Thread: Automatic Part Centering for Turning Operations on Lathe

Hello,

We have a large, cylindrical part that we need to center automatically in a lathe for a turning operation. We would like to automate the loading, centering and unloadin of this part.

Since there are tight tolerances on this particular part, it needs to be centered near zero. Operators typically indicate runout to within 0.0002" to 0.0003". They do this by chucking the part, then loosening the plate holding the chuck to table of the machine. The operator then places a dial indicator on the part face, rotates the part to find the high spot, then whacks the plate and repeats until the runout is close to zero. At which point, the screws are tightened and the part is centered.

My question for this forum is we would like to automate this process. Are there any methods available to automatically center the part to near zero using a special chucking or moveable plate type system?

It would help if we knew the size and length of part and if the od is round to start with. W&S turrets had a slick tool with 3 rollers that would act like a steady rest with a centerdrill built in. There was a knob on the tool that you could finger tighten the rollers to the od of the part.
Jim

Yeah, more info on the part. A collet should easily hold that tolerance, but you said large, so that is probably out.

I would think a good chuck with soft jaws could hold .0005 runout.

More info on the part, per the others. Agree a collet would be the #1 way to go. You can make your own collet and housing if need be.

I guess large is relative. We are talking about a part that is approximately 50" in diameter, to be clamped on diameters below 30" in various stages (the part is machined on one side, flipped, machine on the other, flipped, etc.). This is in fact a large integrally bladed rotor for an aircraft engine and, because it has been optimized for weight, has relatively thin walled sections where it is to be clamped.

How long is part and how much weight ?
Jim

Originally Posted by Red X

I guess large is relative. We are talking about a part that is approximately 50" in diameter, to be clamped on diameters below 30" in various stages (the part is machined on one side, flipped, machine on the other, flipped, etc.). This is in fact a large integrally bladed rotor for an aircraft engine and, because it has been optimized for weight, has relatively thin walled sections where it is to be clamped.

.
you are indicating a 50" diameter part to within 0.0003" ??? (not 0.003" ???) at 0.0003" the slightest dirt or chips that get in a chuck scroll between chuck jaws and scroll will cause run out. With a 4 jaw independent jaw chuck each chuck jaw can be adjusted individually or loosen back one slightly and push on forward or opposite one. Big problem I see is a heavy 50" diameter part needing to be held tight to back of chuck or it will want to fall out or get at an angle.
...... Is this a horizontal lathe (regular type lathe) or a vertical lathe?

Originally Posted by DMF_TomB

.
you are indicating a 50" diameter part to within 0.0003" ??? (not 0.003" ???) at 0.0003" the slightest dirt or chips that get in a chuck scroll between chuck jaws and scroll will cause run out. With a 4 jaw independent jaw chuck each chuck jaw can be adjusted individually or loosen back one slightly and push on forward or opposite one. Big problem I see is a heavy 50" diameter part needing to be held tight to back of chuck or it will want to fall out or get at an angle.
...... Is this a horizontal lathe (regular type lathe) or a vertical lathe?

This is surely a VTL. A horizontal lathe that can swing 50" would be massive.

What about a dedicated fixture you could clamp in the chuck? Depending on what is being done, clamping to a face plate could be more effective than a chuck. We use a few such fixtures. They are great for flexible parts.

The fixture would need to be indicated once. Then the parts should repeat.

Yes, you are correct, this is a VTL. Also, yes we are careful to completely blast the surface of the jaws and part with compressed air to remove chips prior to turning.

For its size, the part is not that heavy. The part starts with a weight of around 500lb and is reduced to around half of this weight. Since this is a rotor that will be spun at very high speed in the engine and will be flipped back and forth to machine the front and back faces, we do not want to take chances with the accumulated stackup of different runouts. Hence we set it near zero each time. This is possible with an operator, however we want to automate this.

My concern is that if we leave it up to a chuck and it is off center, then there is no way to bring it to center. I suppose I am asking, is there an off-the shelf moveable backing plate that we can buy to shift the chuck to zero with a robot or electronic actuation system? We can design something custom, but I would rather not waste time designing something if there is an existing system out there.

We asked a high-end workholding company about this and they told me that I had been watching too much Star Trek

I don't know about a way to adjust the part after clamping. I think the only hope is to get the TIR down in the chuck itself.

Diaphragm chucks can achieve >.0001 TIR. I don't know if one can be built for such a large part, but I suppose anything is possible.

Originally Posted by ewlsey

I don't know about a way to adjust the part after clamping. I think the only hope is to get the TIR down in the chuck itself.

Diaphragm chucks can achieve >.0001 TIR. I don't know if one can be built for such a large part, but I suppose anything is possible.

Thanks! Will check them out. I see Northfield Precision has a site for these. Will ask and see if they have anything that suits our needs. If you think of anything else, please let me know.

Now that we have more information available...

Yes, it is possible to do this electronically. Your process is so unique that I doubt you will find an off the shelf item for this. I'm used to dealing with this situation as off the shelf solutions for our manufacturing processes are also very rare. There are, however, some precedents for what you want to do. They are in production in measuring instruments such as form measuring systems. "Auto tilt/center tables". These machines probe the part then tilt/shift the spindle of the table to align the part to the measuring axis. Such a table is certainly a possibility for your application, but you would most likely have to develop it.

The process would be to mount the part on the table as you do now, taking care to align it somewhat close. Then probe the part and feed that information into a servo system/controller that would align the part to the machine axis. You'd need at least 3 servos, two for center, one for tilt and a clamping mechanism.

Originally Posted by Tonytn36

Now that we have more information available...

Yes, it is possible to do this electronically. Your process is so unique that I doubt you will find an off the shelf item for this. I'm used to dealing with this situation as off the shelf solutions for our manufacturing processes are also very rare. There are, however, some precedents for what you want to do. They are in production in measuring instruments such as form measuring systems. "Auto tilt/center tables". These machines probe the part then tilt/shift the spindle of the table to align the part to the measuring axis. Such a table is certainly a possibility for your application, but you would most likely have to develop it.

The process would be to mount the part on the table as you do now, taking care to align it somewhat close. Then probe the part and feed that information into a servo system/controller that would align the part to the machine axis. You'd need at least 3 servos, two for center, one for tilt and a clamping mechanism.

Hi Tonytn36,

Yes this was what I was thinking of. Chucking the near zero as possible, then designing an intermediate table / plate between the chuck and the table that can be shifted / tilted slightly through servo motors to bring it to zero. We will have probing on the machine so if we can connect the two together, then the runout can always be brought to near zero.

My concern if we rely solely on the chuck is that if there is a problem, there is no way to correct other than manually intervening.

Just wondering if they have systems like this available already. I do not want to invest time, effort and money into something like this only to find out we could have bought the thing for a fraction of the price. Since it seems like a system such as this is far into the realm of custom jobs, I think we will try and design something ourselves.

Many thanks for all of your answers. If you know of anything that can perform operations above, I would be interested in hearing about it.

Thanks