Interference fit between steel shaft & aluminium collar.

[cmberry20]

Not sure if this is the right forum to post this in, so apologies if it's the incorrect place.

Its being 20 years+ since I've had to calculate an interference fit so I'm incredible rusty with the maths.

I have an Nickel Iron hollow shaft with a coefficient thermal expansion of 0,000008mC. The shaft has a 118,525mm/118,475mm Outer diameter with a 78,275/78,225mm internal diameter.

The shaft needs to go into a Aluminium collar, 7mm thick with a CTE of 0,0000247mC.

I'm trying to determine the size of the internal diameter of the collar to meet the following criteria:

The operating temperature of the assembly is -40 to 200C, so the aluminium collar will still need to 'grip' the shaft at those high temperatures.
During assembly, the shaft can be cooled to -15c & the collar heated to +250C.

I've did a quick calculation & I get the internal collar diameter to be approximately 0,53mm smaller than the shaft. This seems like a VERY big interference, but I know that the expansion the aluminium is 3 times that of the nickel-iron shaft.

If anyone has a spare 5 minutes & is a lot more clued up than me on thermal interferences, could you see if this figure seems about right?

Thanks.

Regards
Chis

 

[desertfox]

Hi

What's the od of the aluminium collar?
I'm not sure your fit will work because when you make the interference what are the stresses in the aluminium and steel a room temperature?
Interference fits are based on both materials remaining within their respective yield stresses, if you check the stress in the aluminium with a 0.53mm interference, will it be above the yield stress for the aluminium?

 

[cmberry20]

The outer diameter of the aluminium collar is 132.3mm.

I will try to calculate the stress, but will need to look up how to do this as its been a long time since I had to do any calculations like this.

 

[desertfox]

Hi

I did a rough calculation on just forcing the average diameter of the collar to grow by 0.5mm and the stress in the ring was around the 86.5 N/mm2 on average, before subjecting it to any additional stresses by varying the temperature.
I looked up a yield stress figure for aluminium and found it to be around 95N/mm2, which suggests to me that the joint doesn't have sufficient safety margin during its temperature cycle.

Desertfox

 

[loki3000]

i don't think the yield strength matters here a lot, unless we expect the part to be disassembled undamaged.
the question is, is there enough interference @200°C, since the parts have different expansion coefficients and is the aluminium part past the ultimate strength point @-40°C (although this doesn't matter for compression at this temperature, cracks are not good @200°C).

 

[desertfox]

I think the yield strength matters a lot in terms of keeping the mechanical integrity of the joint.
Read the paragraph just above the sub title "assumption" on this link.

http://www.fea-optimization.com/ETBX/interference_help.html

 

[Ron]

Any concern about ultimate removal? Seizing is likely.

 

[cmberry20]

^ Theres no concern about removing the shaft from the body.

Once its in there, I want to stay there. Removal should only be possible with destuction of the body.

 

[desertfox]

Hi cmberry20

Are there any external loads imposed on the joint other than the temperature cycle?