Weld strength estimation on a drive shaft. 3

[Dafilosofer]

I am designing a drive shaft for an electric car that would be facing both bending and torsional stresses at the point of attachment of my sprocket. Various books are suggesting formulas and tables for weld thickness. Am confused as to which is the safest as am new to welding engineering. Your suggestions would be most appreciated.
A sketch of the problem is attached.

 

[CoryPad]

Any chance of using friction welding so that the weld is across the full section of the shaft?

 

[Bribyk]

Your weld isn't likely to break, the shaft itself is most likely to break at the weld(s).

 

[desertfox]

Hi Dafilosofer

One method would be to treat the welds as a line and workout the Section Modulus Zw for bending and Polar moment of inertia Jw for torsional load.
From this you can find the resultant force per inch on the weld then divide the answer by the allowable force per inch of the given weld size as laid down in standards, the result should be less than one.
I attach a method of calculating the weld stresses for a hitch which I did yesterday for another post on here.
The analysis is not appropriate for your case but it should give you the idea.
Have you considered fastening the sprocket be a different method ie- Shrink fit, keyway, I mention these as when you weld your sprocket on, the heat from the weld may cause distortion of the shaft and in addition you may need to stress relieve the weld to eliminate residual stresses that will be present after the welding process,especially if you have to do a fatigue analysis for this drive shaft.

 

[Dafilosofer]

CoryPad - No, I cant. I leave in Ghana - West Africa. The common welding process is electric arc.

Bribyk - How do you mean the shaft is likely to break? Is it from heat distortion. I can provide a heat sink.

desertfox - You have been most helpful. And yes, i did consider other ways. However, I was informed that a welded joint would be a viable and lasting joint if i could figure out how to accomplish it without destroying the shaft.

Thanks guys... I really appreciate this.

 

[Dafilosofer]

CoryPad - No, I cant. I live in Ghana - West Africa. The common welding process is electric arc.

Bribyk - How do you mean the shaft is likely to break? Is it from heat distortion. I can provide a heat sink.

desertfox - You have been most helpful. And yes, i did consider other ways. However, I was informed that a welded joint would be a viable and lasting joint if i could figure out how to accomplish it without destroying the shaft.

Thanks guys... I really appreciate this.

 

[hydtools]

Does the shaft fit through a hole in the sprocket and the sprocket is then welded to the diameter of the shaft?
What are the shaft and sprocket materials?

Ted

 

[Dafilosofer]

Yes, the shaft fits through a hole in the sprocket which is then welded to the diameter of the shaft. The shaft is of the mild steel (that's whats available ) and the sprocket is nitrided steel. I hope this helps.

 

[Tmoose]

Depending on the rpm and service life of the shaft, millions of reversing stress cycles will be applied in less than a week. That >>can<< make the subtle geometry of the finished weld joint and the quality of the weld important for good fatigue resistance.

When Welding anything to anything I count on some resulting distortion. Problems I'd anticipate are Bent shafts, wobbling AND eccentric sprockets.

The nitrided surface may need to be ground away to produce a good clean weld.
"To weld a tuftrided or nitrided crankshaft, the journal or thrust needs to be underground first to remove this surface hardening. Then the weld can be applied."

http://www.crankshaftco.com/steel-crankshaft-hardening.html

 

[kslee1000]

desertfox is correct that the design stress is the resultant stress from torsion (shear) and bending (compression/tension). I think the combined shear & tension is the most critical case especially for connecting dynamic components.

Can you minimize the bending by providing more supports? Also, have you considered to provide shear keys to assist in resisting stress induced by torsion? I have no clue on behaviors of mechanical setups, just curious.

 

[Dafilosofer]

Thank you all so much. From all the possible problems i might face if i weld, i have to reconsider. I cant do anything about the bending as my design is size constrained. However, by redesigning the shaft to include a key-way, i might avoid distorting the shaft or inducing stresses which my reduce its life. I hoped to have made the sprocket and shaft a single unit though.
Thanks.

 

[hydtools]

I looked further back into my past references, esp Spotts, Design of Machine Elements. Using your loads and maximum shear theory when loads are fluctuating I calculate a factor of safety of about 1.0. The good news is that the maximum shear stress varies inversely as the shaft diameter cubed. So increasing the shaft diameter by 1.26 times (cube root of 2 or doubling the FS) the FS is about 2. Making the shaft 35mm doubles the FS. Or the cube root of whatever FS increase you want to use.
For not much change in shaft diameter you can significantly change the shaft capacity.
Be careful adding a keyway, that also adds a stress concentration factor to the shaft.
You could do both: increase the shaft diameter and use a keyway.

Ted

 

[hydtools]

Correction to my last.
FS varies directly as the cube of the diameter, not inversely.
Sorry,

Ted

 

[Dafilosofer]

thanks... Dont the books provide factors for keyways? I use the ASME compliant equation for shaft design.

 

[Bribyk]

You might want to look at some kind of taper-lock coupling. It won't change the strength of your shaft much by adding stress-raisers and would work similar to Corypad's friction-welding idea. You also have the advantage of adjusting the chain alignment if things don't fit quite right. They are also pretty cheap and readily available from most bearing vendors.

 

[desertfox]

Hi Dafilosofer

I was just having a look at your sprocket and shaft you posted at the start of the thread.
If you need to do a fatigue analysis on this shaft I should point out the the combined or principle stress on the shaft due to the bending and torsion loads is in the order of 176-177N/mm^2 which may be on the high side for your mild steel shaft when considering fatigue loading,so whichever method you choose to fasten your sprocket to the shaft it looks to me as though you might have to increase the shaft diameter.

regards

desertfox

 

[Dafilosofer]

thanks Desertfox, will revise my calculations. seems as i was assuming too much.

 

[desertfox]

hi Dafilosofer

your welcome

desertfox

 

[Dafilosofer]

I replaced the weld with a key and completed the design... my shaft diameter is now 33mm. You guys saved me some major machining time.(and material)