Propshaft strength ? 1

[PGCobra]

Hi,
I have a car I use for Drag Racing.
Can someone tell me if the material I have used on previous occassions is suitable for this new application.
The material is T45 steel tubing. 50.8mm od x 2.9mm wall thickness.
The propshaft length is approx 30".
The drive train on start is subject to an instantaneous hit of power - now this could be in the order of 600 lbsft at 4,800 rpm, but this is an estimate only as it depends on the slip speed of the convertor, but is, I think, a good estimate.
However if anybody can come up with the torque figure that could twist this propshaft this would be a good start - I can then build in certain ' fudge factors ' known to exist in this form of racing.
Thanks in anticipation.

 

[tbuelna]

You need to provide much more information. The 600 ft-lbs @ 4800 rpm torque number you quote is probably the mean engine torque at the crank, right? The peak torque the driveshaft experiences is likely twice that value, further multiplied by the transmission gear ratio.

You've also stated that the driveshaft is constructed from steel tube, so I'm assuming there are yokes welded to each end. The driveshaft is probably not heat treated after welding, so these weld zones will likely be the weakest point in the structure. In order to do a proper stress analysis, you will need to provide details of the weld joint and materials used.

And finally, if you want a truly thorough analysis, you will need to check for any critical torsional frequencies in the system. These can add significantly to the stress in the drive shaft.

If all of this sounds like too much trouble, just make sure you have a very sturdy driveshaft hoop on your chassis and lots of money in your checking account!

 

[PGCobra]

tbueina, thanks for your reply.
Yes the 600lbsft torque is at the flywheel and will be applicable when setting off in 1st gear which has a 2.48:1 gear ratio. Final drive ratio is 3.25:1.
Yes the "steel tube" is no ordinary steel and is highly resistant to twisting forces. The exact given material detail is :- BS4T45 to 4T100. when anything is welded to it no special heat treatment is neccessary. Ordinary MIG steel welding rod is sufficient, I understand - but I am currently waiting for confirmation of this.
Torsional frequencies - well I have no info at all, but all I can offer is that the propshaft is Dynamically balanced at 3,000rpm when completed.
Hope this helps for the moment. I'll come back when I have more info from a specialist welding contact I know.
Thanks.

 

[bentheswift]

The torque that the driveshaft sees will be the torque generated by the engine multiplied by the first gear ratio, right? As long as the wheels can support such a torque (which they likely can't), but that is easy to guess very accurately if you know your tires, wheelbase, and cg location.

Once you know the torque applied to the driveshaft, the problem reduces to a simple beam loaded in torsion.

A quick google search reveals that BS4T45 is the equivalent of a 13xx-series manganese steel, and I'm sure you can easily find out the material properties of your specific alloy.

 

[NormPeterson]

Do those "fudge factors" specifically include some consideration of fatigue? What might last for 10 runs might well not make 100 and probably not 1000.


Norm

 

[PGCobra]

Hi Guys,
thanks for all your info.
I have just been guided to the figures/fomulation I need :-
Roarks formulae for Stress and Strain, which quotes as
2 x Torque x outer radius / 3.142 ( outer rad to the fourth power - inner rad to the forth power ).
Done the calcs and I'm just inside the Shear Stress/Torque Load for my tubing. However as Norm infers, as I am inside the Shear Stress by only 12%, I think that fatigue will definately come into it.
Now buying larger tubing.
Oh, by-the-way, 'bentheswift' the tyres have a footprint and compound formulation which ensures that the tyres will support the torque.
Thanks again.

 

[Tmoose]

Real driveshaft tubing is rolled sheet welded. The consistent wall thickness is a big plus for maintaining balance. Nothing that poor/unknowing workmanship can't overwhelm.

 

[GMIracing]

I am not quite confident in your loading inputs:
600 ft-lbs ("your" stated engine torque) x 2 (TC stall ratio) x 2.48 (trans 1st gear ratio) = >2900 ft-lbs

as for impact loading, the Torque Converter should absorb a good portion of any shock loading that is subjected to the system apon launching.

The most frequent failure mode I have seen in drive shafts is a buckling mode, and not fatigue shear (steel choice has a small effect on resistance to buckling). Mostly, impact overloaded shaft during high traction launch with a manual trans caused the shaft to buckle. Of course, you have a short driveshaft making the buckling failure mode less probable.

and yes, welding will change the material properties in the Heat affected Zone (HAZ).

 

[PGCobra]

GMIracing, thanks for your input. Now I am confused.
I recently talked to my Torque Convertor manufacturer who never mentioned anything about a Stall Ratio and not one to the factor of x2 ! This makes a hell of a difference to my final anticipated torque output to relate to my propshaft supplier. Luckly they have not yet started on manufacture.
Impact Loading was also never mentioned - all they said was that the Convertor would absorb +/- 60 HP on launch.
I really need to get a handle on shock loading figures - is there a fomulae for such ?
I have always used manual transmissions before so I am a complete newby to the world of "Auto Transmissions". My TH400 has a Transbrake fitted, so I expect a instantaneous hit of Torque to the propshaft on launch - the propshaft is a nominal 30" in length.
I understand your comments of a " buckling mode" as I have previously been on the receiving end of 3 such failures - I would describe them as being more like a screwing of the propshaft tubing, between the UJ's, similar to a corkscrew. These have been on manual transmissions. Thankfully I had fitted propshaft hoops to prevent being launched into oblivion when the corkscrewing action pulled the propshaft out from the gearbox tail-end.
By using T45 tubing, obtained from a friend of mine who works in the aerospace industry, this solved all my problems - 2250 tons/sqinch torque loading ( ie max pressure tubing takes before twisting. ).
Do you have any other design recommendations for me please, in the light of what I have just mentioned ?
One last point regarding welding - if I use T45 tubing (only obtainable in the UK I believe) then I would have used A15 Filler Wire to BS2901 part 1. If the welding is done by oxy/acet then no stress relieving would be required.
GMIracing I don't know if you are in the UK or not, but if not I know 4130 Chromoly Tubing is used a lot in the States( 1,970 Tons/sqinch Torque Loading / 41.30 Tensile / 480 Tons/sqinch Yield Strength ). Chromoly tubing also requires heat treating after welding.
Regards.
Regards.

 

[Fabrico]

A torque converter can multiply torque up to 2.5 times. The actual amount is dependant on many variables but you can count on a minimum of 2x. Whatever shock a torque converter can absorb under normal conditions will be nullified by being in a high-stall situation and using a trans-brake. Application time of a trans-brake is on par with side-stepping a very strong clutch.

All this has been beat to death for over half century and you should not have to design much of anything. Lots of places make drive lines for that purpose or can at least provide some information toward making your own.

http://www.markwilliams.com/driveshafts.aspx

 

[GMIracing]

PGCobra,

Sorry for not getting back to your questions till now, and most likely you are already past the start of your manufacturing phase.

As Fabrico said, modern Torque Converters can multiply torque many times over. I know of some companies that have production units over 3x torque multiplication at the stall point. Special aftermarket manufactures could be even higher than this. Including the transmission multiplication, this will result in a significant amount of torque increase at launch.

Given your mentioning of „Transbrake“, (and most likely Drag Slicks) I am quite certain that you have a significant impact load acting on the entire drivetrain during launch and during every shift event, but yes the hydrodynamic feature of the TC should absorb some of the spikes. It is rather difficult to say how much of a shock load you will see. (depends on Inerta of Rotating assembly and the stiffness of the componets with the Torque Converter being the most uncertain variable to predict)

Since you stated that you have been experiencing buckling as your previous failure mode, the easiest design recommendation do increase resistance to buckling is to increase your tube wall thickness or tube diameter (geometry change).

For buckling, you must change the geometry or modulus of elasticity:

According to Norton Mechanics of Materials:
Pcr=(pi^2*E*A)/Sr^2
Pcr = max load before buckling
E = modulus of elasticity (all steels are similar)
A = cross sectional area
Sr = slenderness ratio

Sr = Leff/K

Leff = length / 0.5 (for fixed-end conditions)
K = radius of gyration

K = square root of (I/A)

I = moment of inertia

Material Strength is simply not a factor of the buckling equation, and increasing it will not help you for a buckling failure mode.

 

[tbuelna]

PGCobra,

When designing drivetrain components for a recip. piston engine powered vehicle, you must design for the maximum instantaneous torque produced by the engine and NOT the mean torque value (600 lb-ft) that you quote. Depending upon the engine's number of cylinders, firing order, engine kinematics, component inertias, peak cycle pressures, etc., the instantaneous output torque at the flywheel varies throughout the crank's rotation. And the max instantaneous torque will be significantly higher than the mean torque. Unless you have some sort of dampener or torque limiting device in your driveline, every component between the flywheel and rear tires will be subject to the peak instantaneous torques (4 times per crank rev. with a 4 cycle V8).

With regards to your analysis, be sure to include all of the combined loads the shaft is really subject to: torsion, bending due to friction in the U-joints and splines, and hoop stress due to the CF loads. You should also apply a generous Kt where the shaft abruptly changes section, like at the yoke weld joints.

Good luck!