Thread Starter


we can't put the GT F9 on Slowroll. the speed droop down until 0%. the 2 Solenoids checked. the sequence in MK V working well. we have installed 2 gauges one for the inlet pressure and one for the working pressure. the inlet pressure droop from 100 psi to 90 psi when the 20TU1 is energized. the working pressure remain 0 psi.

Any help appreciated
This discussion presumes that slowroll on your Frame 9E uses Aux. L.O. Pump discharge pressure ported through variable guide vanes of the Torque Converter to provide the force to keep the shaft rotating.

The two biggest reasons a unit won't stay on, or "falls off", slowroll are:

1) Low L.O. Header Temperature

2) Improper setting of the Torque Adjustor mechanism

If L.O. Temperature is less than approximately 25 deg C (approx. 80 deg F), the oil viscosity is too high and the slowroll apparatus (Aux. L.O. Pump discharge pressure through the variable guide vanes of the Torque Converter) can't overcome the friction to keep the shaft spinning and the unit coasts down to zero speed.

The other problem is when the Torque Adjustor mechanism isn't being driven to the "full" torque setting.

Another problem is when the 20TU-2 solenoid (the one which usually ports the Aux. L.O. Pump discharge to the Torque Converter) is leaky or doesn't move full stroke (usually caused by worn mechanism or seals or dirty oil). (20TU-1 is usually energized/de-energized to pressure and de-pressurize the Torque Converter during starting, coastdown from purging, during cranking, and during firing/acceleration.)

Finally, torque converters do wear out. And Torque Adjustor mechanisms do wear out.

Have you tried manually operating the Torque Adjustor, opening the guide vanes until the manual wheel won't turn any more? (This is usually past the "full torque" limit switch position, so don't use a lot of force on the handwheel!)

But, it's likely a combination of one or two of the above. And, the usual culprit is L.O. Header Temperature.
Thanks for help,

Really I don't know if the torque converter is working well. From the command panel we can LOWER & RAISE the T.C also from the MKV when it is on AUTO mode. Yes the T.C was fully open and we have manipulated the hand valve four rounds more. The Lub. oil T° was 47°C and the inlet pression is 110 psi. When the speed slow down the 20TU2 is energized at 90 rpm but the speed droop to 0%.

we still looking a round for any thing else.
It does sound as if the "controls" portion of the torque converter is working properly. Have you tried replacing 20TU-2? Also, there is usually a manual button on the end of the 20TU-2 mechanism which can be pressed (though 20TU-2 is usually not in a very accessible or convenient location!).

Does the unit seem to take longer to accelerate to purge speed and when accelerating after warm-up?

Most sites don't put pressure gauges on the torque converter until after they have issues, so they don't have any normal readings to compare to.

The only other issue I could think of would be a compressor or turbine rub, or even new, tight bearings or generator oil seals.

My bad for not asking when this problem started. Did it just start one day, or after a maintenance outage? If it was after a maintenance outage, what was done during the outage?

Slowroll doesn't really provide much torque to the turbine rotor (and compressor and generator--since they are all coupled together); it's really just, as you noted, something that's started when the unit is decelerating and adds just enough torque to keep the shaft spinning slowly (usually between 30 and 50 RPM or so). If you plot a normal shutdown below approximately 100 RPM, you will likely see that the speed will dip slightly and then "recover" and increase slightly (just a few RPM) as inertia is overcome by the torque provided by the "slowroll" apparatus. If there's any resistance to rolling friction from any portion of the unit then the unit will "fall off" slowroll.

Have a look at the Starting Means P&ID, Piping Schematic. 20TU-2 is energized when the unit is to be on slowroll which directs oil from the Aux. L.O. Pump discharge to the variable guide vanes of the torque converter, which should be near or at maximum opening to produce the highest torque output.

There might even be a possible problem with the Aux. L.O. Pump. You should be able to use a clamp-on ammeter to measure the current drawn by the Aux. L.O. Pump motor and compare that to nameplate to see if the pump is working properly. Hopefully you have some normal running data to compare against, but if not, it's common for the Aux. L.O. Pump to be drawing near rated current when the unit is on slowroll.

If new bearings were put in the machine during a recent maintenance outage, or if the unit has recently experienced a high vibration event, or if new compressor blades (rotating or stationary) were installed, or if new turbine buckets and shrouds were installed, these could all cause rubs which could cause a unit to fall off slowroll. Hydrogen oil seals improperly installed have also been know to cause problems for slowroll (and resulted in damaged seals and shafts!).

Lastly, if there has been some recent change to the turbine compartment vent fan set-up or some dampers are not working properly and are allowing cool air into the turbine compartment when the unit is on cooldown after fired operation then the compressor casing could be cooling faster than the compressor rotor which could also result in a rub. It's even been found that someone opening turbine compartment enclosure doors during cooldown, especially when the unit has just been shutdown, can cause compressor rubs if the ambient is cool enough.

You haven't said how old your torque converter is, or when the last time any maintenance was done on it. Again, they do wear out and require refurbishment. They also do break. Voith are fairly quick to respond to calls for assistance.

How old is the turbine L.O.? I think there is a strainer in the suction of the torque converter, but that's probably not used when the unit is on slowroll.
I have similar case to share. But I believe my case is not so bad as compared to BB.

One of my f9e with Mark V has the problem of failure to breakaway from standstill after the unit successfully put into off cool down.
My TU1 and TU2 solenoids were working properly. The torque adjuster actuator was set properly. Even the 'high torque' starting was tried. Nothing seem to work. The A.O.P pressure was normal at about 8bars. The TC discharge pressure gradually did build up until 14bars and maintained there but the unit still failed to break away. Usually this discharge pressure would only maintain for few seconds and the unit would breakaway from standstill but in my case it does not happen. What was done after that was manual jacking rotation of the rotor at the load tunnel. After about 180deg of turning, the unit was successfully break away.

Evaluating on where the problem is quiet troublesome as every time we experience this problem, the angle of manual jacking rotation is not the same (i,e. sometimes is 180deg, 360deg, 90deg). Sometimes it takes about 1 whole day to put the machine on slow roll.

My bet until today is on the torque converter itself. Lets see what turns up next ;)
Hi all,

On the name plate of 20TU1 and 20TU2, the max pressure rate is 3000 psi. when we checked on a running unit with the same T.C, the max pressure rate was 5000 psi. So now we are waiting for 2 new solenoids then see if we get the slowroll.

this unit was stored 10 years a go, we achieved a major inspection just for the turbine. no thing was done on the generator. The generator has a Hydrogen oil seals. the Lub oil pump is a new one.
Regarding the problem, after discussing with a lot of people who claims that the problem might not be with the torque converter but with other parts on the train, I thought of a small test.
What if we uncouple the torque converter from the acc gear box (mine is frame9e driven by cranking motor) and initiate a slow roll command from Mk V?

I believe if the torque converter is working well, then the output shaft of the torque converter should rotate. If the torque converter is faulty is any way, then there should not be any rotation of the output shaft (please correct me if I'm wrong).

Any one conducted similar test as this or is this test workable to asses the functionality of the torque converter?

Is there any other matter to look into when doing this test or any other test that can be done to verify on whether the torque converter is faulty or there is something else hiding from our vision? ;)


You're not the original poster, are you? You've just added your problem to this thread, haven't you?

You say this is only happening on one unit at your site. You say you've increased the torque adjuster setting to "maximum". You say you can get the shaft to break away from zero speed by using a jack to assist with the torque being provided from the starting motor through the torque converter.

You haven't told us when this problem started occurring. After a maintenance outage? Suddenly?

HAVE YOU OBSERVED THE CURRENT(S) DRAWN BY THE STARTING MOTOR DURING THE START ATTEMPT? They should typically be in the range of 150-160% of rated motor nameplate current draw. So, if the starting motor nameplate rating is 98A per phase at 4160 VAC, then the current draw, per phase, during break-away attempt should be in the region of 150-160A. Each phase should be drawing nearly identical amperage.

To answer your question: No. I don't believe that by uncoupling the torque converter you will prove or disprove that the torque converter is working properly or not. With no load on the torque converter, even if it's not working it will likely still rotate.

Doesn't it tell you that if you provide an assist from a jack to help overcome the inertia of a rotor at rest that either the torque converter isn't working properly or there is something mechanically binding the shaft?

It's likely that the reason differing amounts of rotation when using the jack are required is that sometimes there is enough of a "bump" from the jack to help overcome the inertia of the unit shaft, and at other time there isn't enough of a "bump". In other words, if you could get a faster stroke from the jack it probably wouldn't take nearly as long a stroke to get the shaft turning.

You haven't told us how old the torque converter is, or how long since it's last maintenance. Have you checked the strainer in the suction of the torque converter? The valve in the suction line of the torque converter to see if it's fully open?

But, no; uncoupling the torque converter isn't going to tell you much. If the unit can stay on slowroll after shutdown, and then just can't break away from zero speed after it's taken off slowroll (when the shaft is properly cooled), then it's either a starting motor problem, a torque converter problem, a 20TU-1 solenoid problem, or a mechanical binding problem. It's as simple as that.

I witnessed a test once of an uncoupled torque converter on a Frame 7E machine. When 20TU-2 was picked up, the output of the torque converter rotated in the OPPOSITE direction that it should have from a zero speed condition! It was possible to manually stop the torque converter output coupling (with gloved hands!), and then start it rotating in the correct direction (with one's hands!) and it would then continue to rotate in the proper direction. All of this was done only with 20TU-2 energized (the "slow-roll" solenoid). And, the output rotation was very slow, less than 50 RPM. With the Aux. L.O. Pump running.

When the torque converter was re-coupled and the unit was started, everything worked fine! Starting, cranking, slow-roll, everything.

This was done during a major inspection, just out of "curiosity" because there was some time available and we were bored. We did NOT energize the starting motor with the torque converter uncoupled (we had only tied back the half of the coupling with wire and had not removed it completely and were worried it would fly off and hurt someone or something if spun at a couple thousand RPM by an 800 HP starting motor!).

Let us know what you find out.

Yes I'm not the original poster for this thread but I'm having similar problem.

This problem started after about 3 months following my hot gas path inspection last year when we replaced the torque converter with refurbished unit. When the unit was commissioned after the hot gas path inspection, we failed to get to breakaway from standstill twice. But the third attempt was successful and nothing else was done. The headache only came during the offline wash few months later ;)

The current drawn by the starting motor for this unit is enormous at 200A. The normal amperes during rotor breakaway is about 150A only. The amperes during unit cranking also remains at 200A throughout the cranking period. The current in all phases were identical.

I cant think of any reason that may cause the shaft is mechanically stopping the shaft from breaking away from standstill as I cant observe any abnormality from the bearing metal temperature, lube oil temperature, bearing housing accelerometers and also proximity probes. Say the rotor is bow in any way, I'm sure that I could observe it from proximity probes.

I'm not sure on how old is the torque converter but before the replacement with the refurbished unit, everything was fine without any problem. The working unit was also sent to refurbishment and we did also tried to replace it recently but the result is the same. My only concern is that both torque converters were refurbished by the same non OEM local company.

The strainers are fine and both 20TU1, 20TU2 were also replaced with new spare although both of them were working fine. The unit was also only put into OFF COOLDOWN after the wheel space temperature permissive were met.

To add into discussion, I have 2 pressure gauges installed at the torque converter. I'm not sure on where the internal tap point for this pressure gauges as it was a modification work done after the unit initial commissioning very long time ago. However from what I can observe, the reading of the 1st pressure gauge is the same as the AOP discharge pressure at about 8bars. I assume that this would be the pressure of the incoming supply inside the torque converter. The reading of the 2nd pressure gauge will only go up after 20TU1 energizes and will build up as the torque adjuster motor operates until about 12-14 bars and this reading would drop sharply when the unit successfully breakaway. This would only take about less than 5-10 sec. At my condition now, this 2nd pressure gauge reading maintains at 12-14 bars until the breakaway failure which comes very much later.

As you suggested that uncoupling the torque converter could not prove anything, I would abandon this test. As the last resort, I think swapping the torque converter from another unit would be a good idea. However this could not be performed at the moment due to outage time limitation ;)
So, this isn't a new torque converter, it's rebuilt; it wasn't rebuilt by the OEM; and the problems started after this torque converter was installed after a hot gas path inspection, which shouldn't have affected the bearing conditions (presuming it was a typical hot gas path inspection). You seem to have said you sent the original torque converter to the same non-OEM shop for refurbishment, and then you re-installed it in place of the refurbished torque converter but had no better results.

I would imagine that it might be possible to draw excessive current and still transmit less than normal torque, but it would seem difficult. It would surely seem that something is wrong with either the method and/or parts used in the refurbishment or the method and parts used in the installation.

Removing a known, working torque converter from another unit and installing it in this machine would likely tell you if you have a torque converter problem or a rotor binding problem. But, that would certainly seem to entail a lot of work, and if the installation process is suspect then you might be introducing the same problem with the change-out.

It would certainly seem to be more beneficial to have a manufacturer's representative visit the site, and cost-conscious, too.

I do recall seeing some sequencing that did NOT have the torque adjustor drive mechanism at full torque when 20TU-1 is energized during starting.

Have you tried manually putting the torque adjustor mechanism to full torque position before trying to start the machine?

There should be a several second delay between the time the starting motor is energized and 20TU-1 is energized. This is to allow the starting motor to come up to speed before "pressurizing" the torque converter. Some torque adjustor mechanisms had faster slew rates than others. If the torque adjustor isn't at maximum when a START is initiated, but does start traveling when START is initiated but isn't at full torque position when 20TU-1 is energized, then you might try increasing the time between when the starting motor is started and 20TU-1 is energized.

But, if you haven't done so already, the torque adjustor mechanism should be set to full torque (purge torque) setting prior to energization of 20TU-1.

I don't have any experience with exact pressures on torque converters. As has been said before, Voith is usually very helpful in my experience (at least in North America, Europe, and the Middle East). But, the pressure build-up (not the value but the existence of the pressures) would seem to be consistent with my understanding of how 20TU-1 works, to "pressurize" the torque converter to allow it to transmit torque.

Your purchasing/sourcing organization should be calculating the cost of the efforts you have made to try to make this non OEM-refurbished torque converter work properly. I'm sure it's enormous, considering that you've said you've spent half a day more in some cases trying to get the unit to break away from zero speed. Certainly, any monies saved over purchasing a refurbished unit from the OEM have been more than erased; probably by several multiples.

You didn't say what the starting motor nameplate rated current draw was, but 200 A sounds a bit excessive.

Have you contacted the shop that refurbished the torque converter?
Yes i did try to manually put the torque adjustor mechanism to full torque position before 20TU1 energizes, but its still the same. I'm not sure on what the next step that would be done on the torque converter refurbishment works. But if there is any new findings I will certainly put it up here.

wish me luck ;)
Doesn't sound like the local shop that did the refurbishment(s) is much help; hopefully, they haven't been fully paid yet for the work(s). I'm greatly becoming a huge fan of holding back 10% minimum, sometimes more, and not paying the money if the work is not satisfactory.

How about getting the OEM involved? (I think I probably know the answer to that question.)
Interesting discussion. Hope you don't mind me joining this discussion. It looks like there is no problem with TC or controls (for original problem posted). Can tight mechanical seals (or bearings) on any of the turbine bearings cause fall off to zero speed? I am asking this as it is only the L.O. pressure (around 110 psi) through 20TU-2 and torque converter keeps the machine on slow roll. Even little resistance from down train can cause fall off.
Responding to sd's 02-Dec-09 (23:18) post... repeating your observation...

"I cant think of any reason that may cause the shaft is mechanically stopping the shaft from breaking away from standstill as I cant observe any abnormality from the bearing metal temperature, lube oil temperature, bearing housing accelerometers and also proximity probes. Say the rotor is bow in any way, I'm sure that I could observe it from proximity probes."

Sometimes rotor "bowing" can manifest itself as a shift in clearance, but only if there are multiple (2 minimum) proximity probes at each bearing. Also, there might be a difference in lube-oil pressure at opposite-end bearings?

Q1) Was a lube-oil pressure difference noted between DE and NDE bearings?

Q2) Although difficult to attain, did you take any NDT sonic measurements at each bearing during breakaway?

Regards, Phil Corso
Hi all,

Still the slow roll problem persist, after the team on the site had installed a new T.C and de-coupled the generator.

Please can anybody comment on this?
Okay; now I'm confused.

BB started this thread, and sd added his problem. We haven't heard from BB in a while, and now BB is posting again. BB's last input was on 15 Nov, and he stated they were awaiting two new 20TU solenoids with higher max pressure ratings, and that the unit had been stored for 10 years, had just had a major inspection during which nothing had been done to the hydrogen-cooled generator, and the L.O. pump was new.

I think sd's problem was that he couldn't break the shaft away from zero speed with the torque converter. I think he could maintain the unit on slowroll, but when it was taken off and went to zero speed, it required a bump from a jack on the shaft to help get the shaft turning. sd's problems started after a hot gas path inspection during which a refurbished torque converter was installed.

I think BB's problem was that the unit wouldn't stay on slowroll. He said that the L.O. temp was 47 deg C and that they'd replaced 20TU-2. BB's problems started after a major inspection of a unit which had been stored for 10 years. They have manually opened the torque converter adjustor by four more turns. 20TU-2 is energized at 90 RPM, and still the unit wouldn't stay on slowroll.

So, if we're back on BB's problems, and I've summarized the issue correctly, now it seems that they have uncoupled the generator from the turbine rotor and the torque converter is still unable to maintain slowroll speed.

BB, please confirm or clarify if you are the original poster, and what the conditions were when the problem started, and what's been done in an effort to resolve the problem (from the beginning to now). Because, I'm confused!!!!

The torque converter, when being used to maintain slowroll speed, just provides enough torque to maintain some speed, usually 30-50 RPM. This is presuming the L.O. bearing header temperature (downstream of the L.O. cooler) is above approximately 25 deg C. BB stated the L.O. temp was 47 deg C, but that's awfully high for the bearing header temperature, and would seem to be more like the L.O. tank temperature.

I have seen MANY sites where the cooling water circulation through the L.O. coolers was maintained during slow-roll operation causing the L.O. bearing header temperature to drop below 25 deg C. The design of the L.O. Temperature Regulating Valve, VTR1-1, is such that there will always be some flow of cooling water through the valve, so it will never completely close in the event of failure. This design means that if cooling water flow is maintained after the unit is shut down, it's possible for the bearing header temperature to drop significantly below 25 deg C (regardless of the L.O. Tank temperature), making it very difficult for the torque converter to maintain slowroll. So, we need to know what the L.O. bearing header temperature is (usually signal name LTTH or LTTH1) when the unit is falling off slowroll.

If you've removed the load of the generator rotor from the "train" and the unit still won't stay on slowroll and the L.O. bearing header temperature is above 25 deg C and you've replaced the torque converter with a "...new..." one, then it would seem the possibilities are pretty limited:

1) Insufficient L.O. flow to the torque converter;

2) Insufficient flow of L.O. through the torque converter;

3) Excessive "drag" from the Accessory Gear or the turbine bearings;

or some combination of the above.

We haven't heard if the turbine bearing metal temperatures are normal or elevated, and we don't know what the turbine bearing drain temperatures are, and I'm talking about when the unit is at rated speed. Elevated temperatures would be an indication of tight bearing clearances.

You said the L.O. pump was "new". Is that because the unit was stored for 10 years, and is basically new, or because the pump was replaced recently with a new one? Are you referring to the L.O. <b>pump</b> or the L.O. pump <b>motor</b>.

Are you sure the "new" pump is the proper rating (flow-rate)? Have you recorded the current drawn by the L.O. pump motor when the shaft is coasting down below 200 RPM but before 20TU-2 is energized, and then below 90 RPM when 20TU-2 is energized?

But, again. I'm still pretty confused because of the two (apparently) different sites with problems. The title of this thread and the original posting was about a problem with a Frame 9E maintaining slowroll. Then, we got into a discussion about not being able to break a Frame 9E shaft away from zero speed.

And I'm not sure where we're at right now.

One last thing about slowroll. If you plot the speed of a unit coasting down to slowroll, you will see a dip in speed as the torque converter "catches" the shaft, overcoming the inertia of the shaft as it decelerates, and then slightly accelerating the shaft back to it's "equilibrium" speed (approximately 30-50 RPM, depending on L.O. bearing header temperature, and under normal circumstances). I believe BB said they had at least one other unit on site so this should be easy to confirm.

The point is that the torque being provided by the flow of L.O. pump discharge pressure through the torque converter isn't really very high. (This should be borne out by the current drawn by the L.O. pump motor on a unit which is NOT experiencing slowroll problems when 20TU-2 is de-energized versus when 20TU-2 is energized at about the same speeds.)

So, if the L.O. flow through the torque converter isn't high enough (either because the flow from the L.O. pump is restricted, or the torque converter adjustor isn't allowing enough flow through the torque converter when 20TU-2 is energized, or there is some restriction in the discharge of the torque converter when 20TU-2 is energized) or the mechanical forces in the accessory gear/turbine/generator are too high (tight bearings; cold L.O.; combination of the two) for the torque being provided by the flow of L.O. through the torque converter when 20TU-2 is energized), then the unit ain't gonna stay on slowroll.
hello all

I suggest you to inspect n#3 bearing from load tunnel. you should know n3 bearing (if tilting pad bearing is used) have the most break away torque and have an important roll in break away.
CSA: Dont worry. You are still on track. ;)

BB is the original poster for this where his unit could not go into slowroll. We are not from the same site as well. My problem is that my unit could not breakaway from standstill. No issues for getting unit to go into slowroll after cranking or shutdown. I'm very sorry to cause this confusion as I thought it would be better to add into this thread as both problems might be caused by same device/situation.

Phil Corso: Yes they are 2 proximity probes installed in each bearing and both readings are relatively small and same as the other unit. Sorry I did not manage to get the lube oil pressure difference at NDE and DE during the previous shutdown. There were no NDT sonic measurements taken at each bearing during breakaway as well. I will certainly look into these things during the next "opportunity" ;)
Hi all,

I'm sorry to cause this confusion. After a M.I, we tried to put the GT on Slow roll but without result. same lake sd [sic] problems.

- The T.C is new one and is working very well,the commands (LOWER/RAISE) from the local panel or from the MKV are working properly. this was also confirmed by one T.C specialist.

- The L.O pump and motor changed with new, the discharge pressure is 110 Psi.

- the 02 solenoids 20TU1X and 20TU2X are new,on the nameplate the pressure rate for these solenoids is 5000 Psi, and they are working properly from the MKV.

- the generator was uncoupled but the result is the same.

- the test was done with correct L.O temperature value (60°C)

this, is a summary about what was done up to date. The team on the site is still working on it.

Thanks and Regards,
How can you say that your problems are the same as sd's? He said his unit has no problems maintaining slow roll; his unit just can't break the turbine away from zero speed.

Your unit, on the other hand, has no problem breaking the unit away from zero speed. But, your unit won't maintain slow roll, even with the generator rotor uncoupled.

HOW are they <b>THE SAME</b>?

Your problem started after a major inspection. I think the issue might just be with the tilting pad bearing used in the #3 bearing, just as was mentioned earlier.

You still haven't responded to the questions about the bearing metal temperatures for the turbine bearings. Are they higher than before the major inspection, or lower, or the same? Because higher temperatures could mean tighter bearings which could mean more friction, which could mean the torque being provided from the new torque converter with pressure from the new L.O. pump is not sufficient.

And, if you've had a torque converter expert look at it and say there's nothing wrong with the torque converter, and it worked before the major inspection, then, it's about time the mechanical department step up and take responsibility and start investigating how the turbine was put back together after the major inspection. Because it sure seems like there is more "drag" than before the major inspection, even with the generator coupled!

And, you didn't cause the confusion, other by not responding sooner. sd caused the confusion by also thinking that his problems were similar (HOW?!??!!!!???) to yours. Because both the units are Frame 9Es? And they both seemingly involve the torque converter?

I need another drink.

But, I just had another thought. Is this problem occurring only when the unit is hot? Or, even with the unit is cold? (Probably even when the unit is cold, if you tried maintaining slowroll with the generator uncoupled.) Sorry. I need another drink.