WELDING A BEARING

Gentlemen, there have been a couple of questions on bearing removal ! Here's a trick. Faced with the outer race of a bearing in a blind hole and no way to get behind it. CAREFULLY run a ring of weld around the track - and let it cool off glowing and then turn the holder over - it will fall out.

Trying to get a bush out of a blind hole. Find or make a piece of bar that is a good fit in the hole and at least twice the length of the bush. Fill the hole full of grease and place bar in hole and hit hard witha hammer. Hydraulic shock will push the bush out along the bar ! Simples. Noel, leader of the blackhand gang..

Quite right Noel!

Both methods work a treat.

As the weld cools, it contracts and pulls the outer race inwards so that it falls out of the housing.

The "Bung and Grease" method acts as very effective hydraulic press!

Where needs must and all that!

Howard

I beg to differ Howard it is not so much the contraction as the shrinkage of the molten metal that "shrinks" it ! Shrinkage is huge compared to contraction. Noel.

Molten metal can shrink (thinking of castings with insufficient head of metal) but whatever a melt might shrink, it ain’t going to do anything other than shrink while molten. It is only when solid that it can produce a force and that is the contraction with temperature drop that loosens the bearing (or cylinder liner in the case of removing dry liners from engine blocks).

Doughy bread is a better ‘hydraulic‘ medium than grease in a lot of situations - it is less likely to spurt out of a poor seal between ram and bearing race. I used damp bread to remove a needle roller cage only last week. There was only a thin lip on the drawn cage to act as a seal.

Edited By not done it yet on 11/03/2021 07:53:41

Hi, OK, the way it works is; when you put a bead of weld on the track, the race gets hot, now as you know hot metal expands, but of cause it can't expand it's diameter due to the restraints of the housing it is in. So the expansion is in the only direction it can be which is the width of the race. However when it cools, it contracts and the contraction is not constrained by anything, therefore it will contract in all it's dimensions and of cause the added weld metal will also make it contract even more, hence the diameter is reduced, all be it by a small amount. If you could heat the race to a high temperature very quickly by another method and allow it to cool, the same thing would happen. You can't stop metal from expanding when is heated, as I have said before elsewhere, tighten a bar of steel in a vice as tight as you can, then heat the bar to even a dull red and allow it to cool and without anyone touching the vice, the bar will fall out of the jaws.

Regards Nick.

The main thing is that IT WORKS!

Howardi

That brings back nearly 50 years of memory. Removing the Morris Minor spigot bush from the centre of the flywheel.

Steve.

If you could not use heating, would the pipe-freezing spray used by plumbers work? Or crushed ice mixed with salt (the latter drops the mixture's temperature quite significantly)?

I know the temperature difference hence size change with these cooling methods is far lower, but it may be just enough.

Hi Nigel, I doubt you would be able to cool it fast enough before you start cooling the housing also. The weld method reduces the diameter of the bearing race on cooling permanently.

Regards Nick.

Edited By Nicholas Farr on 12/03/2021 07:39:06

I think you would have to go a loooot colder than that to get enough contraction. EG valve seat inserts in car and bike engines. They are installed by chilling with liquid nitrogen which has a temperature of about minus 300 degrees F. Then to get them out you heat the head to about 400 degrees F to expand it and hit the valve seat inserts with cold water sprayed through a special puller tool.

Salted ice will get down to maybe minus 5 degrees. But it acts so slowly, compared with nitrogen or oxy/propane torch, that the temp differential between the bearing race and the surrounding metal will be minimal. Same with installing bearings. You can heat the housing, especially aluminium motorcycle crankcases etc, with a propane torch to 400F and bearings will fall in and out almost by themselves. But freezing a bearing overnight in the fridge make no noticeable difference to the fit. There just is not enough temperature difference.

Salted ice will get down to maybe minus 5 degrees.

That is -5 degrees Fahrenheit, of course. But not for the average job in a workshop!

Using liquid nitrogen, at around -190 Celsius, to remove anything shrink fitted generally relies on the differential in coefficients of expansion. Shrink fitting with liquid nitrogen is a breeze - if one has the need and the kit to do it.

So a TIG torch run around the bearing outer, with no applied rod, will be as effective as using a stick or MIG?

I'm glad you didn't tell me that before I tried it! I've done several rear wheel bearings on Vauxhall Omegas and can tell you that if you place the bearing after a night in the freezer it goes about 2/3 of the way in(they're 47mm thick) with a single blow from a large hammer. After that it won't move without a press or forcing tool and a lot of effort.

Hi Norm, Not tried tig, but it is NOT just the temperature difference and the contraction, it is the shrinkage caused due to the change of the metals state from liquid to solid and the considerable change in dimension this causes.

The reason the bar in a vice trick works is that the bar is in comppession, once it is heated red it becomes plastic, absorbs the compressive force, cools, contacts and is now fatter but shorter, contracts and falls out !

All good fun. Noel

Glad it worked for you. It has never once worked for me. Maybe it was a very light press fit. Lowering the temp of steel by 30 degrees C in the freezer compared with raising an aluminium or even cast iron housing by 200 C is chalk and cheese. Could do the sums but its too late at night here. But even in steel/iron you will get about 6 times the thermal expansion with the torch as you will by freezing according to those temp changes.. Then there is the greater coefficient of expansion of aluminium to add to it.

But hey if it works, do it!

When the metal is all molten and not contained it can shrink as much as it likes, and it won't have any impact on the diameter of the bearing.

It is only once the outer layers of the weld have solidified enough to create a sealed system that solidification of the metal can reduce its diameter.

A temperature drop of 1100C will shrink a steel weld by about 0.7% (please someone check my maths!) which may create enough inward tension to loosen a stuck bearing when combined with the stress of temperature cycling breaking any cold welds.

An obvious test would be to weld a bearing and invert it as soon as the weld solidifies. Does it fall out straight away or does it need to cool as well?

Neil

I have a 'reliced' bass -not to everyone's taste, but the Fender 'Flea' jazz bass is a very affordable way to get a 1963-spec instrument with excellent build quality, sound and playability.

The finish is meant to replicate Flea's own bass, but some of the markings look a bit artificial and i've done some discrete retouching to make it look less so.

Instruments with a nitrocellulose finish that are subjected to temperature cycling over their lives developed a crazed of 'checked' finish.

I inflicted one of these on my bass by alternating a hot air gun (at a low setting of about 200C) and plumbers' freezing spray.

The cracks were highlighted with woodstain.

Interestingly as the hot air gun moved over treated areas, the pattern would disappear.

Not to everyone's taste, but I play this bass as much as all my others put together... if nothing else it's nice not to worry about banging it into things!

Neil

Hi Neil, whenever I've had to do this, it has to cool significantly before it will move. Occasionally I've had to go over the first bead a second time after it has cooled in cast iron housings, I guess it depends on how tight a fit it was in the first place.

Norm Norton, I haven't ever tried just running around with a TIG arc and have wondered if you did it with a fairly hot arc using a large Tungsten electrode, it would work, the idea is to get as much heat into the race in the shortest possible time.

Regards Nick.

Edited By Nicholas Farr on 12/03/2021 11:27:01

By the time I had put the electrode holder down, opened the vice and inverted the casting it just fell out ! It was in no way cold. To be honest I could not believe what had happened, I had 6 to do and they all behaved in this way. A job I had thought might take all afternoon was done in 15 mins. Noel

The effect is small, but obvious. It's very short lived, which I suspect is due to the small temperature difference. But then I doubt that dunking a double row bearing in liquid nitrogen will do it any good....