What specifically are you going to inspect on the raceway?

I am going to be looking for surface defects which are detected during manual visual inspection. These include unclean (unground) surface, rust or stain marks, rough surface etc.

Nowadays they have imaging systems that do that kind of work, it is all automated and very specialistic/expensive. However when you produce that many bearings on a line in a day then it might be worthwhile. Still, it wont beat manual inspection by experienced workers. Cant help you unfortunately with a spec on the imaging system but it might be worthwhile to get in contact with a company like Mitutoyo: http://www.mitutoyo.co.uk/Mit/mws.jsp?ProductLit.

You could use coated hard(ened) steel rollers to drive the bearings as it will be relatively low speed. Get in contact with Molykote (Dow Corning) to ask about this application: http://www.dowcorning.com/home/MolykoteHome.aspx

Alternatively you could get the rollers plated, there will be many companies in your area that could help you with that. Quality of plating should be high. Also, load you apply to drive the bearings should be low. Remember that you are inspecting before lubricating the bearings (I know I am probably stating the obvious) and that all this is done in a cleanroom environment.

You would be better served with a tri-roller setup like those used to inspect gage pins. All 3 rollers could be driven electrically or hydraulically and with 3 points of contact, only 1 roller has to move to accommodate different sizes.

You are right - the only hitch is ensuring sufficient friction between these 3 rollers and the bearing race. The driven race should not have uneven rotation.

"I am planning to design an automatic taper roller bearing outer race inspection machine."

Good plan and idea, there are many necessities to inspect if you try to obtain a very good bearing.

"and presses against the race"

Not a good idea, this will deteriorate the raceway and deposit dirt and contamination.

Missing information:

Which bearing quality are you looking at? (there are defects you can see with naked eye and have a depth or below 50nm, other defects you will not see and have out of spec of above 5µm!)

Only defects of the raceway to be detected or also radial and axial runout?

There are defects that are material related, others that are related to heat-treatment, others to the turning and grinding processes, others to vibrations, some to cleanliness and lubrication, some to remaining grinding grit in the raceway and some to the devil itself that is plaing havoc from time to time.

How many bearings per day?

The automation you can get everywhere, the bearing related specialities are much rarer and more difficult to find.

I once had a big problem with bearings being worn out after 200h of running (4 years continuous expected) that was traced to hard dendritic crystallites growing from the surface of the retainers by influence of a new chemical etching process that was going on in the same room as the retainers were stored after electropolishing. These crystallites were only some few micrometer high, like trees with branches, not easily seen in an optical microscope. But these crystallites were hard and tough so could grind balls and raceways.

I designed some specialised bearing testers and tests for precision ball bearings - not an easy task.

RHABE

"and presses against the race"

Not a good idea, this will deteriorate the raceway and deposit dirt and contamination.

// Can I use some material which will not contaminate the raceway?

Missing information:

Which bearing quality are you looking at? (there are defects you can see with naked eye and have a depth or below 50nm, other defects you will not see and have out of spec of above 5µm!)

Only defects of the raceway to be detected or also radial and axial runout?

// Only surface defects on ALL surfaces (for the outer race OD, both faces and also the taper ID)

How many bearings per day?

// About 3 per minute or 4000 per day

The automation you can get everywhere, the bearing related specialities are much rarer and more difficult to find.

I once had a big problem with bearings being worn out after 200h of running (4 years continuous expected) that was traced to hard dendritic crystallites growing from the surface of the retainers by influence of a new chemical etching process that was going on in the same room as the retainers were stored after electropolishing. These crystallites were only some few micrometer high, like trees with branches, not easily seen in an optical microscope. But these crystallites were hard and tough so could grind balls and raceways.

// Thanks for the tip

You may touch the outer circumference of the outer ring or the inner circumference of the inner ring but never touch the raceways!

The soft rubber wheels of roller-blades may be a good choice.

High friction, low contact force, no particle embedding.

Delrin should work too but as Teflon can embed hard particles and then scratch.

Think about contacting the wheels to ring at the radius between axial surface and cylindrical.

Drive either by DC-motor to one of the (3) wheels or by a belt.

We suspended the to be tested ball-bearing in airbearings.

This was necessary as torque/angle and torque/speed was to be checked.

RHABE