Shrink fit (thread on part)

[Mannes Karsten]

Hi,

Trying to figure out how much thread will deform during shrink fit. We're reverse engineering shaft sleeves for a compressor rotor and 2 of the sleeves have thread on them.
Bore diam. is around ∅110mm - thread is 4 1/2" - 12 UN (external)

Will the thread be viable if we were to design it to be machined before being shrink fitted onto the shaft? Or would we have to place the rotor in a lathe after the components have been fitted, and then machine the thread?


Thank you.

 

[EdStainless]

What is the calculated interference of your fit?
If the OD thread ends up that much larger will it still be in tol?
In the past with modest shrink fits we have pre-machined by assuming that we will see half of the interference as growth. If the tol of the thread is greater than the interference then this is safe.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[racookpe1978]

DO NOT SHRINK FIT THREADS.

I've machined and installed a number of field-fitted shrink-fit valve inserts, valve seats, coupling liners, shaft coupling sleeves, etc for valves from 60 inch down to 2.125 coupling sleeve diameters. The installation process is tricky, uncertain, and prone to slips and problems - even with a good crew under as-best conditions possible.

The installation requires super-cooling (with solid CO2 pellets or liquid nitrogen) the "too large part" - and sometimes heating the "too-small" parts as well! - then PAIDLY and SMOOTHLY sliding the two overlapping parts over one another BEFORE either gets too wamr/too cool to slide anymore. You need a 0.002 to 0.004 clearance on diameter to slide on, you will have 1-2 minutes to pull the pieces from the chill box, lift them to position, fit them on the shaft or in the valve, then slide them into their FINAL position accurately.

That 1-2 minute operation cannot realistically be done with threaded machined pieces.
Cross-thread the parts? What do you do?
How do you know you've started the threads accurately?
Can't get them up and aligned perfectly in 30 seconds or less?
Gall or too tight a fit? YOU CANNOT TEST FIT BEFORE HAND.

You cannot use lubricant or either the heated nor super-chilled threaded surface - there is no time to apply the lubricant before the delta-T goes away.

Even fitting a key into the slot during the interference fit is tricky, prone to frost-burns on the fingers inserting and trying to hold the key in place.

Now, on the other end, how will you get them apart later?
How will you get the pieces apart WHEN you fail trying to install them, and they heat up and lock into position while only partially threaded ?

 

[Tmoose]

RAcookpe has interpreted the question to mean the threaded surfaces are involved in the shrink fit directly ( I think).

(I think) EDstainless seems to believe the threads are external to the sleeve, and the concern is what happens to them dimensionally when the sleeve is expanded by the shrink fit installation.

https://media.tenor.com/videos/2e86f9c87274e1c57ef080fc3794dd9c/mp4

 

[EdStainless]

Thank you for translation Moose

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[racookpe1978]

If the threads were external to the sleeve (or the shrink fit part of the sleeve or insert), then my experience indicates that the two metals (one cooler, one heated or at environment/room temperature) exchange enough heat to become jammed within 1-2 minutes of touching each other. Then, depending on mass of the two, delta T between the two, and amount of the area of contact, 30 minutes or so to more or less become even in temperature.

Thus, if I were inserting (for example) an interference fit tapered pin with an external straight threaded section for a locknut, I would expect to be able to put the locknut on within 30 minutes to an hour after insertion of the interference fit taped pin into the tapered and reamed hole. If the threaded part were an internal thread with very tight tolerances inside a shrink-to-fit sleeve, the sleeve would also become reasonably equal to the outside part temperature within 15 - 30 minutes of insertion (less thermal mass of the shrunk sleeve, good contact area all around the outside of the sleeve, large thermal mass of the outer and heavier part. The 0.002 to 0.004 shrink overlap will almost certainly not affect the normal Class B, Class C male thread-female thread gap.

The shrinkage factor around a internal threaded shrunk-fit part "might" affect a very precisely threaded fit for certain parts - But even 1200 MegaWatt steam turbine coupling bolt fit requires a minimum 0.001 to 0.002 clearance. And that's for the smooth, lubricated slip-in fit bolts, not for a threaded fit.

Go up to a 0.010 or 0.020 shrunk fit overlap? Different.