Removing bearing housing

[Roger HulettParticipant @rogerhulett83124]

Any tips on removing a steel bearing housing from a rather fragile aluminium casing ??? Thanks

[Roger HulettParticipant @rogerhulett83124]

[michael darbyParticipant @michaeldarby61557]

Could do with a photo to establish requirements.

[Clive FosterParticipant @clivefoster55965]

Dropping the whole thing in boiling water for a while to get everything really evenly heated then shaking the steel part out has worked for me.

If you think it might want little more effort make an expanding handle style thing to grab inside the steel. I have a section of Rawlstuds (expanding concrete anchors) about the place for this sort of thing which I use with a simple slit alloy drum made a nice fit in the part to be removed. The Rawlstud expands the alloy drum to grip the part.

Clive

[Tim StevensParticipant @timstevens64731]

Do be sure before you start that the steel tube is not cast into the light alloy component. If it is, you are likely to destroy both parts before they separate.

This is a good example of 'If only I knew more I could help more' …

Cheers, Tim

[martyn nutlandParticipant @martynnutland79495]

Hello Roger

This is a very common situation amongst Austin Seven restorers (my own enthusiasm).

As Michael says a picture of the part might help, but as Clive intimates, heat is the one essential ingredient.

Depending upon what you are trying to do and its structure (which is why a photo might help) you may be able to make an extractor that will pull or push the bearing out. But even then, with old or fragile aluminium you need to heat things up. This, for example, is how one can remove Austin Seven main bearings from the aluminium crankcase.

Alternatively, if you can access the bearing with a brass drift – i.e. if it is not against a flange or lip – you may be able to knock it out. But again, heat is essential. Don't bash it cold. Old-fashioned brass stair rod may be suitable for the drift.

Try to heat the part to around 80°C and aim for generalized heat not local. A domestic oven is ideal for the purpose. In the Austin Seven context their steel hubs will give up their steel bearings quite readily when heated and you will probably be surprised how aluminum, once expanded by heat, will let go a steel bearing.

Good luck, just don't bash it cold!

Martyn

[mechman48Participant @mechman48]

Tim

As an aside, & no insult intended to either party, your avatar, uncanny… Jeremy Corbyn…?

George

[Gordon WParticipant @gordonw]

If the steel housing is no longer required, and it is intended to come out – get a bit of scrap steel that will fit inside the sleeve and stick weld ( or MIG ) steel to housing. This shrinks the housing and heats the ally, and provides something to hit or pull.

[Roger HulettParticipant @rogerhulett83124]

Herewith photo of bearing housing. I did try the boiling water method,but to no avail. The dimensions are 60mm x 60mm x 35mm. It is from a 1919 motorcycle and neither part can be destroyed.

[michael darbyParticipant @michaeldarby61557]

I would suggest that on something so small, it would have been cast in. can you not bore it out and make a press in end,then pin it in a couple of places to secure it?

[John Stevenson 1Participant @johnstevenson1]

So it begs the question, if both parts are good hence not being destroyed then why does it want splitting ?

If, and I'm guessing here the steel bearing housing is worn then it doesn't matter destroying it and just bore it out ??

[Tim StevensParticipant @timstevens64731]

This looks to me like the end casing of a magneto. If so, the steel bearing 'housing' may well be the outer race of the bearing itself. It will have a groove where the balls run, and there are pullers which expand into this groove so that the two can be separated. One way to do this without a 'proper' puller is to re-fit the ring of balls in its cage, and push a gently tapered steel bar into the hole (from below in your picture). This will force the balls into the groove and should give enough purchase to tap the outer race out. Don't go mad, though, as you can split the race and destroy the housing if you go too far. Old light alloy can be very much weaker than when it was made, after 100 years …

Alternatively it should be possible to make a puller rather like a 3 or 4 jaw chuck to fit snugly into the bearing, and with a sharp ridge around the extreme edge of the jaws which will locate behind the radius on the inside (lower as shown) edge of the bearing. There is not much room there, so it needs careful making, but you can be sure that there really is a radius in there. The photo is too blurred to show this.

Hope this helps

Cheers, Tim

[Tim StevensParticipant @timstevens64731]

Perhaps it would help if I explain that 'magneto bearings' are a special type of angular contact bearing, in which the inner race has a cage full of balls which normally stay assembled. The outer race (ie the bit we are interested in) comes away with the end of the magneto, rather in the same way that a tapered roller bearing comes apart.

This may help those of our colleagues who have never rebuilt a magneto in the dark, at the roadside, in the snow.

Cheers, Tim

For a picture, see:

**LINK**

[Gordon WParticipant @gordonw]

I've re-built more than one mag in wet fields etc. I'm still assuming that the bit to be removed is the small dia. part in the middle, if so then welding a bit of steel in is the easiest way, or any of the heat and tap methods. If it is in good condition why remove it ?

[colin hawesParticipant @colinhawes85982]

I would try heating it to around 200 deg C to see if that would release the bearing before using force on it. Colin

[Roger HulettParticipant @rogerhulett83124]

It is in fact the end casing of a Runbaken magneto. The reason I need to extract it is so that I can cast a new end casing for another mag, I could of course cast it whole as it is and then machine out the surplus aluminium to fit the new bearing housing,but I had hoped that someone might have an idea that would save the extra work. I am only a novice at engineering work and if I were to stuff it up it means starting the whole casting process again.

[duncan webster 1Participant @duncanwebster1]

Do I understand you're going to use the original as a pattern. If so the new casting will be about 2% too small overall, and have no machining allowances. For a one off I'd be tempted to machine it from solid rather than messing about with pattern making.

[Swarf MakerParticipant @swarfmaker85383]

On later magnetos the outer races of the bearings were insulated from the casing with dense insulating paper sleeves to avoid re-circulating earth currents being passed through the bearings and causing arc erosion. It may well be the case that this magneto has the same interface, in which case the differential heating method will not work – the paper grips both surfaces too well. Also, too high a heat and the insulator gets destroyed! I've sent you a personal message.

Edited By Swarf Maker on 05/02/2016 14:43:54

Edited By Swarf Maker on 05/02/2016 14:44:52

[Speedy Builder5Participant @speedybuilder5]

Just a bit of advice, if you are 'servicing' a magneto, do not leave the rotor out of the field magnets too long, as it acts as a 'keep' for the magneto. If left out for a number of months, you will start to loose magnetism in the fixed magnets.

[Jonathan MilnerParticipant @jonathanmilner22148]

Try heating up in the oven, appox 200C ,like Colin suggests, and then slap it down firmly on the bread board,so the housing drops out.Good luck.Another way would be to drill a couple of small holes through the casing to the bearing housing and try punching or pressing out,but again using heat.Hope this helps

[Roger HulettParticipant @rogerhulett83124]

THANKYOU to all who contributed. The suggestions were most useful.Success was achieved by repeated heating to 200c and then placing on a steel rod with the same diameter as the bearing.The contraction of the bearing soon released it from the housing. There was some type of insulation between the two parts. No damage to either part.

[Chris Evans 6Participant @chrisevans6]

Not sure where you are based but I use a good and helpful aluminium foundry up here in the West Midlands. It is possible to "Wax " the original component to allow for some of the contraction. I would still as suggested machine from solid in a good spec aluminium though.

[Author]

Posts