axle without having all kinds of crazy wheel rotations? Run a solid inner portion to a CV joint that pivots?

[SusProg3D]

The NASCAR type fully floating rear axle has a driveshaft splined at both ends. The inner end into the spool and the outer end into a drive plate.
I think that they crown the splines slightly so that there is enough clearance so that the driveshaft can run at a slight angle.
Crude, but effective.

[rjwoods77]

Crowned splines or ball drive. JR, we cannot do that on our setup because you need to have a fully floating rear end which is a car thing and not anything like we do it and a differential/spool with separate drive axles right and left. There is no way to do a cambered/toed housing with a solidaxle running from tire to tire. It is used mostly to help combat the effects of banking from tracks on tires. More importantly there really isnt a need to do so either. Our rear suspension design trades off lateral cornering grip for better longitudal grip with the benefits of weight and pacakging balancing out the the difference. If you really are interested then you need to investigate speedway engineering and/or winters quick change websites. Good for having knowledge but is is absolutly not applicable to FSAE dues to weight/packaging and again not needing to camber and toe the rear tires. There is alot of metal involved in a full floater rear end.

[VFR750R]

i don't think camber is only needed on banked tracks. I think it's actually more sensitive to it at tracks with less vertical loading (ie short tracks and road courses).

And from the stories i heard, when the first guy figured out how to put camber in the rear housings...he wore out the whole field for half a year until other teams figured out how to do it too.

Not to wear You out rob, but with 25hp how much advantage are you going to be able to make of that extra longitudinal grip? How much lighter is a solid axle with bearing carriers and 4 links over a double a-arm spool setup? 10lbs maybe? By itself the lap times won't notice. And with deflection your dynamic camber will be postive in all the corners. Good for drifting.

About the only advantage is not hurting ones head trying to design the right geometry into the chassis.

[rjwoods77]

Its more like 30hp but its supposed to be 50hphttp://fsae.com/groupee_common/emoticons/icon_smile.gif My comments come more from where the car will end up with development and not in its current configuration. Conjecture yes but the car has shown a lot of potential to do big things once the weight is out of it, power is upped a little and the handling bugs (roll gradient to high, roll bars needed. Once again anyone can say any of this about their car but the difference with ours is we know where we are and how much there is to improve and how we can get there vs seeing the "standard" car do the standard car things. I hope that makes sense because I am too tired to explain furthur. The car was never designed to dynamically outperform. It was designed to do 75 percent of it with much higher reliability, much cheaper on the cost report which makes it really possible to get a win. It was also designed to take the comps purpose to heart. Does this make it the most competitive no but it is still a guiding purpose to work with. The rear suspension weight is really a wash when compared to a d-arm but where you gain is not having to have any frame members behind the main roll hoop which drops a bunch of weight and the cost is much much less. That said there is about 15 more to lose of the rear suspension this year with some good design work (numbers already run) and about another 10lbs in frame members (numbers already run) that don't need to be there and 10lbs to lose on the flywheel. The car was a super fat 440lbs this year. 35lb weight loss without even breathing. With other work I am super positive that the car can get down to 375lbs if the planets align. That with 50hp and a cvt and a cheap cost report for the win. At that point the rear axle makes more dynamic sense especially for corner exit. Ill agree its all ass talking but I really believe in the concept and think it can do the deal in the long run. With an RMIT or UWA doing the setup I am sure they could do some amazing things.

P.S. I'll start watching NASCAR when they set the COT's on fire and race Aussie Super Touring V8s with NASCAR engines in them. It's a sin against god to race cars in NASCAR that aren't road cars modified like the good ol days. Oh yeah. DIRT is racing, asphalt if for getting therehttp://fsae.com/groupee_common/emoticons/icon_smile.gif

[suraj mathew]

this is to rjwoods... u wer sayin about how u know the more effective ways of connecting the drive shafts to the differential... wat r they?? n im jus an amateur... not as sound as u r wid the techies.. so... in simple language please... http://fsae.com/groupee_common/emoticons/icon_smile.gif

[The AFX Master]

you would get some "simple language" answers if you use "real language" instead of a Blackberry Slang

[Benn]

simple eh

http://simple.wikipedia.org/wiki/Differential