PDA

View Full Version : fitting bearings to a shaft advice

marbles
28-11-2018, 10:51 AM
Hi,

I'm making up a rotary indexer. In the headstock build i'm using a ER32 collet shaft extender. The shaft is 32mm OD and bearings 32mm ID but they are too tight a fit at the moment to slip on.

The difference as measured by a collegue (shouted across the room) is 5000th of an inch, or 0.0127mm in my world.

Will heating the bearings allow me to get these on or should I be looking at getting the shaft turned down, or another method?

Thanks

Clive S
28-11-2018, 11:28 AM
Hi,

I'm making up a rotary indexer. In the headstock build i'm using a ER32 collet shaft extender. The shaft is 32mm OD and bearings 32mm ID but they are too tight a fit at the moment to slip on.

The difference as measured by a collegue (shouted across the room) is 5000th of an inch, or 0.0127mm in my world.

Will heating the bearings allow me to get these on or should I be looking at getting the shaft turned down, or another method?

Thanks

Try putting the shaft in the freezer overnight But have you got an extra zero in there

marbles
28-11-2018, 11:46 AM
Try putting the shaft in the freezer overnight But have you got an extra zero in there

Thanks, will do.
Highly likely I have an extra zero 😀

Neale
28-11-2018, 11:49 AM
Just done some quick sums based on rough coefficient of expansion of steel. If your diameter difference is correct, then you will need a temperature difference of around 55C to get enough expansion/shrinkage. So, bearing in oven to, say, 100C, and shaft in freezer to give you around -15C. Difference should then mean that the bearing will drop on with a very small amount to spare, but will grab quickly as the temperatures equalise.

Saying "5000th of an inch" is a bit difficult to interpret. I think you actually mean "half a thousandth of an inch" or "half a thou" or 0.0005", which is the imperial equivalent of 0,0127mm. Shows one advantage of the metric system!

Good luck,

spluppit
28-11-2018, 02:33 PM
I misread the original post and as Neal said expressing it as 5000 of an inch is misleading. That interference is fine if it's measured correctly. It needs to be pressed on squarely using tallow. Pressing things square here is the key. Put a shoulder on the end of the shaft if possible that is a slide fit on the bearing then press. This will keep the bearing square. Also use tallow if you have any.

marbles
28-11-2018, 07:31 PM
Just done some quick sums based on rough coefficient of expansion of steel. If your diameter difference is correct, then you will need a temperature difference of around 55C to get enough expansion/shrinkage. So, bearing in oven to, say, 100C, and shaft in freezer to give you around -15C. Difference should then mean that the bearing will drop on with a very small amount to spare, but will grab quickly as the temperatures equalise.

Saying "5000th of an inch" is a bit difficult to interpret. I think you actually mean "half a thousandth of an inch" or "half a thou" or 0.0005", which is the imperial equivalent of 0,0127mm. Shows one advantage of the metric system!

Good luck,

Yeh I think your right about the "half a thousandth of an inch" or "half a thou" or 0.0005" :)

marbles
28-11-2018, 07:37 PM
Put a shoulder on the end of the shaft if possible that is a slide fit on the bearing then press. This will keep the bearing square. Also use tallow if you have any.

Thanks for the advice. I'm a tad light on tallow, never had a calling for it, must be something else I could try as a sub. I'll try the freeze and shrink thing first and see how that works.

m_c
28-11-2018, 07:38 PM
Hi,
The difference as measured by a collegue (shouted across the room) is 5000th of an inch, or 0.0127mm in my world.

Are you sure he didn't mean 5 thousands? Aka 0.127mm.
In which case, it's going to be one very tight pressfit, which will most likely lock the bearing so it doesn't spin.

marbles
28-11-2018, 08:36 PM
Are you sure he didn't mean 5 thousands? Aka 0.127mm.
In which case, it's going to be one very tight pressfit, which will most likely lock the bearing so it doesn't spin.

Thinking back I think he said "5 thou". Ok, if we assume its 5 thousands with bearing expansion and freezing the shaft will that not give the spacing to get them on?

m_c
28-11-2018, 09:46 PM
Thinking back I think he said "5 thou". Ok, if we assume its 5 thousands with bearing expansion and freezing the shaft will that not give the spacing to get them on?

Not likely.
Even if you did manage to get it on, it's highly likely the bearing will seize due to the pressure expanding the inner race. I don't even think an increased clearance bearing would even be able to handle that much expansion.

Neale
29-11-2018, 07:29 AM
Yes, 5 thou is definitely too much. Half-thou is not quite as much as the usual "thou per inch" rule of thumb, but then you are heating the bearing and I'm not quite sure how hot you can reasonably make that. The other useful trick, assuming that you are fitting two identical bearings, is to turn down the central area of the shaft to give a thou or two extra clearance everywhere except the areas where the bearing rings are going to sit. That means that the bearing that has to go on furthest has less chance to grab/jam partway down. You could also make it easier by finding a reasonably close-fitting tube that fits over the shaft with a square end and use that to push the bearing (pushing on the inner race) to try to maximise the chance of it staying square.

I'm not quite sure why Spluppit recommends tallow specifically over a more conventional (and easier to find in the workshop?) grease - I would be interested to hear. That's not a negative comment - I have just not heard of that before and I'm curious.