DC Motor Comm Stall Marks. Odd replacement motor behavior. 4

[rawelk]

Background Info
A Reliance B507ATZ frame, 250HP, 1750 RPM base speed (2300 RPM fully field weakened) 500VDC/400A armature, stabilized shunt, 300 VDC shunt field (4.84A hot/1.1A fully weakened)DC motor is used to turn a plastic extruder screw. This extruder is equipped with a melt pump between the screen changer and downstream adapter zones, and feeds a 50" wide sheet die. A 250HP, 500V/400A Reliance MaxPak+ S6 drive is used for variable speed operation.

Recently it developed stall marks on the commutator (burned, low bars positioned at brush rigs). My suspicion is either one or both of the following conditions were the cause:

1). Extruder was started using the production department's nominal speed setting of 20% before barrel and adapter zones were fully at temperature.

2). Extruder was started with 100% speed reference under same extruder temperature condition. This can happen if the melt pump suction pressure controller is not placed in manual and provided a reasonable setpoint before starting; during shutdown output will rise to 100% when in auto.

Questions (part 1)
1). Are there any rules of thumb, or analytical methods for determining how long such a motor can sit stalled (no armature rotation at up to full load amps) before stall marks are created?

2). If going the analytical route ... if memory serves, motor commutator bars are not pure copper, but rather a copper/silver alloy. What percentage of silver might one expect?

3). Another factor - I'm thinking comm bars are tempered in some fashion, and have an upper temperature (500 degrees F comes to mind) above which tempering is compromised. If this is the case, how big a problem is losss of temper, and what are it's manifestations?

Odd motor behavior
Background - 20 years ago, we used both Reliance B507ATZ frame 400HP motors (on 6" diameter extruders) and older, obsolete "U" frame 300HP motor on 4-1/2" extruders. In the course of time we installed 300HP motors (in B507ATZ frame) to replace the "U" frame motors, and (when purchasing new equipment) specified B507ATZ frame motors for both 250HP and 300HP machines. This allows us to keep a single 400HP motor in stock that can bolt up onto any of the other extruders.

In the past I've never had any functional problem installing the 400HP motor for temporary service while the failed 250 or 300HP motor was being repaired.

However, this time, when the 400HP spare was put in place of the stall marked 250HP motor it didn't perform.

I was able to set up the drive for proper operation throughout base and weakened speed ranges under low load conditions. Wierd thing is - sometimes the drive would act as if in current limit (speed would drop very low - about 520 motor RPM) at about 350 amps, and other times would run just fine under the same set of conditions. Current feedback was tracking armature amperage properly (as measured with a DC clamp). The only thing of note was it's been abnormally cold outside (and the motor's clean air supply is mixed with outside air). The Reliance field engineer we called upon (a 20+ year vet) couldn't figure out what was happening either.

The symptom disappeared after reinstalling the repaired 250HP DC motor.

Reliance doesn't endorse running the 400HP motor from a lower HP-rated drive, but the question is why have we been able to do this successfully up until now?

Bob

 

[jbartos]

Suggestions marked ///\\rawelk (Industrial) Feb 3, 2004
Background Info
A Reliance B507ATZ frame, 250HP, 1750 RPM base speed (2300 RPM fully field weakened) 500VDC/400A armature, stabilized shunt, 300 VDC shunt field (4.84A hot/1.1A fully weakened)DC motor is used to turn a plastic extruder screw. This extruder is equipped with a melt pump between the screen changer and downstream adapter zones, and feeds a 50" wide sheet die. A 250HP, 500V/400A Reliance MaxPak+ S6 drive is used for variable speed operation.

Recently it developed stall marks on the commutator (burned, low bars positioned at brush rigs). My suspicion is either one or both of the following conditions were the cause:

1). Extruder was started using the production department's nominal speed setting of 20% before barrel and adapter zones were fully at temperature.
///OK, if this implies that the plastic is harder at lower temperature thus causing the motor to be excessively loaded.\\2). Extruder was started with 100% speed reference under same extruder temperature condition. This can happen if the melt pump suction pressure controller is not placed in manual and provided a reasonable setpoint before starting; during shutdown output will rise to 100% when in auto.
///OK, if this implies that the extruder is supposed to start with lower speed to accelerated the plastic mass.\\Questions (part 1)
1). Are there any rules of thumb, or analytical methods for determining how long such a motor can sit stalled (no armature rotation at up to full load amps) before stall marks are created?
///Yes. Safe way to obtain it is from the motor acceleration curve that is sometimes available from the motor manufacturer. Unless the motor is design for a heavy-duty service, e.g. plugging, jogging, etc., motor stallings are not recommended.\\2). If going the analytical route ... if memory serves, motor commutator bars are not pure copper, but rather a copper/silver alloy. What percentage of silver might one expect?
///There is no need to worry about silver, since the silver is the better conductor than copper.\\3). Another factor - I'm thinking comm bars are tempered in some fashion, and have an upper temperature (500 degrees F comes to mind) above which tempering is compromised. If this is the case, how big a problem is losss of temper, and what are it's manifestations?
///The commutator manufacturer would not compromise the product by any copper processes negatively impacting the commutator electrical and physical properties. If done by someone else than manufacturer, then the commutator may become defective.\\Odd motor behavior
Background - 20 years ago, we used both Reliance B507ATZ frame 400HP motors (on 6" diameter extruders) and older, obsolete "U" frame 300HP motor on 4-1/2" extruders. In the course of time we installed 300HP motors (in B507ATZ frame) to replace the "U" frame motors, and (when purchasing new equipment) specified B507ATZ frame motors for both 250HP and 300HP machines. This allows us to keep a single 400HP motor in stock that can bolt up onto any of the other extruders.
///As long as the motor has the same input and output characteristics, mechanical vibration limit, and proper mechanical alignment, there should not be any problem.\\In the past I've never had any functional problem installing the 400HP motor for temporary service while the failed 250 or 300HP motor was being repaired.

However, this time, when the 400HP spare was put in place of the stall marked 250HP motor it didn't perform.
///There may be several causes, e.g. power supply quality, the load characteristics, the motor output characteristics, etc. different.\\I was able to set up the drive for proper operation throughout base and weakened speed ranges under low load conditions. Wierd thing is - sometimes the drive would act as if in current limit (speed would drop very low - about 520 motor RPM) at about 350 amps, and other times would run just fine under the same set of conditions. Current feedback was tracking armature amperage properly (as measured with a DC clamp). The only thing of note was it's been abnormally cold outside (and the motor's clean air supply is mixed with outside air). The Reliance field engineer we called upon (a 20+ year vet) couldn't figure out what was happening either.
///The proper analysis or troubleshooting may get involved in some cases. Usually, different parameters of motor in more complex controls cause different motor functioning.\\The symptom disappeared after reinstalling the repaired 250HP DC motor.

Reliance doesn't endorse running the 400HP motor from a lower HP-rated drive, but the question is why have we been able to do this successfully up until now?
///Something may have gone wrong in the drive or in the motor power supply. This would require a root-cause analysis.\\

 

[TomG33]

RAwelk,
I guess the first question to ask here is "How do you know they are Stall Marks ?"

Usually commutators are made up of many segments each shaped like the outside part of a long skinny pie segment,

These are packed together with mica insulation between them and then processed somehow with heat and preasure to make the complete commutator, which is then built into a support of some type.

In a proper Stall damage situation, the high current through the relatively few bars causes them to heat up excessively, thus expanding and because of the wedge shape, they will force themselves out, a little. These will not go back in. The result is that under each Brush arm there will be some commutator segments that are higher then the others by a few 1,000ths of an inch.

If you try to run the motor after this, The high comm segments will cause destructive damage to the brushes and usually they make heaps of noise and spark excessively.

Your description of the problem does not indicate that this is the case.

DC Machines develop a commutator pattern that we refer to as "Pitch Bar Marking" which eventually turns into Pitch Bar Burning". THis usualy starts out as burning marks on the commutator, spaced out evenly around the commutator, and equal in number to half the number of brush arms. As the burning progresses they appear in the intervening places so that the final appearance is "one for each brush arm".

These are frequently erronousely referred to as "Stall Marks" because they appear to be under each brush arm and a stall appears to be the only possible explanation.

We see these marks frequently on MG Set Generators, which run at fixed speed with an almost zero possibility of a physical stall.

These are particularly hard to troubleshoot because the causes can be numerous.

When a certin electrical or mechanical disturbance exists which is in phase with the motor armature rotation, excessive arcing , metal transfer and associated burning occur every time the same commutator bars pass under the cathode brushes. Each burn mark can involve more then one commutator segment.

Once the marks start, the burned commutator bars will become lower then the rest of the commutator causing the brushes to "Bounce" as they go over the low part. This bouncing causes more sparking which causes more burning and thus deeper burn marks, The only way to fix it is to resurface the commutator. ( machining, not stoning )

The sad part about this that the initiating situation may only be transient, but once the burning starts it continues, even after the cause no longer exists, until the commutator surface is fixed.

The things I would be looking for would be

1 A bad connection in the armature
2 an unbalanced Shunt field
3 Brush Preasure that is too light
4 Mechanical vibration at motor speed.
5 Chamical contamination of the cooling air.

Perhaps this is enough information to get you started

I cannot answer your comment about the larger motor under performing, apart from commenting that if the motor was the problem, then the drive would tend to put more current into it to keep it at required speed, If both the speed and the current are falling, Chances are You have some sort of drive or reference issue going on.

Does the Drive use a speed feedback of some sort? Tacho generator, Pulse Tach?

Regards
Tom

 

[rawelk]

I should have mentioned that the 250HP motor was seen arcing fairly severely one morning, and was immediately shut down. Unfortunately, before I arrived it was decided to hand stone the motor (which, as Tom observes, doesn't do any good) and muddied troubleshooting.

I'm going with the stall mark theory for now because:
1). So far as I'm aware there were no drive issues preceeding the observation of brush arcing on the morning of the downtime episode.
2). Two severe marks were located across from one another (which I'll refer to the 0 and 180 degree positions for sake of explanation) one brush width wide each, and a pair of less noticable marks at the 90 and 270 degree positions.
3). The motor shop turned and undercut the comm, and replaced the motor bearings, but (except for checking for insulation failure, etc.) did nothing else.
4). This motor has been installed and is running well with 'black' commutation - no arcing.
5). Before removing the 250HP motor we replaced the brushes (which were down to about 1 cm of usable life remaining) to see if increased brush pressure helped. It didn't, but I observed only light arcing at motor speeds lower than approximately 1600 RPM. At higher speeds arcing (and brush chatter) became progressively more severe. Also, brush bounce was readily felt while pressing against the back of the brushes during seating (I used an Ideal brush seating stone while running the motor rather than using a sandpaper method).

So far as drive speed reference in the second case - neither I nor the Reliance tech found any issues there (BTW-speed feedback is via a 600 PPR Dynapar encoder translated to 0-10 VDC via their model FV2 pulse-to-voltage converter). With the original motor reinstalled the drive is performing properly ... one of the reasons we're scratching our heads on this one.

On the other hand, pitch bar marking isn't completely ruled out, but I would suspect arcing caused by it would have a more gradual onset.

A peripheral observation is a good possibility extruder operators were often trying to start the machine before fully at temperature. In the course of troubleshooting the drive problem noticed the melt pump body zone was slow to come up to temperature, and an operator unsuccessfully attempted a startup - motor wouldn't turn - with it nearly 50 degrees under setpoint.

I re-tuned that zone's PID loop for faster response, and am in the process of stressing the importance of proper startup and operational procedures, but (if attempting startups with cold zones is as prevalent as it appears) it points more heavily to stall marking.

Here are a couple of photos of the comm in question.

Thanks for your valuable comments.

 

[jbartos]

Suggestion: To avoid the improper motor start and commutator damages, the more sensitive motor protection might be implemented; at least on temporary basis to pinpoint the culprit.

 

[TomG33]

Bob,
I'll leave you with it.
I would love to know the outcome (if you ever get one)

These things are always perplexing, and it is great to get feedback from a particular Case.

Good Luck
Tom

 

[UKpete]

rawelk, this is an interesting one and the photos remind me very much of a problem I saw some years ago when investigating problems on a new traction motor design. About 10% (out of 3000) motors of a particular type were affected, with bar burning very similar to that in the photos. The motor was 4-pole 50kW 5000rpm for a rapid transit vehicle.

Firstly, in answer to your original questions:

1. Stalling: No, but our motors often survived extended periods stationary at over 1000A when there were difficult starting conditions. I never came across or heard of any ill-effect from this. In fact, when you watch a dc motor starting at very high amps there is nothing to see although obviously temperatures are going up, especially if it’s a self-ventilated machine (no cooling air at standstill).

2. Commuator copper: I can’t remember much about the copper type used at the time except that the grade we used was called combarloy. I believe you are correct that silver-bearing copper is used e.g. C11500, this has 0.054% max silver i.e. next to nothing. According to an old reference book silver copper is used for commutators because it has a high softening temperature and a higher creep resistance at moderate temperature (copper alloys are particularly prone to creep at temperatures as low as 100 to 150°C).

3. Comm seasoning: When manufactured, they are seasoned by heating and spinning, stopping and re-torque the vee-ring bolts several times over. The critical item is the mica separators and the micanite vee-ring insulation, the bars have to bed properly into these hence the seasoning. There is a picture of one on this site (extreme bottom of page).

http://www.supertechgroup.com/products.htm

These were shellac bonded, although I know resin glass was also used.
The maximum operating temperature after a full load test for the commutator was always 120°C, I suspect this to maintain a proper commutator film. A commutator that had run hot was recognizable by it’s colour. The ends of the bars (off the brush track) in your photo don't show any discolouration due to overtemperature - hence I don't think it is a stall.

Regarding the problem with the control of the 400hp motor, I know very little about dc motor control, but possibly it is to do with closed-loop stability? The larger motor will have longer time constants, mechanical and electrical (e.g. the flux takes a finite time to change due to eddy currents in the frame).

Going back to the original problem, I concur with TomG - I suspect that it is a commutation problem that has led to electrical bar erosion due to localized sparking in a runaway effect, just as TomG describes. The non-burnt bars in the bottom of your photo also show that the film is poor and there has been general sparking. If you’ve still got the old brushes they can tell you a lot. A commutator that has gone uneven will cause pronounced wear marks on the sides of the brush where it has rubbed against the brush box, likewise there is more wear than usual on the brush tops where they have worn under the tensator spring.

It’s too late now but it would have been obvious if a profile had been taken, either with a device that fits in the brush box or a dial gauge with a mushroom head. Measurements comparing the profile on and off the brush track will indicate what is mechanical movement and what is electrical erosion.

A typical spec for a commuator of that size is:
Newly turned: TIR 40µm, max bar-to-bar 5µm
Service limit: TIR 200µm, max bar-to-bar 13µm (commutator requires re-machining)
(0.001” = 25µm approx).

So if it starts to spark again, take the commutator profile regularly and monitor the situation.