Commutator Undercutting and Chamfering

[safspir]

Hi,

I work on large DC propulsion motors. There has been a lot of problems with commutators lately. Personally, I feel this may have to do with outdated AC to DC rectification issues. But, that would be for another thread. What I would like to talk about is favorite ways to undercut and chamfer the commutator bars. Our management sticks with the single edge chamfering tools and hacksaw blades, which require a very trained eye to keep the cut even and bar width equal. I came here from a different school, which advocates a double edge, (triangular or square), tool to make a cut on both adjacent bars, and a depth set, width set, under cutter. Your thoughts? Also, while on the subject, they only sand in the brushes with an 80 grit, saying that anything finer produced no better results. I am old school, with a relentless thirst for new knowledge, dealing with a "we've always done it this way", mentality. I would appreciate a discussion from the members here. Looking forward. Cheers.

 

[Skogsgurra]

I do a lot of work on DC drives. But mostly when there are problems with the thyristors, gating units and control as such.

Sometimes, however, I catch a bad commutator. This one came directly from the winder shop:

http://gke.org/pub/files/DC motor commutator w all errors.jpg

I got mica removed from the bar's walls and the edges broken. It did run after that. But it is too deeply cut, between 4 and 5 mm and that is nothing I could remedy.

A side note: that shop had also managed to get excitation wrong. Same polarity on all poles.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[electricpete]

I'm not all that familiar with the topic.

Do you by any chance have access to EASA's "Fundamentals of DC Motors... Operation and Repair Tips"?

All of page 6-6 is devoted to undercutting techniques. They contrast V-cut and U-cut. They seem to look unfavorably on a tool that might produce a V cut, since that will eat too much of the sides of the bar before it gets deep enough. They prefer ideal U cut except at low speeds (less than 200rpm) "where the self cleaning characteristics of the V-cut are helpful". The V cut introduces a chamfer into the bars, but too much of one. The U cut requires another step to chamfer the edges of the bars.

There's a lot more info on that page.



=====================================
(2B)+(2B)' ?

 

[Skogsgurra]

A V cut shall, in my view, never be used. U or square and then chamfer. V is just incredibly bad practice. Why do they mention it at all?

I don't have the EASA document, so I can't say if EASA has a reason to discuss V cut at all.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[zeusfaber]

There has been a lot of problems with commutators lately.

Have you also been losing refrigerant from your CW plants?

A

 

[electricpete]

A V cut shall, in my view, never be used. U or square and then chamfer. V is just incredibly bad practice. Why do they mention it at all?

The EASA picture shows a pure Vee extending from bevel at the top corner of the two bars down to the apex at bottom center in the mica. Again, they say the only place it is used is low speed (<200rpm) "where the self cleaning characteristics of the V-cut are helpful". The quotes indicate their words, not mine. I can't explain exactly what they had in mind, why it's beneficial or whether there is unacceptable reduction bar compared to oem. I've never dealt with a slow speed dc machine for that matter but I gather there are some significant differences. EASA are to my knowledge the best available source of repair guidelines for motors in general.

Op mentioned use of triangle file, which is what drew my attention to that portion. I see now it's clear his intent is that is just to use that triangle file to bevel the corners of the two adjacent bar edges evenly and maybe put a line at the center of the mica... then he would use a U cut further down into the mica. That makes good sense to me fwiw.

My suggestion is for op to get the document I cited. I am not in a position to convey everything that's in there because I don't understand it all.





=====================================
(2B)+(2B)' ?

 

[edison123]

In our shop floor, we use U for the mica undercut (about 2 mm) and bevel both edges of the copper slightly with a V tool. The undercut sides are cleared of mica. The winder's skill is important here.

Muthu

http://www.edison.co.in

 

[safspir]

Thanks for the input folks,

I guess it just comes down to personal preference of chamfering tools. No input as to the brush seating grit? Again, we use only 80 grit paper, nothing finer.

For those who are into this stuff, we recently had an interesting comm problem. Someone left a tool inside one of our large DC propulsion generators and it damaged the comm in it's travels. Our machinists turned the comm in place, powered by the diesel prime mover. They said that their total runout was less than 2/1000. We profiled it, both with our old roller profiler, and our new ruby tip one. The profilers both agreed, total runout of 5/1000, with high point "lobes" in a pentagram pattern. These generators originally had a bearing at both ends, then coupled to the diesel. Someone in the past decided it was best to remove the second bearing (cooper), between the gen and the diesel. We analyzed the readings, taken at 0,90,180,and 270, then compared them to the profiler results. Our contention is that the machinists readings were taken at an average spot, not taking into account the highs and lows identified by the profiler, off of these set points. So, we got the order to stone the comm, which we did. Shiny in spots, not so much in others. Brings to mind thoughts of vibration, harmonics, etc.. Now we're resetting the brush holders, perfect tape and all that. We'll see what happens when it is put on line. I'll keep you updated on the results.

Again, thanks to all of you for your input on this.

 

[Skogsgurra]

Was the rotor connected to the engine when you ran the profiler? Thinking crankshaft and pattern...

What profiler do you have? Any filters in it?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[electricpete]

No input as to the brush seating grit?

http://www.reliance.com/prodserv/motgen/c7090/

Rough seating can be done with 60 or 100 grit sandpaper. Final seating should be done with fine sandpaper.

As I recall emory cloth should not be used due to metal particles. Garnet paper is better.



=====================================
(2B)+(2B)' ?

 

[motorspert]

With regards brush seating grit, we have always been told to use 100 grit if the surface is highly polished. A wise old engineer once told me that too highly polished surface causes problems building a skin on the surface which can lead to high brush wear. We work to an ideal surface finish is no better than 0.635 micron N6.