Problem with ratcheting gear 20CS on gas turbine

B

Thread Starter

BigTom209

Well, first time poster here. Hopefully you guys can lend some insight into the problems we're having at one of the remote plants.

Woodward control system, feedback to the control system from 33CS and 33HR

We can't seem to get this ratcheting gear to function as it should, we have the cover off and basically what's happening is at the start of the cycle we can see the 2 cylinders close hard pressing the jaws together, then it turns backwards (ie unlocking the jaw) for about 3 seconds, then forward for about 6 seconds to where the jaws JUST get into the locked position and starts turning the shaft, then it reverses again for about 2 seconds where now the jaws now are completely unlocked, then 33HR is triggered, our woodward system kills the 120VDC to the 20CS and the 88HR and now since the 2 cylinders are spring loaded the jaws completely seperate and the cycle continues.

We've had a couple GE "specialists" come to the site and basically didn't know a whole lot, the one recommended we replace the 20CS so we did and got a brand new module with the 33HR included/installed (so assuming it's calibrated properly) and from what the operators tell me it worked perfect for about 1.5hrs after the new one was installed and now is failing just like before.

We don't have any pressure transmitters or anything on the lines, all we have is a gauge on the input prior to the FH1 filter, at startup it bounces around 500+PSI and is fairly erratic, then at the end of the cycle it shoots up to 1350 where the VR5 pressure release is set at and slowly decreases to zero.

Currently, we've installed a block to hold the jaws in place so they can't retract and this works to keep the ratcheting going, but it means when we want to fire up the unit we have to send someone out to pull the block... PITA. But when we do have the block in there is easily enough hydrolic pressure to turn the shaft.

So, I'm not sure where exactly to start. We checked the 33CS switch and are definitely getting a signal from it and reading the few manuals we have we set the switch to activate about 2/3rds into the clutch teeth engaging, and we can see 33HR switch off/on during the cycle...

I know my details might be a bit vague but bear with me this is the first ratcheting gear I've ever looked at plus first time on a Woodward system and have only worked with both for about 6 hours total!

How exactly is this gear supposed to function? I expected to see the teeth close, the gear turn forward locking the teeth, then at the end of the cycle maybe retract slightly to release the teeth but not enough so the cylinders would push the jaws completely apart, then stop and wait 3 minutes to repeat the cycle? Is that somewhat accurate or am I missing something
 
When did the problem start? It's presumed (because you didn't say so) that it was working fine and then this problem started. After a maintenance outage?

I'm presuming the ratchet mechanism uses a self-sequencer, but you didn't say that. The self-sequencer controls the flow of oil to the ratchet pistons. This discussion presumes a Frame 5 or -6 turbine with a self-sequencer. Some very old Frame 5 and -6 turbines and some Frame 7E/EAs have a slightly different type of ratchet system that doesn't use a self-sequencer, but we don't know what kind of unit and/or system is in place.

What is the position of the manual valve which controls the flow of hydraulic oil to the jaw clutch engaging cylinders? It's usually at the base of the Accessory Gear Box, near the Hydraulic Ratchet Pump.

The normal sequence of operation is that when the ratchet operation starts 20CS engages to actuate the closure of the jaw clutch and the self-sequencer should also be "turning" the ratchet in the forward direction assisting with the engagement of the clutch teeth. In a Speedtronic control system, 33HR drives a logic signal, L33HRF, which should be a logic zero *during* the forward stroke.

At the end of the forward stroke (and that's usually signified by a brief pressure increase when the pistons reach the end of travel) 33HR should change state to indicate the ratchet has reached the end of the forward stroke. In a Speedtronic control system, 33HR should change state and a logic signal, L33HRF, should go to a logic "1" at the end of the forward stroke, and remain a logic "1" until the retraction stroke has completed.

When the retraction stroke completes (again signified by a brief pressure increase when the piston reaches the end of travel), 33HR should change state and the ratchet should travel a few seconds in the forward direction (to try to make sure the clutch teeth stay engaged when 20CS is de-energized) and then stop, and the three minute timer should start. (Actually, I've seen some control systems that start the three minute timer when the 33HR changes state after the retraction stroke is complete, so the timer is counting while the short forward stroke is completing.)

And at the end of the three minute timer, the process should repeat.

I've seen a similar problem (not exactly the same) when the manual valve (used for testing) was accidentally kicked out of position and wasn't fully in either position. It's not labelled very well, is usually covered in dirt and oil, and most people don't even know it's there. It should be shown on the Starting Means Piping Schematic and it's position is important.

I have also seen similar problems when the self-sequencer isn't working properly and is failing or intermittent, but that doesn't explain why replacing 20CS would have a positive impact for a brief time and then start acting up again.

The filters in the hydraulic ratchet system are usually neglected. There's not usually a differential pressure gage or switch, and when they are allowed to get so dirty they rupture, a lot of dirt is released directly into the self-sequencer.

So I hope this helps; there are a lot of ratchet mechanisms and configurations out there. But, the basic principles of operation are the same: engage clutch during a forward stroke, keep clutch engaged during the retraction stroke, start a forward stroke to hopefully keep the teeth engaged during the three minute off time. Then repeat. The big difference is whether or not there is a self-sequencer controlling the flow of oil to the ratchet pistons or whether the turbine control system is being used to detect ends of strokes and changing flow directions (usually via a solenoid).

It's not uncommon for the turbine shafts of Frame 5 and -6 machines to roll "in reverse" at the end of a forward stroke (because of the backlash in the reduction gear), but the intent of "bumping" the ratchet in the forward direction for a brief period after the retraction stroke is to try to use the slope of the teeth to keep the clutch halves engaged during the three minute off time.

20CS should be keeping the clutch halves engaged during the ratcheting operation. I have to wonder if the oil flow to the ratchet pistons is stable enough, and if 20CS is really keeping the clutch halves engaged during the ratcheting operation (forward and retraction strokes). If the pressure builds up and shaft moves forward and the pressure dies off quickly and then the backlash of the generator shaft through the load gear causes the turbine shaft to "reverse". But, I believe 20CS should be forcing the clutch halves together all the time. So, that manual valve's position is still questionable. And so is the flow through the hydraulic ratchet pump.

I don't have a Speedtronic elementary to look at so I'm not really able to comment on what 33CS does during the ratchet operation. If it's changing state during the ratchet operation, then it's probably contributing to the problem. But, I believe you've probably adjusted the switch correctly. There should be a bit of a "lag" between when the switch changes state as the clutch halves are engaging, and when it changes state as the clutch halves are disengaging. In other words, the switch should change state at a different physical place when the clutch is closing and when it's opening, and it should have to travel a little farther "out" when opening to change state than it did when it closing, to have a little "wiggle room" in the engagement. But, 20CS should still be holding the clutch halves together during the ratchet operation, in my opinion and recollection.

But, it would still be nice to know when this problem started. After a maintenance outage? After some cleaning in the area of the hydraulic ratchet pump and manual valve? And, again, this is all presuming a Frame 5 or -6 with a self-sequencer, and a DC hydraulic ratchet pump, for the most part.

It's an unfortunate fact that many companies employ very young, in some cases recent college graduates, as field service personnel. (They'll travel because they haven't started a family yet and don't have a home to pay for and keep up, and feel they don't have a lot of experience to work in a factory environment--for a few months.) The training they get is good, but not great, and usually is only oriented towards the new units, and for GE that's F-class units these days, and the older equipment is just barely mentioned. But, this problem (young, inexperienced field service people) isn't peculiar to GE, but people just love to bash GE. Especially at the labour rates they charge. Try to imagine what it's like to be sent out to a site to work on something you know very little about but are expected to be an expert on because you're from the original equipment manufacturer. It's not a comfortable situation. (And it's not their fault. It's screwed up that people with experience feel they're too good and too valuable to travel. Its really bass-ackwards.)

Last, please write back to tell us what resolves the problem. "Feedback is the most important contribution!"(c) here at control.com. It's what makes being able to search past threads helpful when others can see what worked and what didn't.
 
Sorry for the late reply, been on the road a few thousand miles this last couple weeks to various plants.

Anyway to fill in some gaps, talking with the operators it's been working perfect since I believe they said '87 or something like that so either way quite a large number of years. The only recent work done in the area was on the diesel directly behind the gear but it was minor work so shouldn't have anything to do with the sequencer plus it didn't start immediately after. They have a maintenance schedule that they change the filters on the oil system every year (or so they say, other operators at other plants say they are a little lax with their maintenance) but I guess this last outage there was troubles on the grid so they only were given a 4hr window to do everything and basically the filters were not changed.

Also, the old guy says this is a GE Frame 6

Now, this is all that was done prior to failure according to the operators. The electrical operator said he noticed the problem first when there was issues with the DC Pump motor for the hydrolics, so he changed the brushes on the motor... then there was a problem with the 33HR switch they found it only picked up every other click so they changed that... then with the problem still there they changed 33CS just in case even though it tested fine, the pinouts for the NC/NO contacts were different for the new switch but they do note that in the book and swear they check/rechecked it...

So, after changing everything and nothing worked they brought in a few guys locally who basically played around with 33CS but it didn't work, then brought in 2 GE pro's one of which said replace the sequencer itself, they did that to the tune of I think $15K, worked perfect for 1.5hrs then now it's screwed up again.

One thing I will say is from what I gathered from the operator and this boggles my mind... they didn't change the filters yet again even when they changed the sequencer... so I mean you have it all apart, but not change the filter 8" off to the side of the new sequencer???? ... ...

Anyway, hopefully that's not the issue just a blown out filter and contaminated oil...

So that's where we left it since we were needed elsewhere. But I did compile a video you can check out which shows the pressure, the countdown to starting, 33HR contact, and also the gear itself and how it's operating. The video is kindof crappy and so are my editing skills but it's worth a look. The videos are very close to in sync. And the pressure gauge, each major mark is 500PSI, it's to blurry to read in the vid.

Let me know what you figure from what I've said and the video... I still find it strange the teeth move backwards so much at the start THEN go forward for quite sometime, then backwards again. It's like the sequencer is out of sync?

http://s29.photobucket.com/albums/c267/BigTom209/?action=view¤t=2-18-201011-30-23PM.flv
 
Another thing I just found out and should mention, originally this was installed with a speedtronic system but it's been changed out with the Woodward going in so it's no longer there. And looking over the logic we also use the NOT of the 33HR signal so if you're watching the video remember that. I was reading over your comment when it mentions it should be a logic 0 during the forward stroke where as ours was a 1... but we used the NOT so it should be alright.

Here is a few other pictures, the schematic and actual install which might be useful

http://i29.photobucket.com/albums/c267/BigTom209/Meadowlakeschematic.jpg

http://i29.photobucket.com/albums/c267/BigTom209/P1000155.jpg

http://i29.photobucket.com/albums/c267/BigTom209/P1000157.jpg
 
Some character isn't "translating" correctly in the video reference (the one between view and t), so I can't see the video.

Most of these ratchet mechanisms sort of "pulsed" during the forward stroke. They almost "hum", lightly then deeply, then lightly then deeply, continuously and very periodically during the forward ratchet stroke. If one has a DC clamp-on ammeter on the Hydraulic Ratchet Pump Motor lead the current can be seen to fluctuate with the same frequency.

If you're looking at the clutch during the forward stroke, it rotates a little then stops for a brief moment then rotates a little again then stops for a brief moment, in unison with the pulsing humming sound. It always appeared to me that the pressure from the pump had to build up to overcome the inertia of the turbine rotor and then when the rotor moved (and the ratchet gear moved) the pressure would decrease slightly and then start increasing again until it moved just a little again. This continues for approximately 1/8th of a rotation (more or less).

This continues until the forward stroke is complete, and then the pressure builds up very high when the forward pistons reach the end of the stroke and can't move and then the self-sequencer changes the direction of flow and actuates 33HR (I think that's what happens to 33HR). The flow is switched to the retraction piston and when it bottoms out the pressure builds up very high causing the self-sequencer to switch the flow back to the forward pistons and de-actuates 33HR, and then the "pulsing" humming starts again and the ratchet moves another 1/8th turn (approximately) if you're holding the manual jog push-button (43HR) in continuously through a second ratchet forward stroke.

All during this 20CS should be picked up, directing flow through VM2 to keep the jaw clutch halves closed.

VM2 is in the lower right corner of the first .jpg and it appears to be in the right position for jaw clutch operation (from my aging memory; watch who you're calling the "old guy"!). When it's in the other position it will not allow oil to the clutch closing cylinders, CS1.

Which brings up the question: What happens if you move VM2 to the opposite position and wait for the ratchet to operate? The ratchet side of the clutch should still rotate, but 33CS won't be actuated because the clutch halves won't be forced together. Does the clutch still rotate backwards?

I've never seen the jaw clutch halves rotate in the opposite direction during a forward stroke; that's just very, very odd. I've seen the inertia of the generator rotor, through the backlash of the load gear, cause the turbine shaft to rotate backwards, slightly disengaging the teeth, but that's the reason that before the ratchet stops (automatically) it's supposed to move a little bit in the forward direction to try to make sure the jaw clutch teeth are/remain engaged so that when 20CS is de-energized the clutch halves remain engaged. But, I've only seen this phenomenon on a very old Frame 5 on a 50 HZ application.

I have to believe that if replacing the self-sequencer (VH-14, with 20CS) worked for a brief period of time that the new one is "gummed" up or fouled up and isn't working again. Especially if the filters haven't been replaced recently. The fact that the brushes were replaced on the Hyd. Ratchet Pump Motor also rings a bell for me. It's another thing that has "changed". Has anyone used a DC clamp-on on the ratchet pump motor to see if it's drawing anywhere near rated current during the forward stroke? It could even be a problem with the ratchet pump itself.

There's also a relief valve, VR5, which might be causing part of the problem, though why replacing VH14/20CS would solve the problem for a brief time is still a mystery.
 
Hi,

We have the DC motor on our ratcheting system (Frame 5).

I read whole the post briefly and I'm not sure if the problem we had is similar.

So, we experienced the problem with the ratchet system that was working for some time at the beginning of the cool down, than after a while, about three cycles it stopped working leading to trouble alarm (on all four turbines we have).

It was found that the shaft was jamming (we operate in desert conditions, high temps at summer).

The solution implemented is to operate the ventilation fan during cool down for 3 minutes only and then stop it for 20 minutes. If temp. goes too high the auxiliary fan motor kicks in and stops below certain temp. value (feature also added).

We were also checking everything, pressures, 20CS, 33HR, having no clue what was the real cause. Surprisingly it was not being equally cooled hardware. Forgot to mention, we observed the shaft moving a little in reversed direction too.

Do you have the problem right after the shutdown? Have you tried the ratchet sequence some days after the cool down?

Regarding the DC motor, we observed 3.5 Amp during forward stroke having no trouble and the highest 5.5 Amp before we got jammed.
During the reverse stroke the current is lower 2-2.5, can't remember exactly (not have it recorded like I have for the forward stroke).

L33HRF - in the MKVI here the signal is called HYDRAULIC RATCHET IN FORWARD STROKE and is also indicated by logic 1. As far as I know it should be indicated with logic one as per it's description. If something is to be indicated by logic zero, then Z letter is added at the and of signal name eg. L33HRFZ. (check the starting command for Emergency Lube Oil pump wich is driven by open relay, you will find L4QEZ).

If the problem you have is not related with then sorry for disturbing.

This post may help other sites having similar problem to ours.

Regards
minister
 
Minister,

Temps maybe somewhat of an issue, but not overheating. We operate in Northern Canada and have ranges from -50C to +45C (or -58F to 113F for the Americans in the crowd) and right now we're in somewhat of the heart of winter, all be it a mild one we're still having to actually heat the turbine bay/oiling system to ensure we can start the unit(s). So our temps are very closely monitored and regulated especially in winter

And I'm not sure I must of made a error in my comments, the rotor shaft is not turning/reversing at all. The only part of the jaws that is turning is the ratchet side. If the ratchet jaws actually lock together there is enough pressure to easily turn the rotor shaft though. Sorry for the confusion

I'm not sure why the video link didn't work before? Try it again, if I copy/paste it into IE seems to work fine?

http://s29.photobucket.com/albums/c267/BigTom209/?action=view¤t=2-18-201011-30-23PM.flv

Anyway, keep the ideas coming! I'm pretty sure I'll be making another trip up to the plant so the more tidbits I pick up to test the better. I've even got the engineers who designed/built the 20CS reviewing the videos I made and so far we're all coming up stumped. I have a feeling it's going to end up being some minor hidden issue that's causing all the grief... just have to find it
 
The clarification should help you a lot. If it's not the turbine side of the clutch that's reversing or rotating backwards, then it is most likely a problem with the self-sequencer, VH-14, which includes 20CS and a plunger to actuate 33HR. Because the only reason that the pistons which are providing the force to the starting means side of the clutch should reverse direction is if the flow of oil from the self-sequencer is not constant.

You have said that when the self-sequencer was replaced that the operators reported the system worked properly for approximately 1-1/2 hours or so, then started acting up again.

I wouldn't rule out VR-5 or the pump. The brushes were changed; shouldn't be a problem, but has been many times on DC machines. The pumps do wear out, and a check of the current drawn during forward strokes would help, especially if you could get similar data from another site (I believe you said there is more than one turbine and/or site).

There's still an odd character in the URL in the posting (perhaps the control.com moderator could help with that, please?). I receive email notifications of posts, and the odd character is there, as well, though not the same character as in the control.com version. There appears to be a Euro symbol between view and t about two-thirds of the length of the URL. I've tried deleting everything after BigTom209, and then Photobucket asks me for a guest login password for BigTom209's account/site. When Photobucket does this, it replaces the Euro symbol with %80. I've tried MS-IE, Mozilla Firefox, Google Chrome, Opera, and Safari, and the URL doesn't work with any of them.

But, something is causing the self-sequencer to port oil to the ratchet pistons to cause it to rotate in the opposite direction. I would still like to know what happens when VM-2 is put in the "test" position and the ratchet is operated. When VM-2 is in the test position, the starting means side of the clutch should rotate relatively smoothly (no pulsing) to the end of the forward stroke, then stop while the pistons are retracted (the "ratchet" action which should prevent the clutch half from turning), and then start rotating in the forward direction smoothly again.

If the rotation is smooth and doesn't reverse when there is no load, then I would suspect the problem to be with the self-sequencer. I believe it's the high pressures which occur when the pistons reach the ends of their strokes (forward, then retraction) that cause the self-sequencer to switch flow to go from forward to retraction to forward. So, when there is no load (because the clutch is not engaged when VM-2 is in the test position) if the self-sequencer works properly, but when there is load when the clutch is engaged (when VM-2 is in the normal position) then I would suspect the self-sequencer. But, if the pump/motor is having a hard time when the clutch is engaged then that might also be a problem--not enough flow causing the pressure to build up and causing the self-sequencer to change flows.

If the starting means side of the clutch reverses when the clutch is disengaged (when VM-2 is in the test position), I would still suspect the self-sequencer.

I'll admit, I've never had one of the ratchets apart or even looked at an internal drawing other than the schematic representation on the Starting Means Piping Schematic drawing, so I don't know how the gears disengage during the retraction stroke. But if the disengagement of the gears a function of reaching the end of the forward stroke, and the oil being ported to the pistons changes direction before the piston reaches the end of the forward stroke, then it would seem possible that the clutch could be rotated backwards. But this is just a theory on my part, again, because I've never seen a ratchet disassembled or a drawing of the internals; never had a problem that required tearing into a ratchet mechanism or looking at the manufacturer's drawings of a ratchet mechanism.

By copy to minister, be very careful about the generalizations with regards to the use of the 'Z' character in signal naming. 'Z' has several meanings; it can indicate a timer (time delay on energization); it can indicate a "drop-out-to-run" condition (which requires a closed contact, normally-closed contact of a de-energized to energize a motor starter; and in some cases, it has been used to indicate an "inverse" sense (but that is usually very limited and not considered good programming practice). The units in C-eh-n-eh-d-eh (a small bit of humor; sorry, couldn't resist) have Woodward controls, which may or may not use the same signal-naming conventions. There's also that pesky "inversion mask" thing that confuses the daylights out of many people that can ruin a person's day and understanding.
 
In reply to CSA: I suspect the video URL problem is a character encoding problem. Microsoft uses a proprietary character set (Windows 1252) which is *almost* but not quite the same as the standard ones (ASCII or ISO UTF-8). The euro character happens to be one of the differences.

As long as you are only exchanging data directly between computers that use MS Windows, you don't notice the problem. As soon as you go out on the Internet where it has to try to inter-operate with the rest of the world however, it's a disaster. When you cut and paste or copy data, you can end up with different characters as your software tries to "guess" what the character really was meant to be. There may be nothing that the Control.com operators can do if that is the actual byte stream being sent to them.

The following is the URL that I tried, and it took me to a web page where the video supposedly is (BigTom209's Default Album). I say supposedly, because I don't have Flash installed, so I can't play it anyway. Try the following URL (remove any spaces from it the forum might insert):

http://s29.photobucket.com/albums/c267/BigTom209/?action=view&current=2-18-201011-30-23PM.flv

You should not see a euro symbol anywhere in that URL. This bypassed the log-in screen, so I assume it is working OK.
 
http://s29.photobucket.com/albums/c267/BigTom209/?action=view&current=2-18-201011-30-23PM.flv

http://vid29.photobucket.com/albums/c267/BigTom209/2-18-201011-30-23PM.flv

Yeah hopefully we can get this video to work, much easier to understand what's going on when you can see it in a video opposed to trying to describe it. Talking with the guy who built the sequencer they're thinking there maybe a problem with the ratcheting gear itself and the Fs or Rs line isn't providing enough pressure to switch the internal valve... so I'm trying to get the Engineers here to approve installing pressure gauges on all 4 lines going to the ratchet... then we can see what F/R/Fs/Rs are all doing.
 
Thank you, M. Griffin! I tried several substitutions, but didn't think about the good old ampersand.

As for the Flash thing, is it a security concern or a space concern or a memory concern? Or just that you don't visit many sites on that particular PC that require Flash? (I'm just curious, by nature.)

Back to the bigger issue at hand, I've never seen a ratchet run backwards before, so this will be a first. I keep going back to the fact that it was stated that after the self-sequencer was replaced (in my mind, the self-sequencer is the "block" with 20CS installed on it and the plunger which actuates 33HR) that the ratchet worked okay for 1.5 hours or so.

I also don't know where the pressure gauge was installed.
 
In reply to CSA: You asked why I don't have Flash installed. The reason is simple. There little to nothing that I want to look at on the Internet that actually requires Flash (e.g. I can just download Youtube videos), and the Flash plug-in is probably the single biggest cause of web browser problems. So, why bother with it?

In the long run, things like Flash (and Silverlight) don't seem to have much of a future. Newer web browsers can play video directly in the browser without it. When the video web sites finish converting over to using video tags, that will leave mainly just ad delivery, and who is going to install troublesome plug-ins just for that? I won't be at all surprised if Flash (and Microsoft's Silverlight Flash clone) are dead and forgotten five years from now.
 
M. Griffin,

Thanks! I don't seem to visit many websites that require Flash, myself, because every time I do I seem to need a Flash upgrade.

And the one I needed to watch this ratcheting video was for a 'Security Concern'.

This is the nature of technology, though, isn't it?

Thanks!
 
You seem to have obtained a good deal of information from the self-sequencer designer that would be very useful to myself and others.

Can you take a few minutes and summarize some of the details of how the ratchet mechanism works, how the self-sequencer interacts with it, and what the four tubing lines are and do?

Thanks; it would be very much appreciated!
 
Well I've learned a fair bit just by studying the schematic, reading what limited info we have on the system, and talking a bit with the designers/builders of the 20CS sequencer. I'm assuming the link to the picture of the schematic worked? Seems like this website put a space at the end "schematic" so maybe some people were not able to see it, but here is the link again just make sure no spaces are inserted when you copy the link

http://i29.photobucket.com/albums/c267/BigTom209/Meadowlakeschematic.jpg

Anyway from what I've read it's very simple, the sequencer runs off 120VDC which feeds the contacts as well as a coil pickup which on the side of the sequencer is the silver cylinder sticking out the side, when the pressure builds up enough the coil picks up opening the port letting oil flow to both of the pistons (CS1) which in the video you can see close the jaws together, and also feeds oil to both the shuttle valve and reversing valve (internal) and these valves are controled by the "pilot operator" which is kindo f like a switch shown ontop of the valves in the schematic. At first oil should flow through the reversing valve out the F (Forward)line driving the jaw forward, once the stroke is completed the Fs (Forward Stroke) line should build up enough pressure (200PSI+) to drive the pilot operator on the shuttle valve which in turn controls the pilot operator on the reversing valve, switching the oil to now flow through the R line now reversing the jaw... this switching of the reverse valves pilot operator also triggers the 33HR contact, once the reverse stroke completes the Rs line should build up over 200PSI and switch back to the forward stroke again shifting the pilot operator on the shuttle valve, in turn shufting the pilot operator on the reverse valve triggering 33HR and now switching back to the forward stroke.

From what I've read in the documentation for our plant the forward stroke is upwards of 10 seconds long but the reversing is only a couple, basically to release the pressure on the teeth but not allow them to completely come apart, then we again switch to the forward stroke for probably only a second then kill the power. This way next cycle we're already in the forward stroke.

We need to make sure the teeth don't completely separate since those 2 pistons are spring loaded so if the teeth separate like in the video the pistons push out and disengage the jaws, we need there to be some slight overlap so they stay together.


So that's how I understand it works. What's inside that hydraulic ratchet? No idea, we have zero information on it... not even sure who makes it or how it comes apart. But, if there is problems internally and those Fs or Rs lines aren't functioning properly or whatever then yeah it's going to cause problems.

We (the controls/electrical) have suspected that ratcheting gearbox from the start.. but the mechanics swear up and down it's our problem. I'm not sure how it works but I'm guessing it's a big job to remove/inspect?? It's got a one way clutch inside I believe so when running on the diesel/hydraulics it locks but once the turbing speed overcomes the diesel/hydraulics it can break free triggering 33CS and the control system shuts down the sequencer/diesel

So... anyone know how to inspect that big ratcheting gearbox???
 
I was going to reply right away but was waiting to make sure my solution was the correct one. I spoke again in great detail with the guys from Young & Franklin about that sequencer and figured it's got to be functioning like it should based on all the videos I took and looking at the 33HR tag switching etc... it is functioning. We know the 33HR and 33CS switches are good and setup properly, the control system was working... it's got to be inside that hydraulic ratchet. So I went digging in the library and found the binders for the station and the service manual for the Twin Disc torque converter (with ratchet installed on the side)

Basically how it works either the diesel or the hydraulics has the ability to turn the jaws, the hydraulic is designed I guess mostly for cooldown period as well as startup to break the static friction for when the diesel starts, our system has been changed so we ratchet all the time every 3 minutes except for startup where we set it to continuously ratchet until the diesel takes over and spins the rotor up to speed.

Anyway now I had the part number for the company and started calling all the service departments listed on the Twin Disc site, eventually after many confused phone calls I found a old service guy with 40+ years exp and who's worked on these before. I mentioned our symptoms and sure enough he said it sounds like the inner cam is worn out and now allowing it to freewheel which is why in the videos you can see the jaws move backwards which they should never do.

So, a brief run down of how that works if you look at the hydraulic schematic there is the F and R lines that goto the right piston as well as the Fs and Rs. The sequencer through the reversing valve which is controlled by the shuttle valve can put pressue either way, and at startup idealy it pressurizes the Forward line driving the piston down, at the end of the travel I'm not sure exactly which but the piston moves out of the way and signals either the FS or RS line indicating the stroke is finished, this switches the shuttle valve in the sequencer, which switches the reversing valve, and therefore changes the stroke to push the piston in the opposite direction.

This piston in the center is geared and as it moves up and down there is a center circular one way cam that acts on the torqe converter, this cam as it rotates there is an inner cage that holds 22 or so rollers, as the piston rotates it rotates this cage which moves the rollers up the jawls and wedges them into place locking the cam and turning the torque converter. Once the forward stroke is finished and the reverse starts the cage is spring loaded so the pressure moves off the rollers and they roll back and basically freewheel for the reverse stroke so the shaft doesn't turn.

Clear as mud?

Anyway I spoke with a guy who had a Frame5 unit and problems with his gear and he sent a bunch of really good pictures showing the guts of the system and how it all works, hopefully again these pictures will work

http://s29.photobucket.com/albums/c267/BigTom209/Frame5/

Once we tore into ours we found the problem, one of the bars on the inner cage broke free and likely got wedged behind the roller and snapped it in half. The unit was full of metal and in pretty rough shape, I also noticed 5 other bars on the cage were cracked so likely not much longer we would've had even more issues

http://s29.photobucket.com/albums/c267/BigTom209/DamagedCam/

So, luckily a service tech within 8 hours away had a rebuilt cam squirreled away in the back room and we were able to get it in and installed within a few days. Likely we'll have to take a major overhaul of that torque converter cause judging by the horizontal play in the bearings likely that torque converter needs a complete overhaul as well... another big job.

So that's where it stands, I could go into more details but I think I covered it good enough :)
 
BigTom209,

Many thanks for the feedback and the photo's.

This is another great example of how blame is usually assigned to the control system because it's the least understood component, and with all those wires any problem must certainly be inside that panel!

Again, thanks for the feedback. I hope we were able to help you prove the problem was not in the control system.

You have certainly explained the ratchet system very well from a controls perspective. It's primary purpose is for rotor cooldown, operating once approximately every three minutes. And it's secondary purpose is to assist with breakaway of the turbine-generator shaft from zero speed during starting (that "inertia" thing, which usually is overcome with the "bump" from the ratchet during application of torque from the starting means (diesel engine or electric motor)).

Thanks again, also, for the pictures, and the video! It will be a great help for many people!

 
Top