DIY carbon fiber axles/driveshafts? Also: anyone have a broken 350z driveshaft I can have?

Robbie said:

In reply to ¯\_(ツ)_/¯ :

10,000 ft/lbs is what the shaft has to be built for. If the shaft radius is 1 inch, that means radius is .5 inches. So a joint at .5 inches needs to be able to handle 20k lbs in shear. Even with perfect 4500 psi glue, you need minimum 5 sq inches of surface area in the joint. The whole shaft may only be 8-12 inches long, so pretty quickly, your carbon fiber shaft is really just a carbon fiber sheath on a metal shaft.

Not following your math but get the same conclusion.  Seems like 10,000 ft-lb is 120,000 inch-lb.  So with a radius of 0.5 inches you have a shear force of 240,000 lb.  That means you need 53 sq inches of glue area.  With a 1 inch diameter, you need about 17" of shaft to get 53 sq inches.  And that is just on one side.  So, it's a carbon fiber sheath on a metal shaft - just as you said. 

How are you going to come up with a fitting to put a U joint on a 1" tube? It would probably be a lot easier to extract the yokes/joints from an off-the-shelf shaft, and that's probably going to be 3" at minimum. If I'm doing my math right, that means you don't even need an inch of shaft with glue on it. Or did I divide where I should have multiplied?

In reply to Keith Tanner :

I think there are two applications being discussed.  One is for an axle which would be about 1 inch and doesn't need u-joints.  It would join shafts with different splines to adapt different CVs in a swap situation.  The other is the driveshaft that would be bigger diameter and would need the u-joint.

Keith Tanner said:

How are you going to come up with a fitting to put a U joint on a 1" tube? It would probably be a lot easier to extract the yokes/joints from an off-the-shelf shaft, and that's probably going to be 3" at minimum. If I'm doing my math right, that means you don't even need an inch of shaft with glue on it. Or did I divide where I should have multiplied?

Sorry, I'm flipping back and forth in this thread between axles for the motorcycle engined Fiat build and the driveshaft for the Ford powered AMC build.

1 inch shaft comes from me being able to semi easily machine the existing axles ends I have to fit into a 1 inch tube.

The driveshaft problem would be a 3 inch thing for sure.

In reply to Titan4 :

So, I think you are right. Nothing like being off by more than an order of magnitude!

However, this is a place I don't know how to gut check. Does an axle shaft really expect to see 10k (or more) ft lbs? Just looking at splines and CV joints that seems like a helluva lot.

Titan4 said:

Robbie said:

In reply to ¯\_(ツ)_/¯ :

10,000 ft/lbs is what the shaft has to be built for. If the shaft radius is 1 inch, that means radius is .5 inches. So a joint at .5 inches needs to be able to handle 20k lbs in shear. Even with perfect 4500 psi glue, you need minimum 5 sq inches of surface area in the joint. The whole shaft may only be 8-12 inches long, so pretty quickly, your carbon fiber shaft is really just a carbon fiber sheath on a metal shaft.

Not following your math but get the same conclusion.  Seems like 10,000 ft-lb is 120,000 inch-lb.  So with a radius of 0.5 inches you have a shear force of 240,000 lb.  That means you need 53 sq inches of glue area.  With a 1 inch diameter, you need about 17" of shaft to get 53 sq inches.  And that is just on one side.  So, it's a carbon fiber sheath on a metal shaft - just as you said. 

(sad trombone)

Robbie said:

In reply to Titan4 :

So, I think you are right. Nothing like being off by more than an order of magnitude!

However, this is a place I don't know how to gut check. Does an axle shaft really expect to see 10k (or more) ft lbs? Just looking at splines and CV joints that seems like a helluva lot.

Well...

I know from experience that a 1/2" square cross section piece of hardened steel will break when exposed to about 400-500 ft-lb.

I have broken 27mm diameter axle shafts by accelerating in 2nd gear with an engine that produced, at best, 170ft-lb, with a 2.12 gear in the transmission and a 4.78 gear in the differential.  That is 1722 ft-lb at the axle, plus whatever mutiplication factor for shock loading.  But it was also on dirt.

Time for SWAG and extrapolation!

Here's a single lap shear bond line. The stress peaks at the outer edges of the resin, and is less in the center. So a long bond line does about nothing to increase the strength of the joint: You still get large stress peaks at the ends of the bond line, and the central area of the joint is unloaded... dependant on the relative stiffness of the substrates and adhesives. 

What's it all mean? If you need 120,000 in*lb of torque transfer (ultimate), and you have a 5000 psi resin (ultimate) and a generously long 1.5" bondline, you'd need about a 4" diameter tube so that you have a large enough circumference to get the amount of bond area you need. 

You could create a smaller diameter joint that can handle the torque, but you would have to get clever about it, and do some really cool stuff with tow winding... and it'd start to get a little bit experimental and perhaps a little bit more expensive than just buying a tube, and gluing it to another tube.

In reply to gencollon :

Can you elaborate on this? You're saying that the bond area would be 4*3.14*1.5 = 19 square inches - how do you come up with that? I also think you are saying that a longer bond line than about 1.5 inches doesn't really add to the strength of the joint anymore? (this would be the depth of the joint in the tube - right?)

maybe this is another way to look at ftlb on the axle shaft: A car at a constant 1G of acceleration would complete the 1/4 mile in about 9 seconds flat. .5G gets you there in just under 13 seconds. 

So say we can generate 1G max. that has to all come through the axles. Assume tires to be about 1 ft radius. if the car weights 1500 lbs, we need 1500 lbs of thrust to achieve 1G of acceleration. So that would be 1500 ftlbs through the axle. triple that for safety factor and you're at 4500 ftlbs for the axle. 

(that would really be an extra large safety factor as the tires are less than 1 ft radius and there are 2 axles, but assuming a spool you could theoretically put all the force through one). So maybe we are more like 50k in lbs instead of 120k?

I expect the biggest load (on a street car, not a rock crawler) is going to come at launch. Maximum torque multiplication and a big shock load when you drop the clutch.

Robbie said:(that would really be an extra large safety factor as the tires are less than 1 ft radius and there are 2 axles, but assuming a spool you could theoretically put all the force through one). So maybe we are more like 50k in lbs instead of 120k?

Assuming a spool, you can VERY EASILY put more force on the axle than the engine can put out.  In a tight low speed turn, one axle will be dragging, which is negative torque, which means positive torque on the other axle, over and above what the engine puts out.

This is how you can have cars running 9s and faster with a spool, but they will only break an axle while making a tight turn in Reverse in the paddock.  No tire slip such as you'd get under hard acceleration, and by whatever mechanism, you can make a tighter radius turn by backing up.  Perfect storm of axle stress.

Blast from the past. Robbie since you keep notes on everything did you happen to measure the 350z driveshaft u joint to u joint?  I have Brett's old RX8 driveshaft now but my 350Z isn't in a position to measure it's length. I'm hoping the RX8 driveshaft is shorter. 

Rockauto to the rescue.  Looks like the RX8 overall length is 9.75 inches shorter.  Looking at the two the 350z trans side is shorter than the RX8 so I might be able to get it even shorter with my plan.  But if you have that measurement I'd know for sure.

Stampie said:

Blast from the past. Robbie since you keep notes on everything did you happen to measure the 350z driveshaft u joint to u joint?  I have Brett's old RX8 driveshaft now but my 350Z isn't in a position to measure it's length. I'm hoping the RX8 driveshaft is shorter. 

Well, racec4r might still have a 350z driveshaft in his garage. Iirc though from my research, the rx8 driveshaft is significantly longer than the 350z unit.

Search eBay for "u-u" and you'll find a ton of driveshafts and their lengths. Including rx8 and 350z. Manual vs auto might have differences in length too.

In reply to Robbie :

Thanks. The only 350z was an auto but comparing that to Rockauto and the eBay RX8 u-u shows that 9.75 to be about right. Looks like I'm building my wheelbase on the length of the RX8 driveshaft.