Run outs

Hi to all. Just wanted to know from opinions here what are acceptable range of run out that is ok. I know it is dependent on how satisfied you are or the nature of your work. From my point of view its hobby driven. Nothing critical to a point of perfection but of course I also would like to achieve decent work results.

My main worry is when I used my ER32 collet chuck (chuck and backplate assembly). Likely this is a Chinese made hobby grade equipment. The initial run out was .02mm (Indicator is a Mitutoyo with .01mm reading). After trying my best to tap it to concentricity I managed to lock it down to about .01mm run out. I have seen video of run out significantly less than that to a point that the DTI is barely moving. Of course they have better quality equipment.

Am I asking too much for the chuck's tolerance of the .01mm run out is ok. I'm telling myself its fine but I would like to know what you guys think. I don't aim for perfection just getting the majorities opinion and if most of you think this is not good enough then I would not use it anymore.

Thanks!

Raphael

Hi Raphael,

I guess it depends on how fast you run your chuck and how accurate your work needs to be. If your Chinese ER chuck has a run out of 0.01mm, I wouldn't complain. I just bought a cheap Chinese ER-16 chuck and it has slightly more run out than yours. The cheap ER collets I have says max run out 0.015mm.

Thor

I should've thought you'd need to be doing some unusually demanding work for 4 tenths TIR to represent a problem - certainly outside the scope of 'ordinary' model engineering. Perhaps some high-revving model aircraft engine parts, but even those only if you have to take them out and re-chuck between ops.

That runout is certainly acceptable for hobby use. If you have an application which calls for higher quality finishes, it could be worth taking the time to get things running at the best possible runout, otherwise just set everything up and mill. It would be worth doing a test to see if there is a noticeable difference between the same tooling with maximum and minimum runout set. I have two er25 collet holders, one has much better runout than the other, both Chinese. I have a few higher quality collets in the most common sizes which help also.

It is kind of hard to say if it is good enough as that varies with the application and tool.

Try doing some of planned the work operations with the setup? If the results you get are at the precision and quality levels you need, and tool life is acceptable, then no problem?

If you just want a number I believe the industrial milling ER holders I got are rated at around half or one-third of that. Then again I also use Weldon tool holders and Jacob chucks which are likely worse - they still work fine for their respective use .

It has been suggested on another thread, that the concentricity of ER chucks/collets can improve if the clamp nut is tightened harder.

Have never bothered to check, but might be worth checking?

Cleanliness of chuck and collet is taken to be the case.

Howard

Raphael

Adding to Howard's comment about tightness, for ER 32 it is worth using a "bearing" nut to ensure that the collet is adequately closed down onto the workpiece. These nuts reduce the effort required to close the collet.

hth

Dave

Check the runout of the chuck itself before going ape on the nut. Use an indicator on the inside taper. .01 runout could be the spindle, the chuck or just a speck of swarf somewhere.

Also adding to Howard’s comments, specs for collets are generally only applicable to the nominal size. Anything smaller will, presumably, not necessarily meet that same specification.

About .02mm runout is about as good as you are likely to get from cheap Chinese tooling or from worn-out old Brit iron either. You've done well to get it down to .01mm. If you want better, buy better at the higher price, or keep buying more cheap ones until you find a good one - if you are lucky.

You said yourself: "I have seen video of run out significantly less than that to a point that the DTI is barely moving. Of course they have better quality equipment."

But being a good machinist is not all about having the perfect equipment. It also is about doing the best with the equipment you have. So if you want perfect concentricity you need to machine the job in one-set up, or between centres. But for most home hobby work such as making model steam engines, working within .02mm is sufficient accuracy. Part of being a machinist is knowing to leave enough clearance on the matching parts to allow for these small variations in dimensions.

If you are turning two or more diameters without disturbing the workpiece, run out will not matter. The diameters that you turn will all be concentric about the centreline.

If the workpiece is moved, then run out may well become a significant factor.

It all depends whether your form of model engineering is making things, or chasing microns.

Howard

Thanks Gents,

I really appreciate the input here.

Collet and backplate are clean and free of any blemish of tiny specks of debri. Fully cleaned with acetone before assembly. Spindle (S7) is also checked and running true and in perfect alignment (also checked by DTI). Bolts are equally tighten as good as I can (no torque wrench to use). The process of checking is via the taper of the ER chuck itself. I have yet to check with collets in place but will wait for the ER nut with bearings as suggested. Again its the cheap Chinese collet (all I can afford now). Im not expecting any super precision accurate reading and I understand what I am buying so risk are on myself. If I can get away with a .01 to .02mm run out as explained here then that I can accept. Maybe if I can afford a better quality equipment someday .... we will see. I'm happy with the inputs here and at .01mm - Chinese and cheap I can live with that.

Great input guys and much appreciated!

Raphael

Edited By RAPHAEL VAL GOLEZ 1 on 11/12/2019 12:30:39

Thanks Howard. I do turning without disturbing the work but at times could not totally avoid taking it out and doing the opposite side. Life would be easier is it is in one go like you have said (without disturbing the work piece) but sadly its an avoidable at times.

I love making things Howard but sometime numbers are also important to get it as close as possible but definitely I'm not the one chasing microns.

Raphael

What Howard says is correct and actually very fundamental to the accuracy of turned work. Whilst you have clocked the nose of your lathe, what you need to do, if you have not done so, is to insert a length of bar into your collet and turn the piece. Then clock the work piece itself, and you will probably find that your small amount of run out will disappear. It is run out on a finished piece which matters after all.

Hi Mike I fully understand this and totally agreed. As what was explained already it will run true once you turned it to length as it will follow the concentricity of the spindle axis. The problems starts when you take things out (For whatever reason you have) and do the turning in the opposite to try match it. If you have run out then it won't match the diameter. As suggested earlier you can do in between center to match the whole length. Most of the work I do are short lengths hence this topic. I don't do turning on long piece to justify doing in between center turning and there is no way I can do that in the small length parts I do. So chucks and collets for me.

Edited By RAPHAEL VAL GOLEZ 1 on 11/12/2019 13:59:41

To ensure that the work is centralised accurately when changed end for end, this is where the 4 jaw chuck comes into its own.

If only a 3 jaw is available, the nearest that you get is to mark the work and the chuck, and when the work is changed from one end to other, align the marks on the work and the chuck. This should position the work in the same place as it was originally.

Probably, I can be accused of being a micron chaser, but not a very good one, when I use the 4 jaw, as I like to get the work centred to better than 0.0005" (12.7 microns) if possible,.

Years of "fitness for purpose" have blunted my love of absolute accuracy.

Howard