Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Motor anti-reverse-rotation device failure (sliding pin type) 10

Status
Not open for further replies.

electricpete

Electrical
May 4, 2001
16,774
0
0
US
Machine Description: Outdoor vertical motor (700hp, 900rpm) driving a pump through a rigid coupling. Rated torque = 5250*700/900 ~ 4100 ft-lbf. Locked rotor torque is probably around 8000 ft-lbf.

What happened: Motor tripped during start on time overcurrent (51 relay). On-site inspection revealed motor would not rotate by hand, even after uncoupled. Sent to repair shop who provided attached photos showing a damaged anti-reverse-rotation device.

Overview of anti-reverse-rotation device operation: Slide 1 is an overview of the parts of the device. Principle of operation: During start, the vertical pins slide up the ramp on the stationary ratchet plate which pushes them into the holes in the rotating pin-holder where they are held I place against the side of the hole by centrifugal force (and friction). When the motor slows down, the pins fall down onto the stationary ratchet plate again but a portion also remains in the hole. If the motor rotates reverse then the pin is pushed backwards against the flat vertical portion of the ratchet plate preventing reverse rotation. By the way normal rotation is CW viewed from the top. There are 12 equally spaced ratchet ramps and 6 equally spaced pins (why are there 12 holes... there were originally 6 holes but during a previous refurbishment they were found damaged so 6 alternate holes were drilled at that time... that previous repair was associated with reverse rotation damage to the holes and the vertical faces of the ratchets found during proactive refurbishment which had never caused any problems with motor operation in the plant).

Inspection Findings: The photos show that one out of 6 pins is broken roughly in half (not much deformation, maybe a brittle failure?). One pin hole has a wallowed entry (on the side that would suggest that damage occurred while rotating forward with pin stuck on ramp)[revised to remove incorrect info]. One ramp is severely damaged. The vertical portion of another ramp is slightly damaged at the top.

The repair shop’s conclusion is that the pin bound in the hole due to rust and lack of clearance (0.005” clearance on a pin which is maybe 0.5” diameter and 1” long). They plan to increase clearance to 0.020”. They also suggested using stainless steel pins to address the rust, but as far as I can tell from the photos the rust is mostly on the pin holder rather than the pins (I’ll ask them about that). I asked if parts could be lightly coated in oil to reduce rust. They said lubrication is not typically used since it would attract/retain dirt/dust. I suggested some kind of dry powder lubricant and they said they’d consider it.

MAIN QUESTION: Based on the info above, do you have any ideas about the cause of the failure or what are the proper actions to prevent recurrence (I'm preparing for a phone call with the shop Friday 1/22 to finalize our repair plan).

My questions for the shop (not necessarily important for responders to this post). To prepare for my phone call I made a list of questions. This is not the main part of my post, I'm just listing it here because it's convenient for me. You can add questions or weigh in on my queestions if you like, but you don't have to read them (all the important stuff is above):
1 - Is there rust on the pins? (I don't see it). What can we do about rust in the pin holes?
2 - What is the silver pattern on bottom of the pin holder slide 2
3 - Slide 4 why does the brown rub pattern seem to be further out at a larger radius on the ramp and then closer in at a smaller radius on the flat?
4 - Is the rub mark on the non-damaged stationary ramps normal / expected outside of a failure?
5 - What about the slightly damaged vertical wall slide 5 at 10:00 position (two to the left of the obvious damaged ramp). It looks as if the pin was starting to fall and caught the top of the vertical wall and then pin pushed up out of the way. Is that normal?
6 - What materials are the pins, pin-holder, ratchet plate? Are any of these parts painted? (what explains the colors on the ratchet plate).
6a. Regarding pin material - Is it expected to have a brittle looking failure without much pin deformation
6b [new]Were the materials changed during the last repair? Is the stationary ratchet plate now harder such that more of the reverse impact is transferred to the pin?
7. How much clearance is between the top of the ratchet and bottom of the pin holder
8 What are exact dimensions of the pin (I said 0.5" x 1" but that's just my guess). And what is the height of the stationary ramp?
9. If we lined up the pin's plane of the failure with the bottom of the pin holder, where would the bottom of the pin be... resting on the bottom between ramps of resting in the middle of the ramp where the damage is or somewhere else.
10. are those clearances per side or diameteral.
11. You'd think the pin breaking would allow the motor to successfully start. Did the pin break before trip but still slowed it down enough to trip?
12. What is the radius of curvature of the bottom corners of those pins? Would a larger radius of curvature be beneficial to help the pin slide smoothly up the ramp during start?
13. Could the proposed larger clearance allow the pin to tilt enough to score the inside of the holes during starting? Maybe larger radius of curvature on top corner of the pins would help prevent that? And also radius of the hole entry?


miscellaneous note - Since the are exactly twice as many ratchets as pins, all six pins will contact at the same time (assuming dimensions are precise enough to permit them all to contact). Other designs have different odd combinations of pins and ratchets which means the pins wouldn’t all contact at the time (one would hit first to stop motion) but that type of staggered design has the advantage that it significantly limits the arc available for reverse windup before contact is made and is therefore probably a better approach in terms of stresses during stopping reverse rotation. But the distinction doesn't seem important because it appears the damage here occurred during forward rotation.

=====================================
(2B)+(2B)' ?
 
 https://files.engineering.com/getfile.aspx?folder=738ac414-8c69-4b3a-8a23-9ca8e890867b&file=AsFoundAntiRotDevice.pdf
Replies continue below

Recommended for you

Just for thought, the anti-rotation should engage immediately the motor comes to a stand-still - possibly the pump started reverse rotation before the pins engaged, ie the pins were in the up position and slow to engage allowing the pump / motor to run in reverse, resulting in very high load on the pins when they finally engaged - - maybe only 1 or 2 engaged.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)
 
> the pins were in the up position and slow to engage allowing the pump / motor to run in reverse, resulting in very high load on the pins when they finally engaged - - maybe only 1 or 2 engaged.

Thanks Artisi. Damage that occurs during start / forward rotation would be on the ramp (where we see most of our damage as in slide 5 12:00). Damage that occurs during reverse rotation would be on the vertical wall (where we see only minor damage slide 5 10:00).

A DETOUR to talk about damage we saw during a previous refurbishment which was fully repaired long before the current event. In that previous refurbishment the location of the damage was on the vertical wall and indicates there was at one time more stress from stopping reverse rotation than the device could handle. photo attached to the current post shows that previous since-repaired damage.

But let's set aside that previous failure (long ago repaired) and focus on the recent event and recent findings:

The first indication of a problem was during start (although maybe that was not time of initial failure).

The main damage we see to the stationary ratchet plate in the current damage is on the ramp (indicating forward rotation) as slide 5 12:00. Although there is also a little bit of less-severe damage on the top of a vertical wall of the stationary ratchet plate (slide 5 10;00 position) which presumably would have occurred during reversing. Maybe that could somehow cascade to later produce the rest of the damage during a subsequent start? I don't rule it out but I can't visualize the sequence of events that would be involved.


The location of the damage on the pin-holder hole (slide 3, remembering normal forward rotation of this pinholder piece would have been CW from top = CCW when viewed from the bottom which is the side this photo is taken from) all suggest that the damage occurred during forward rotation / start*
[ul]
[li]* Note my assumption above in analysing pinhole damage was that the damage occurred while the pin was still intact and in the hole... but maybe there are other scenarios like the broken pin had already broken and damaged the hole on it's way out in which case it may have been pivoting and the top of the pin could have caused damage at that location during reverse rotation... but that scenario seems unlikely to me because I would have expected additional damage on the pivot point at the opposite side of the hole in that case (ovalized hole entrance). Further analysis of height of the pin fracture relative to the other parts (questions 7/8/9) might give a clue where the pin was at the time it broke... it may be that the height of the break suggests pin had to be fully dropped at the time it broke which could argue for damage during reverse rotation rather than while climbing up the ramp in forward rotation. Also I'd think a careful inspection of additional marks inside of that hole would distinguish these two scenarios[/li]
[/ul]

=====================================
(2B)+(2B)' ?
 
electricpete said:
it appears the damage here occurred during forward rotation
I have seen pump shaft broken presumably during pump start while on reverse rotation. Here also it appears brute force was applied on ratchet pins and it broke as a result.
I apprehend that there is a leakage in the discharge check valve so there is certain amount of reverse torque available due to reverse flow before start. As the pump starts, initially motor overcomes this reverse torque. After it starts rotating, there is a high acceleration as it experiences almost no discharge pressure till the back side of NRV becomes pressurized.
During the high acceleration period(fraction of a second) the pin has insufficient time to slide back and may have created a notch due to the impulse which prevents further sliding. As considerable amount of water is getting accelerated, the torque is also high acting on a single pin, which results in breakage.
In order to prevent this I suppose the leakage across NRV to be minimized. Also it is safer to use all 12 pins available in design which reduces the probability of failure. It also may be worthwhile to lubricate the pins and ramps.
It is also to be noted that sliding of pin over ramp may be prevented by presence of surface defects. So the surfaces should be checked periodically.

Engineers, think what we have done to the environment !
 
> I apprehend that there is a leakage in the discharge check valve so there is certain amount of reverse torque available due to reverse flow before start.

My understanding is there is no check valve or NRV on this machine. We don't use motor anti-reverse rotation devices except on pumps that don't have those (which are generally low speed high flow low dp pumps attached to pipes which are very large, maybe too large for check valve to be practical). The parallel pumps were in operation at the time of our failed start attempt.

> Also it is safer to use all 12 pins available in design which reduces the probability of failure.

I'll discuss it with shop, but original design was 6 pins, 12 ramps. Double the pins might increase wear on those ramps or have some other unintended effect.

> It also may be worthwhile to lubricate the pins and ramps. It is also to be noted that sliding of pin over ramp may be prevented by presence of surface defects. So the surfaces should be checked periodically.

Thanks. I'm definitely with you on the role of lubrication for preventing binding and corrosion (mentioned it in the op). Periodic inspection is something for us to consider.

=====================================
(2B)+(2B)' ?
 
Are there any other ideas / comments?

I have my phone call a little over an hour from now and we will have to finalize our repair plans for this motor at that time.

=====================================
(2B)+(2B)' ?
 
Phone call is done. There was not a lot of time for all my questions. We will proceed with the same repair plan as above (widen the pin clearance, use stainless steel pins possibly larger diameter, use dry lubricant.

I did find out the pinhole that has the obvious damage was one of the unused pinholes... so that pinhole damage was leftover from before, not the recent event. Also the pins are apparently 1.5 - 2" long. The shop said that close examination of the failed pin showed that it had fine crystal structure for the outer 1/8" and coarser grain structure inside, so they think it was made from case hardened dowel rod, and that may have contributed to the brittle failure.

=====================================
(2B)+(2B)' ?
 
I don't think stainless steel is a wise choice here. Are you going to be using a hardened stainless steel? Is it at least as hard as the existing pin? Stainless steel is extra sensitive to fretting and galling. Drilling the holes oversized and sleeving with oil impregnated bronze bushings would be a better solution.

I suspect fretting is an issue because you mentioned a pin being bound by rust. I see no corrosion (pitting) present in the pictures so the rust could have been the product of fretting corrosion. A bronze or Nitronic (xx) bushing will solve this.

I also suggest milling the holes as opposed to drilling. A drill doesn't provide as precise a location so one pin may end up carrying more load if it contacts first.

Better yet, make the holes square and use square pins. This will give a better bearing surface to carry the load. Drill the corners as connect the dots. This will save you from having to repair the already damaged holes.
 
Thanks TugBoatEng, those are great comments.

I'll ask about the stainless pin hardness. Harder is less susceptible to galling I gather. But also the difference in hardness between two materials is important, right? I think stainless steel galling is more prevalent when both materials are stainless steel than when stainless is combined with other steel. I'm not very familiar with how to prevent galling and I'm open to any more comments on all of that. At any rate we will use some kind of dry lubricant like moly disulfide or grpahite in isopropyl alchohol to help reduce concerns about galling.

I'm not familiar with oil impregnated brass bushings. Googling shows me they are made with a special process to produce a structure that is porous to allow oil in. I wonder if that structure is still ductile like bronze of if it becomes brittle / susceptible to cracking ? It would need to be able to stand up to a pretty good shock during stopping reverse rotation. Bronze or nitronic bushing are other options to look at. I don't think I have enough time to be able to change the plan for the current motor repair, but it'll be something to look at for the next motor.

I agree precise hole spacing and ratchet spacing is absolutely critical to have a prayer of load sharing in this design (other designs with staggered odd numbers of pins and ramps don't rely on sharing load but still benefit from the other pins in limiting the worst case reverse windup arc). It is interesting that in the photo of a previous damage the pins had deformed the vertical wall part of the stationary ratchet plate. It seems there was some ductile deformation going on that may have ended up helping to equalize the loading (regardless of any previous spacing mismatch on the stationary parts) and also perhaps the deformation energy absorbed a bit of the shock so the pins didn't see as much (those previous pins didn't break). I don't know if that was part of the design intent or not. That previous damage did NOT cause any operational problems... it was only found during proactive refurbishment. The same ratchet plate was repaired by weld buildup and re-used and now we ended up with the pin failure. Maybe the weld repair was not as ductile as original metal (although of course the leading theory is still that the current damage occurred during start so if that's true then it wouldn't be relevant to the current damage).


=====================================
(2B)+(2B)' ?
 
I think the failure initiated during the previous anti-reversal. The pin broke in the notch before the one with the gouge. The broken pin kept the remains from dropping completely out of the hole. During the start, as the broken pin advanced to the next slot the remains dropped out of the hole and that's when it wedged and bound. It must have happened this way as this would allow the motor to gain some momentum before being abruptly stopped to create this amount of damage.
 
Thanks. I don't rule out that it occurred during reverse yet but I'm still leaning against that theory. If that were the case you'd think the pin would show some distinctive damage pattern on it like sliding smudge across that fracture surface (or on the side of the pin fragment if it had tipped over). Also during start there is a large electromagnetic torque (4100 ft-lbf) present even without the additional torque from rate-of-change of momentum (in contrast to stop/reversing when the rate of change of momentum is the sole source of torque). But I still have skepticism that we really understand the failure scenario and the factors that contributed. I've asked for some additional vertical measurements (7/8/9) to help piece together the actual failure scenario... let's say when the pin fractured the elevation of that fracture was at the bottom of the pin-holder (there's not much clearance between bottom of rotating pinholder and top of stationary ramp), then work down from that to figure out where the bottom of the pin would've been at the time it failed (is it on the ramp or at the bottom of the ramp against the reverse wall stop). If it ends up that the elevation of the failure would've been inside the hole assuming the bottom of the pin was at the middle of the ramp, then that would seem to contradict the theory of failure during start (I don't think the pin would break at a location that was inside the hole at the time of failure).


=====================================
(2B)+(2B)' ?
 
I looked again at "previous since-repaired damage" photo, and cannot believe the amount of damage shown on the vertical face, you can also see damage on the up-ramp where the retaining pin has gouged into the face. Seems the pump unit was running backwards (more than once) prior to the pins dropping into place. Also see the same on the latest failure although not as bad. To me it seems the pins are being held up and very slow to drop, possible made worse as the speed increases (in reverse) the centrifugal force is also assisting - would also look to rounding the end on the bottom of the pins.

added: the pins should drop immediately the pump comes to rest and prior to reverse rotation starting - that's assuming there is some back flow, unlikely not to be - would except there is a pause before the change of direction - back flow needs to develop enough torque to start rotation, same as start-up torque at which time the pin/s should have engaged

So what's holding the pins up?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)
 
I think the engaging pins should have a spring assisted mechanism to ensure faster engagement.
Also it is strange not to have a discharge check valve. A controlled check valve can prevent water hammer as well as reduce the reverse rotation or when there is a anti reverse rotation mechanism, it can reduce the severity of impact on the same.

Engineers, think what we have done to the environment !
 
electricpete

is it possible to display the actual assembly with 2D cad view in pdf. I having difficulty understanding how this coupling assembles.
problem I see the pins are still engaged while reversing. thus catching the tips of the pins. from previous small designs there is some type of spring
to retain engagement but it is freely allowed to overcome the spring and the jaws of two ratchet faces to disengage with out damage.
normally if it does not disengage properly there is a shear neck on a shaft that is designed to to shear at a predetermined torque.

problem I see here is I don't view the other mating face ratchet that engages with the one shown. the assembly is not dis in gaging properly.
the hole may need counter sinks that allow full disengagement.
edit rounded spherical heads on the pins would also help
 
mfgenggear :this is not a coupling, it's a hold-back device to stop reverse rotation of the pump / drive motor, in forward operation there is no connection between the hold-back and pins.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)
 
Electricpete :,
Question, why is the face of the rotating pin holder and the ratchet hold-back "blued" from high temperature as well as showing signs of rotating contact?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)
 
Tugboat:having looked again on the comp. rather than the phone, you are right, but why, there is no setting out etc on these components?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)
 
> Do notice that the gouge is not centered in the ramp.

I noticed that too... on all the ramps both the current damage and the previous damage. I think the geometry of the ramp is tricky (it's not something that you would intuitively select if you were going to build this thing yourself). If you were to draw a line of "straight up-hill" (highest elevation gradient) on these ramps, I think that line would not be pointing directly circumferential. Instead it would be pointing somewhere between circumferential and radial. Due to that funky orinetation of the ramp, only the outer part of the pin makes contact with the ramp on the way up the ramp. Then when the pin gets to the level top of the ramp, a larger area of the pin centered on the pin axis makes contact and the mark is further inwards.

What is the purpose of having the ramp slanted in that crazy way? I'm not sure.

=====================================
(2B)+(2B)' ?
 
Status
Not open for further replies.
Back
Top