How to cool a piece for a shrink fit?

[RainbowsParticipant @rainbows]

I want to fit a 20mm shaft into 8mm thick plate and achieve a very strong joint but also don't want any warpage from welding or heating the plate up to expand it.

Only thing I was wondering is what sort of temperature is needed? Do I need to dip the shaft in liquid nitrogen? Can I stick it in a fridge?

[RainbowsParticipant @rainbows]

[Roger Provins 2Participant @rogerprovins2]

I've done this on several occasions simply by putting one bit in the freezer for an hour and then warming the other up a bit with a gas blow lamp.

Edited By Roger Provins 2 on 22/03/2016 02:12:29

[John McNamaraParticipant @johnmcnamara74883]

Dry Ice?

It sublimes at -78.5 deg C

**LINK**

You should be able to get it.

Regards
John

[Paul LousickParticipant @paullousick59116]

It depends on how strong a joint you want to achieve and how much load it is to withstand.

CRC Freeze Spray instantly cools small components. It aids in the assembly of close tolerance or interference fit configurations and is a helpful tool for many other engineering and automotive applications.

Loctite Threadlocker Red 271 is designed for the permanent locking and sealing of threaded fasteners and can be used for mounting bushes and bearings in housings.

Paul.

Edited By Paul Lousick on 22/03/2016 05:29:57

[Clive HartlandParticipant @clivehartland94829]

Some years back I had some sockets that were splitting with the torque being applied, 375 ft lb. I tried the sleeve method and turned up sleeves of what I knew to be tough steel and pressed them onto the socket with a 50 ton press, guess what they still split!. It was the square socket that split first, not the multi toothed nut end. The bolts were 7/8 UNF for holding the drive sprockets on a tank. In the end we used impact hammer sockets that were able to do the job.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Rainbows on 21/03/2016 23:34:30:

I want to fit a 20mm shaft into 8mm thick plate and achieve a very strong joint but also don't want any warpage from welding or heating the plate up to expand it.

.

I'm with Paul … Consider using Loctite

That said: For 'a very strong joint' I would normally expect to use 638

MichaelG.

Edited By Michael Gilligan on 22/03/2016 08:14:10

[Ady1Participant @ady1]

I fixed my backgear by cutting a thread inside the cog and hammering it on with a plastic drift but the cog is aluminium and has some plasticity

You are joining two solid steel parts of a relatively large size and it really really needs to work first time, especially if your plate is going a long way down the shaft

You can experiment by putting the shaft in the bottom of a freezer and the metal plate in the oven, then handle the job with something like welding gloves

Having a soft tube drift (copper pipe?) ready to be used to tap the plate into its position would be a good idea

A twitchy bottom job which will either work first time or turn into a nightmare

edit: put the shaft in a freezer bag to keep ice crystals to a minimum

Edited By Ady1 on 22/03/2016 09:24:12

[colin hawesParticipant @colinhawes85982]

In my opinion you are unlikely to get a big enough temperature difference by freezing the shaft. I would expect a difference of at least 200 degrees C . Colin

[Paul LousickParticipant @paullousick59116]

Agree with Colin, you will need a temperature difference of more than 200deg C.

Liquid Nitrogen at a temperature of -200 deg C will shrink a 4" diameter mild steel shaft by only 7.6 thou.

I intend to shrink fit a cast iron cylinder into the steam chest of my traction engine by heating the steam chest in the BBQ to about 250 deg C and freezing the cylinder in the refrigerator to -15 deg C.

Paul.

[duncan webster 1Participant @duncanwebster1]

You need surprisingly little actual temperature difference to just overcome the interference. Typical shrink fit allowance is 1 thou per inch. Coefficient of thermal expansion for steel is 13 * 10^-6/degreeC. Do the sums and you need 80C. Trouble comes when you actually start to put the hot bit and cold bit together, you get heat transferred between the 2 so the shaft gets bigger and the hub gets smaller. This is more pronounced the smaller you go. The 200C advice given above is probably in the right ballpark. If it sticks half way on you will have enormous difficulties getting it apart without scoring the shaft. However 8mm length of engagement on a 20mm shaft is unlikely to result in the plate running true unless it is up against a shoulder. I'd go with the loctite suggestion, even then have a shoulder for the plate to butt against

Edited By duncan webster on 22/03/2016 12:33:52

[Ian S CParticipant @iansc]

You only need about .001" per inch interference fit, I wonder if after leaving the item to be shrunk in the freezer over night, a good squirt with a CO2 fire extinguisher would cool it enough. Personally, I would use a shaft lock adhesive, probably Loctite, its the method used in the automotive industry.

Ian S C

[JAParticipant @ja]

What do you call a "very strong joint"? The limits for a force fit which I would expect to be strong enough are typically +0.00" to +0.53" on the hole (this is for an H7 reamer) and +1.04" to +1.37" for the shaft at 20mm diameter. Experience tells me that this is too much for a small arbor press but cold assembly could be achieved using a hammer. Obviously a suitable lead in on the shaft is required and lubrication helps. Heating the hole and cooling the shaft will always helps.

Loctite type adhesives need to fill a gap which allows the hole and shaft dimensions to be relaxed. This leads to their widespread use in manufacturing (and model engineering). However this relaxation must come with a decrease on concentricity.

JA

[JasonBModerator @jasonb]

JA should that not be 0.00053" as just over +1/2" would be very loose

Same with the shafts. 0.00104" to 0.00137"

Plumbers artic spray that is used to freeze pipes gets things quite cold.

 

 

Edited By JasonB on 23/03/2016 12:38:51

[Michael GilliganParticipant @michaelgilligan61133]

Posted by JA on 23/03/2016 11:10:49:

Loctite type adhesives need to fill a gap which allows the hole and shaft dimensions to be relaxed. This leads to their widespread use in manufacturing (and model engineering). However this relaxation must come with a decrease on concentricity.

JA

.

Sorry, JA … That one has been 'worked-to-death' already …

Loctite anaerobic retainers can work just fine with 'light press fits'

MchaelG.

[John McNamaraParticipant @johnmcnamara74883]

An alternative…..

Cut a step in the shaft making it a couple of mm oversize, a light press fit will do, it will be tightened up by peening.

Then bore the plate to fit and put a 1.5mm 45 deg chamfer on one side of the bored hole.

Press on the plate then carefully ball peen the edge of the shaft into the chamfer, with a small hammer and light taps to maintain control. It projects 2mm, giving you metal to work with.

I would do this in the lathe with the plate resting against the face of the chuck jaws, the main thing is to keep it from canting. Tidy up face of the rest of the projection in the lathe. if looks are important.

If you peened it well you will only see a fine line.
It will be a very strong joint

Regards
John

[JAParticipant @ja]

Posted by JasonB on 23/03/2016 12:27:57:

JA should that not be 0.00053" as just over +1/2" would be very loose

Same with the shafts. 0.00104" to 0.00137"

Plumbers artic spray that is used to freeze pipes gets things quite cold.

Edited By JasonB on 23/03/2016 12:38:51

My mistake at reading my own tables. Sorry.

I am well aware the the Loctite business has been done to death.

JA

[MikeParticipant @mike89748]

Not much help to Rainbows, and possibly dangerous too so I don't recommend it, but I've frozen small items by taking a small gas torch powered by a butane/propane mix, turning it upside-down and opening the valve, and directing the spray of liquid gas liquid gas at the component to be frozen. Don't know the temperature achieved, but I believe Propane boils at around -40C. Do it outside, and don't smoke!

[Tim StevensParticipant @timstevens64731]

An alternative method:

Make the fit so that a light, non-damaging force is needed to push them together at room temperature. But before you do this, bore a hole of rather more than half the diameter in the shaft, and ream to a known size. Make up a steel plug an interference fit, with a radius on the corner one end, harden & temper it. When the shaft is positioned exactly where you want it, drive the plug into the hole – perhaps with hydraulics rather than a mallet. This will expand the shaft while giving certainty on relative locations etc.

Experiment will be required to learn exactly how much interference is needed, etc. Not a new idea, used on motorcycle engine cranks but I forget which firm.

Regards, Tim

[Author]

Posts