Interference fit for drive shaft 1

[spike]

Hello all:
I have a 4" hollow tube drive shaft. At the ends of the drive shaft there is a 2 1/2" solid journal with a keyway on the motor end of the shaft. The journal projects inside the hollow shaft about 6" and 12" beyond the end of the shaft. Two metal discs with an outside diamter matching the inside diameter of the tube and a hole in the center of the disc smaller than the journal diamter are heated in an oven. The discs are then slid down the journal shaft and allowed to cool. This forms an interference fit. The tube is also heated and then slid over the journal assembly and allowed to cool. Thus forming a shrink fit here, which is followed up with welding processes to finish the assembly.

The manufacturer now wants to eliminate the "shrink fit" on the discs and use welds to attach the discs to the journal. This would mean that the disc will have a hole larger than the journal, or a "clearance fit".

I have strong reservations about this. However, I do not have any data to support my reservations. Should we discontinue use of the shrink fit? Your comments would be apprecated.

 

[desertfox]

Hi spike

If you weld the discs to the journal there will be distortion due to the heat from welding, would that distortion make it more difficult to slide the hollow drive tube over the journal and disc assembly.It would be helpfull
if you could provide the torque and any other loads that
the assembly see's in service along with the current interference fit values, given this information we may be able to help you analyise whether you can dispense with the
interference fit or not.


regards desertfox

 

[spike]

desertfox asked for additional data. Here it is:
The torque is 6300 in-lbs.
The discs are A-36 steel. The actual outside diameter is 4.4375, the inside diameter is 2.1810. They are 1/4" thick.

The journal is hot rolled, turned, ground and polished C1018 material. The diameter is 2.1875 + 0000/-.0015.

Thanks for the help desertfox!
spike

 

[desertfox]

Hi spike

Thanks for the info, lets see if i have this correct:-

the two discs are shrunk fit onto the journal and then the tube is shrunk fit onto the two discs and this now as to transmit a torque of 6300 in lbs correct?
I can workout the torque required to spin the two discs on the journal from the info given but what is the interference between the tube and the 4.4375 dia discs?


regards desertfox

 

[BillPSU]

The shrink fits makes this assembly self fixturing before welding improving the concentricity of the tube to the shaft. I would be interested to learn how the manufacturer will retool to do this job to maintain concentricity and balance. You did not mention what function the tube has, but is concentricity critical? Have the manufacturer prove he can maintain the concentricity with his new process. He obviously wants to eliminate the machining of the disks and the tube and reduce his costs. If the assembly is machined and balanced after welding then this change may be possible but beware of the tube being welded to far out of concentricity causing thinning of the tube in the subsequent machining operation.

 

[spike]

desert fox:
The 6300 in lb torque is correct.
The tube is 4.4320 inches in diameter. However, the manufacturer will continue to shrink fit the tube onto the disc.
They want to eliminate the shrink fitting of the discs onto the journal. I am concerned with eccentricity, fatigue, etc. plus I beleive that the shrink fit is the proper way to connect the disc to the journal.
I sure do appreciate your assistance desert fox.

 

[desertfox]

Hi spike

I have calculated that a .125" fillet weld around a 2.1875"
diameter shaft can withstand a force of 1200lb/in and with the torque your transmitting ie 6300in-lb it would generate
a force in the weld of 838.2 lb/in giving a safety factor of about 1.4.The interference fit between the journal and a single disc would only transmit a torque of 3880 in-lb some
what less than you require which is maybe why you finish the existing arrangement with a welding process. The calculation for the latter is only approx because I assumed a friction factor of 0.25 and I assumed the journal and the disc material had the same modulus of elasticity ie 30x10^6 and the same poissons ratio of 0.3 please advise if these are fair assumptions as I have no data for the materials you quote. Regarding fatigue you only need to worry if the stress is cyclic and if so how much it is varying by. Regarding distortion and eccenticity if you go down the welding route then you need to weld the parts together and leave a maching allowance on so that you can stress relieve
the welded assembly and then do a final machine which would
eliminate eccentricity and distortion prior to assemblying into drive tube.

regards desertfox

 

[Tmoose]

Length of driveshaft and more info about loading might be important.
Heavy, long Process rolls are made with endshafts going thru 2 bulkheads like your disks to remove "bending" type loading.
Picture a stub welded to the bottom of a tin can. Very wiggly to the detriment of the can bottom.

I am partial to bolted flanges and welded bulkheads.