Scaping bearings

[IvyParticipant @ivy]

I hope this does not start an argument.

I expect people have different opinions on this. So here goes.

Is it really necessary to scrape new headstock bearings? The reason I ask is because on IC engines with plain bearing cranks the bearings are obviously not scraped and the RPM and stresses are surely much higher than my old lathe headstock.

As an example I have a Kawasaki engine that runs at 10,000 RPM all day long, the bearing sizes are available to buy and fit and forget. I have rebuilt these engines with no reliability issues, or vibrations or noises. Any runout would cause sever vibration at these RPMs.

What do you think.

Ivy.

[IvyParticipant @ivy]

[Dave HalfordParticipant @davehalford22513]

Isn't this more to do with poured white metal bearings from a 100 years ago?

[Robert ButlerParticipant @robertbutler92161]

Ivy

Fit and alignment are critical.

Robert Butler

[HOWARDTParticipant @howardt]

Scraping is done on larger bearing halves to break the surface and create pockets for some oil retention. A smooth bearing surface and a smooth shaft which are both perfectly round will not leave room for oil between the surfaces causing excess wear and heat build up.

[Tony Pratt 1Participant @tonypratt1]

I would say not unless the design calls for it, i.e. when I made & fitted a new spindle to a Myford ML7 the also new bearings shells needed scraping 'in'.

Tony

[IvyParticipant @ivy]

Posted by Dave Halford on 27/05/2022 16:15:55:

Isn't this more to do with poured white metal bearings from a 100 years ago?

Possibly, but is it still needed with modern machining?

[IvyParticipant @ivy]

Posted by Robert Butler on 27/05/2022 16:17:22:

Ivy

Fit and alignment are critical.

Robert Butler

Can you elaborate?

[IvyParticipant @ivy]

Posted by HOWARDT on 27/05/2022 16:19:43:

Scraping is done on larger bearing halves to break the surface and create pockets for some oil retention. A smooth bearing surface and a smooth shaft which are both perfectly round will not leave room for oil between the surfaces causing excess wear and heat build up.

Oil grooves retain oil.

[mgnbukParticipant @mgnbuk]

As an example I have a Kawasaki engine that runs at 10,000 RPM all day long, the bearing sizes are available to buy and fit and forget.

But if you look at the factory rebuild manual, it will tell you how to select the correct size bearing shells from a table – look up the crank & housing dimensions (often coded on the parts with coloured dots or simliar) and it will give you the correct bearing to use to get the desired clearance. Mass produced engine components are graded during production & assembly from the graded parts to get the required running clearances is easily assured.

Not so with a one-off or low volume application, hence the need to "check and adjust" by a suitably trained fitter. A true "fitter" is not just an assembly monkey – they make parts fit together as the designer intended..

Nigel B.

[IvyParticipant @ivy]

Posted by mgnbuk on 27/05/2022 17:25:56:

As an example I have a Kawasaki engine that runs at 10,000 RPM all day long, the bearing sizes are available to buy and fit and forget.

But if you look at the factory rebuild manual, it will tell you how to select the correct size bearing shells from a table – look up the crank & housing dimensions (often coded on the parts with coloured dots or simliar) and it will give you the correct bearing to use to get the desired clearance. Mass produced engine components are graded during production & assembly from the graded parts to get the required running clearances is easily assured.

Not so with a one-off or low volume application, hence the need to "check and adjust" by a suitably trained fitter. A true "fitter" is not just an assembly monkey – they make parts fit together as the designer intended..

Nigel B.

Yes, I know. But if a bearing is made to the shaft that's the same thing isn't it?

The different sized bearings you mention are different in size by "oil film thickness".

[Howard LewisParticipant @howardlewis46836]

Low speed drip fed bearings and high speed, pressure fed bearings is not am good comparison.

The IC engine bearings are fed with a copious supply of oil under pressure and easily build up a dynamic wedge of oil.

The headstock bearings of a Myford are drip fed and run at comparatively low speed, and are probably designed for much smaller clearances. (A high speed IC engine will be designed for circa 0.002 – 0.004" 0.05 – 0.1 mm clearance, to allow a good flow of oil, to provide cooling as well as lubrication).

The purpose of blueing and scraping will be to maximise the area riding on the dynamic wedge of oil; possibly boundary lubrication only a few molecules of oil thick.

A headstock with automotive clearances would be classed a badly worn machine tool..

Howard

[IvyParticipant @ivy]

Posted by Howard Lewis on 27/05/2022 17:35:29:

Low speed drip fed bearings and high speed, pressure fed bearings is not am good comparison.

The IC engine bearings are fed with a copious supply of oil under pressure and easily build up a dynamic wedge of oil.

The headstock bearings of a Myford are drip fed and run at comparatively low speed, and are probably designed for much smaller clearances. (A high speed IC engine will be designed for circa 0.002 – 0.004" 0.05 – 0.1 mm clearance, to allow a good flow of oil, to provide cooling as well as lubrication).

The purpose of blueing and scraping will be to maximise the area riding on the dynamic wedge of oil; possibly boundary lubrication only a few molecules of oil thick.

A headstock with automotive clearances would be classed a badly worn machine tool..

Howard

Thank you Howard, that's answer makes a great deal of sense.

[SillyOldDufferModerator @sillyoldduffer]

Posted by Dave Halford on 27/05/2022 16:15:55:

Isn't this more to do with poured white metal bearings from a 100 years ago?

That's my take too! White metal bearings are easy to make, but need much TLC to get the best out of them, notably making sure metal to metal contact is minimised by good fit and helping oil spread. Scraping is probably still the best way of doing this.

However, because scraping is mind boggling expensive and plain bearings have a multitude of shortcomings, clever engineers have been developing alternatives for nearly two centuries. They've had considerable success and today plain bearings are much less common than they were and scraping is rarely required. Took over 50 years for manufacturers to make ball-bearings smooth and long-lived enough to supplant plain bearings in headstocks, but it was done just before WW2. Since then ball-bearings have come to dominate. They're low maintenance, don't require skilled workers, and can be swapped when they wear out.

The alternatives to plain bearings are high-tech and mostly too difficult to make in a home workshop. I've made plain bearings that work well at moderate loads and rpm as long as they're generously oiled, and anyone who owns a lathe could do the same. Roller bearings are much harder to make well. Although simple in principle to provide long-life under fast heavy loads commercial bearings have to be precision made from tough hardened metal . The balls are closely dimensioned near perfect spheres and the ball races are closely matched to them. And the higher the load and rpm, the better made roller bearings have to be. In comparison home-made copies would be high friction and short-lived due to bump and grind, and only usable at low speed. I doubt anyone on the forum has the technology needed to make a good ball bearing from scratch.

Have a go at making a 10mm diameter sphere plus/minus 10 micrometers in any direction. It's a challenge, especially when you need 20 of them!

Dave

[old martParticipant @oldmart]

I would fit the bearings and after making sure they were not tight, just run them for a short while with plenty of lubrication. Then dismantle and look for tight areas and scrape then lightly and repeat. The design varies a great deal, from clamped and shimmed bearing halves, tapered designs which can be adjusted by axial means and also parallel surfaces which have bushes made like er collets which can be closed down by axial adjustment. Each style needs slightly different approaches. There are certainly no traces of scraping in the spindle and bushes of the Smart & Brown model A lathe. Any scraping like bedways get would be detrimental to a rotating bearing, only the rings of tighter contact require attention.

[IvyParticipant @ivy]

Posted by old mart on 27/05/2022 18:47:05:

I would fit the bearings and after making sure they were not tight, just run them for a short while with plenty of lubrication. Then dismantle and look for tight areas and scrape then lightly and repeat. The design varies a great deal, from clamped and shimmed bearing halves, tapered designs which can be adjusted by axial means and also parallel surfaces which have bushes made like er collets which can be closed down by axial adjustment. Each style needs slightly different approaches. There are certainly no traces of scraping in the spindle and bushes of the Smart & Brown model A lathe. Any scraping like bedways get would be detrimental to a rotating bearing, only the rings of tighter contact require attention.

I was thinking along the same lines as you. I also saw a chap on youtube who fitted new bearings to his Myford ML7 and said no scraping was needed, he showed the results of the change including running temperature at the bearing caps. I am confident he knows what he is doing.

I would like to get a spindle reground on the bearing journals and make new bearings for it but I am not at all confident about scraping, I've never done anything like that.

It's not a Myford, but the same arrangement as an ML7. Split, parallel bearings shimmed for clearance.

Edited By Ivy on 27/05/2022 19:11:42

[not done it yetParticipant @notdoneityet]

bearing sizes are available to buy and fit and forget.

These are denoted as precision bearing shells, not rough cast (or turned) items which are required to be dressed to the shaft – particularly where the underlying mounting may not be perfect.

You may have not noticed that your kwaka shafts/bearings are continuosly pressure fed with oil (not a simple drip feed and total loss system like the bearings with which you are comparing. Indeed the precision plain bearings in engines generally require about 5psi (~1/3Bar) to maintain a lubrication film between shaft and bearing, to avoid rapid destruction of the very thinly constructed bearing white metal deposited onto the shell backing material.

You would not be able scrape those precision shells because the white metal is only a few thousandths of an inch thick! That construction, using far less material (particularly the expensive white metal component), superseded much thicker besrings that were adjustable (to an extent) by shim removal between the bearing halves as wear took place. Precision shell bearings are not adjustable, and when worn out often require the shaft surface to be precision ground to a different undersize to accept a new set of bearings that match the shaft size precisely.

Additionally, engine crank-case bearings (particularly those with intermediate main bearings) have to be very well aligned to avoid uneven wear (or early failure). This is a standard level of precision engineering, these days, and is far better than older designs which, while being modern in their day, was far from the precision of modern equipment.

[old martParticipant @oldmart]

Ivy, if your spindle is getting reground, the bearings could be produced at the same time with bores between 0.0005" and 0.001" bigger and then cut in half afterwards. They would have to be made with the same outside dimensions as the old ones. The better the finish in the bores the less running in they would need. Because the S & B model A bearings are adjustable, (the third type in my earlier post) they can be run until the headstock only gets warm after 20 minutes running at top speed. With shims you could gradually tighten the caps until the bearings got above warm and then go back to the earlier shim thickness. The proceedure will take a lot of time and patience but your lathe will be as new in the spindle department. I would recommend one of the bronze types for the bearings. If there is provision for oil grooves the bearings will last longer. Also there will need to be provision for the thrust taken if using a centre in the tailstock or drilling.

[IvyParticipant @ivy]

Posted by not done it yet on 27/05/2022 20:21:46:

bearing sizes are available to buy and fit and forget.

These are denoted as precision bearing shells, not rough cast (or turned) items which are required to be dressed to the shaft – particularly where the underlying mounting may not be perfect.

You may have not noticed that your kwaka shafts/bearings are continuosly pressure fed with oil (not a simple drip feed and total loss system like the bearings with which you are comparing. Indeed the precision plain bearings in engines generally require about 5psi (~1/3Bar) to maintain a lubrication film between shaft and bearing, to avoid rapid destruction of the very thinly constructed bearing white metal deposited onto the shell backing material.

You would not be able scrape those precision shells because the white metal is only a few thousandths of an inch thick! That construction, using far less material (particularly the expensive white metal component), superseded much thicker besrings that were adjustable (to an extent) by shim removal between the bearing halves as wear took place. Precision shell bearings are not adjustable, and when worn out often require the shaft surface to be precision ground to a different undersize to accept a new set of bearings that match the shaft size precisely.

Additionally, engine crank-case bearings (particularly those with intermediate main bearings) have to be very well aligned to avoid uneven wear (or early failure). This is a standard level of precision engineering, these days, and is far better than older designs which, while being modern in their day, was far from the precision of modern equipment.

Yes, I understand automotive engine assembly and bearings very well.

I made a poor comparison with the bike engine.

I still don't really get why if a bearing is made to the spindle it will be used with it needs "scraping in".

[IvyParticipant @ivy]

Posted by old mart on 27/05/2022 21:00:23:

Ivy, if your spindle is getting reground, the bearings could be produced at the same time with bores between 0.0005" and 0.001" bigger and then cut in half afterwards. They would have to be made with the same outside dimensions as the old ones. The better the finish in the bores the less running in they would need. Because the S & B model A bearings are adjustable, (the third type in my earlier post) they can be run until the headstock only gets warm after 20 minutes running at top speed. With shims you could gradually tighten the caps until the bearings got above warm and then go back to the earlier shim thickness. The proceedure will take a lot of time and patience but your lathe will be as new in the spindle department. I would recommend one of the bronze types for the bearings. If there is provision for oil grooves the bearings will last longer. Also there will need to be provision for the thrust taken if using a centre in the tailstock or drilling.

Thanks Old Mart. We are on the same wavelength.

[old martParticipant @oldmart]

You might be able to make shims out of feeler gauge blades, if you can find big enough sets to cut up. It used to be common for the thinnest blade to be 0.0015" thick, which means with some juggling, you could change thicknesses by 0.0005". The thinner ones would be best used sandwiched between thicker ones. Also the shim thickness each side of the cap could vary by up to 0.002" without harming anything.

[not done it yetParticipant @notdoneityet]

I still don't really get why if a bearing is made to the spindle it will be used with it needs "scraping in".

This is an easy one. Precision and longevity. Replacement bearings may need to accommodate less than perfect shafts. Can you guarantee your shaft to the n th degree? Mass produced shafts, mass produced bearings over umpteen batches over many years of production – some may match.but likely not all. Particularly replacement bearings, where wear has obviously taken its toll on the original bearing (which may have been changed previously.

Even with automotive engines, parts are chosen to match. Pistons (weight , for balance), bores (different ‘grades’ of pistons fitted), piston protrusion (different head gasket thickness for compression and/or not contact when running), bearing shells (particularly for competition engines), valve seat inserts (for correct valve recession) just as obvious examples.

Regarding running in – some items, in the past were ‘run in’ by tightening the bearing caps to such an extent that the bearing surfaces almost melted – then clearance shims were fitted. Those bearings were most certainly ‘relieved’ on the sides to allow lubricant to be picked up as the shaft rotated. Likely lightly scraped to provide oil capture/retention, too.

These processes were clearly needed and understood by both the manufacturers and the engineers of the day. Most motor mechanics became just that – mechanics, not engineers.

Although the average engine may only need a fitter to change service items, the engineering (to allow that) has progressed to that point over many years.

How many physically diagnose a problem these days? Not many. Fitters, to simply substitute old parts with new items, are cheaper than proper engineers. Most modern vehicles use computers to diagnose faults, so technicians can just look up any ‘fault codes’ and fitters can then substitute parts required. Good, but only to a certain point.

Think here of something like F1 technology levels where sensors, transmission and data processors are of paramount importance for absolute optimal performance of the unit for its short lifespan under very arduous conditions of operation. Your lathe is unlikely to perform to the same exacting standard as a F1 engine!

[Nick WheelerParticipant @nickwheeler]

One thing to consider is that the definition of 'fitter' has changed for the worse: now it means fit a new part out of a box and hope the job doesn't come back, instead of make/modify/adjust a part(using skills developed over several years) so that it works correctly. Engineer is another overused word – the man who can't diagnose, let alone repair your boiler isn't an engineer. He doesn't fit the definition of mechanic either, so we get the deliberately vague technician

[HopperParticipant @hopper]

Surface finish and alignment.

New bearings have bored holes up the middle. Rough as guts by tribological standards. Scraping removes the high spots and promotes consistent oil film thickness.

When you have two bearings like on a lathe, the outsides as well as insides are a machined finish. They may or may not be perfectly concentric and aligned. Scraping makes sure they are.

No idea if Myford scraped on new bearings at the factory though. If they were precision made bearings they might have got away with running as is. Seems like a typical Myford cost saving. They were a low cost hobby lathe in their day.

[Bill PudneyParticipant @billpudney37759]

White metal bearings were involved a long time ago when decent ball or taper bearings hadn't been invited. The shells used for the use of engines (car, motorbike etc) require a VERY accurate and ground journal, they should not be scraped "in", they simply require that they are fit in the correct recess and have some appropriate oil. I think that Tony Vandervells concern invited them in the 50s.

Most good quality machine tools these days involve ball, roller or taper bearings, some involve ceramic balls, these are capable of very high r.p.m They can be eye watering cost, because the Careful Concern Company will use laboratory type clean room to achieve the cleanliness involved. Probably not a clean room were they could process surgery, or make electronics, but very very clean nonetheless.

It was my task to scrape in a 10" split front bearing, and a 8" back split bearing, from memory it took over a week, too get it fitted. But then it was VERY slow, maybe 100 r.p.m.

Which by the way is why "Fitters" are called "Fitters", because the make precision things"Fitted", not just use a new part in a cardboard box.

cheers

Bill

[Author]

Posts