Heat Shrink Fit 2

[Phamjt]

I am currently working on analyzing the capability of using shrink fit (Most likely with induction shrink fitting) to heat up a steel hub (1045 steel) so that it can slide onto a shaft. Using the tolerances from the tables obtained in the machinery handbook for a FN5 fit I have an estimated desired deformation of 0.0045. Ideally I would like to have the hub at around 1.1255" and the shaft at 1.125".

Using the formula shown below I should be able to calculate how much I need to heat up the hub, but I was wondering for nominal length L should I use the inner or outer diameter? Also if I were to use either option it would reach over 300 F. Isn't that value above the critical value before the material properties would change? Or is it 300 C and up that I need to worry about?

d = aL(Dt)

d = Total deformation desired (in or mm)
a = Coefficient of thermal expansion (in/in °F or mm/mm °C)
L = Nominal length of the part being heated (the diameter for a cylinder) (in or mm)
Dt = Temperature difference (°F or °C)

(

http://www.inductionheating.jp/pdf/shrinkfit_calcs.pdf)

 

[racookpe1978]

I have not run your numbers, but I have used solid CO2 for that small a diameter with an interference fit of 0.002 to 0.001 Much lower temperature difference than 300 deg C!

For 2-4 thousandths interference fit, I have used liquid N2 successfully to cool the small diameter part. Again, much smaller temp difference than heating the larger diameter part that much.

 

[ESPcomposites]

The L dimension should be based on the position of the interface.

Brian

http://www.espcomposites.com

 

[Tmoose]

How big is the hub? For the bore thermal expansion to be outward the entire hub must be up to a uniform temperature.

 

[rb1957]

why not cool the shaft (as already suggested) ?

another day in paradise, or is paradise one day closer ?

 

[hydtools]

I guess I will asked the dumb question. The hub bore is 1.1255" and the shaft dia. 1.125". Is that not a loose to net fit by .0005" clearance? Not an interference force fit.

Ted

 

[Phamjt]

I guess I will asked the dumb question. The hub bore is 1.1255" and the shaft dia. 1.125". Is that not a loose to net fit by .0005" clearance? Not an interference force fit.

I may have done this wrong but when I was using the FN tables in the machinery handbook for shrink fits it showed the Largest Tolerance for the hub would be 0.0012" while the smallest tolerance for the shaft would be 0.0025". That would give me a tightest fit on top of some clearance to get a reasonable value. I may be wrong on doing things this way but my calculations are very similar as the one in the link below.

http://docplayer.net/41324995-Shigl...sign-tutorial-3-16-press-and-shrink-fits.html

The hub has an OD of 2" so it's not too big.

For now we are trying to avoid cold shrinking as the shaft is chrome coated steel so we want to avoid the possibility of cracking when it heats back up

 

[hydtools]

By the chart you reference: Hole diameter 1.125, +.0012/-.0000; Shaft diameter 1.125, +.0033/+.0025
Nominal diameter 1.1250
Minimum interference: 1.1275 - 1.1262 = .0013
Maximum interference: 1.1283 - 1.1250 = .0033

Ted

 

[tbuelna]

A diametral clearance of just .0005" at assembly is going to make things a bit tricky with a thermal fit. The instant the mating bore and shaft surfaces contact there will be rapid heat transfer, and they will seize together. You probably should use some sort of assembly fixture to keep the parts properly aligned during thermal fitting. And increase the diametral clearance at assembly a bit if possible.

 

[Phamjt]

Yeah it was a very tight fit so I would have to keep doing some calculations to see what would work.

Assuming all that works out though my main concern would be the hardness of the material. If the calculations are right the hub would be around 700F. The only way to safely assemble it would be to design a fixture as you mentioned then have someone carefully slide the shaft into the hub. Heat transfer would definitely be a problem but to retain the hardness won't we need to quench it as fast as possible?

 

[Tmoose]

Off the top of my head I'd expect a 1" Ø bore to expand several thousandths of an inch when heated to 400 F.
400 F is the tempering temperature of many steels over HRC 50, so wouldn't change the hardness much of softer steels

What is the interference fit supposed to accomplish? Transfer of torque, or resist axial forces ? Not release the shaft at 60,000 rpm?

 

[Phamjt]

Maybe my calculations are wrong in calculating the temperature?

I used

d = aL(Dt)

d = 0.045
a = 6.1x10^-6in/in °F
L = 1.125"
Dt = Temperature difference (°F or °C)

The assembly will go into a steering cylinder so I think that's axial forces?

 

[Tmoose]

http://www.engineeringtoolbox.com/linear-thermal-expansion-d_1379.html

1.125 inch heated 70 F to 400 F = .0022" diametral increase.

70 > 500F = darn close to 0.003"

 

[rb1957]

not sure the Chrome plating on the shaft will survive.

hopefully there's a step on the shaft so the hub can be quickly driven home (against the shoulder).

another day in paradise, or is paradise one day closer ?

 

[CouplingGuru]

With something that small, you are better off press fitting. The coefficient of thermal expansion isn't worth the headache especially if it is a production environment. Because now you are dealing with hot parts and they probably won't drop on easily anyway.

And with that type of fit variance you have to worry about yielding the hub that you are installing.
If it is one time creation you are better off getting the exact measurement after plating the shaft, then tolerance your hub accordingly.

It may seem counter intuitive but smaller parts are much more difficult to heat fit, because of the amount of interference require to hold and limited about of material growth from heating.

When it comes to couplings we are always here to help.

http://WWW.PSCCOUPLINGS.COM