Gbx before and after repair and now out shft to extruder lo speed hi torque. Shaft repair went bad? Other suggestions? Time to repair? Next January prefered. Olov

repair.doc (142 Kb, 180 downloads) Data Gbx

Oli,

Without knowing any of the details, I couldn't say much more than something is wrong.

In the after plating spectrum it looks like you have something at about 1070 hz with lots of sidebands. I would guess either 2nd harmonic of high speed gearmesh or more likely a gear natural frequency being rung by something running at the sideband spacing.

The current condition looks like you probably have 1, 2, and 3 x gmf with mountainous sidebands assumed to be running speed.

The current twf is pretty scary looking. I assume that when you expand the twf, it shows impacting at running speed and gmf ringing down. Without having anything to compare it to, I would say that it is very significant and that is a likely cause if your data indeeds supports it.
As far as how long to repair, there are too many variables to make that guess. (I say this after making huge guesses on the gears.)

I'm sure you know what I mean. How bad was the damage to start with? How critical is this equipment? Are their redundant or back-up systems? How does the data compare to normal data (if you have any)?

Good Luck,

Danny

Thank´s Danny, it does seem to be pretty bad but it´s slow speed so it does not fall apart in an instant it will take some time. This beast is the so far only case where expert softw. once did beat me on an evaluation of a fault. Then it was a locked up coupling on this output shaft. Now they swear it was taken apart and regreased in October but who knows. Levels are much lower on extruder side so it´s not likely this time. So I have a special eye for this one. Seems also to me that something is grinding giving a rubbing most likely I guess metal-metal or one gear set is slipping apart so the mesh is real bad. To me it looks like it excite every possible resonance that exist in the complete frame and assembly. It is the input machine for the plant so it´s pretty critical, no spare to be found within months. On old data at normal status it´s well behaved, gear set is 20-25 years old and did look acceptable for that age before. I had a problem when it was new after repair in the input shaft where I asked them to look at the input coupling as it was reported to be worn from the assembly crew but they only did an inspection and declared it ok. So I still have indications also on the input shaft but not this bad. They will now swap product and reduce speed so we will see what difference it will make. Bearings are gathered to have them around when it´s possible to stop. Olov

Olov,

If you blow up the TWF, will meshing be visible in it? Switching to velocity units may help in this respect. Increasing Fmax may also help in detecting a faulty tooth.

Was assembly phase marked and maintained after the gear box was reassembled?

Just an opinion...l

I was reading from experinced people that when gs peak is 7 g (positive or negative) in the TWF for ball bearing or 12 gs (positive or negative) for roller bearing, the bearing has flaws.

In your before repair TWF you had g peaks more than 6g and disassembly shown you bad bearing, isnt it?. I guess it is a ball bearing, isnt it?.

Now your after repair TWF show g peaks higher than 4g, it make to think that new bearing is 66.667% of useful life compared to old bearing if the relation would be linear. Get your own conclusions and share with us.

Hi Oli,
What about extruder alignment?
Had a case where we lost GB, input shaft bearings failed, repaired GB and similar to your case GB looked worst after. Problem solved by aligning extruder and cylinder, not easy job. Extruder's screw was attached to GB output hollow shaft, and waveform showed sign of rub, but we were sure that was from GB, when we pulled out screw, clear sign of rub inside of cylinder. After alignment GB looked normal.
I used laser, but heard stories about using piano wire to do alignment.
Good luck.
Dragan

Hi Miguel,

Ur calculations are interesting!!! How did u get the % of useful life?

Dragan, it was supposedly laser aligned after repair and the axial bearing is on the extruder side. Last time 5-6 years ago it was outgoing coupling and then it was a piano wire man that did the alignment, worked fine. Rubbing do have less signal level on the extruder side, but you never know. Gbx will be opened after directly after easter so we will soon know. Olov

Mystery solved one of the gears on the middle shaft was wobbling on the shaft having heavy pitting and wear in the attachment area and that metal-metal clanking excited every resonance all over the gbx including the old chiller that now was found as debris so someone didn´t look around that hard at the previous refurb a year ago. I guess it was good the time to open it was kept short. Olov

Gentlemen,
This may need to be another thread, but now that the problem is solved, DragonT reminded me of a question...
Do you use the special lasers just for extruders? Or is there a way to make the regular alignment work?(We have a Fixturlaser.) I same up with an idea a couple of months ago for a low-cost alignment laser using a magnetically centering gunsighting laser fitted into an arbor that goes in one end and a target etched in plexiglass set in an arbor at the other end. Having a prototype built right now. How well do you think this will work?

Richard,

I have worked with a lot of gun sight lasers. You will need a jig to center the laser beam to the laser body or the arbor that it is mounted in. Many of the lasers have adjusting screws for beam centering. Regarding the application: the method should work OK as long as visual targeting gives adequate accuracy. There are lasers available that offer greater accuracy using an electronic target.

Walt

Wheel that was about to drop loose is now fixed on shaft with radial expansion gadget w. like 12 bolts, shaft is reworked to make a surface for this and wheel have a bigger hole to make space for it, levels are down but there are still modulation by mid shaft rotating speed and those plates that have easy to excite freq´s still give noise but not as bad. Repair guys swear that gear is now concentric on shaft. Any other suggestions than bent shaft? Or will it settle after some time as it sure now have a brand new wear track to get used to? Any other suggestions? Olov

second_repair.doc (56 Kb, 40 downloads) 2nd repair

OLI,
Other possibility is that middle shaft center bores are not concentric. They might fix gear and shaft, but ask for GB housing bores. To check properly GB housing, they need horizontal boring machine. Ask if bores are remachined, checked/measured.
At this level you might consider do nothing, continue to monitor.
Good luck.
Dragan.

It´s now back to haunt me. After 4 weeks vacation the guy with ears came back to the plant and the gbx sound like a freight train. Most trends in and from the FFT has dropped 10 to 30% since last measurement after installing the quick fix gear mount so it did a good wear in part from the train that it now got inside that sounds pretty bad. But not in vertical and mostly not on bearings towards the motor side they have not reduced or increased slightly. In TWF you can see very nice impacting with spacing matching the intermediate shaft speed. So we opened for a quick look today and not much to be seen. But the quick fix mount have moved axially on the shaft as there are shavings and thus the gear is now not so good centered axially as it was from start. The mounting devices, there are 2 of them side by side should take at least 3 times the load that´s supposed to be. It does not seem to be slipping as it still rotates. So how can it move axially without slipping? What we could see does not explain the impacting, is my guess that the wheel got a crack that we could not see at a quick look possible? I can post some nice impacting tomorrow if someone would like to see. On wednesday we will run full load that is the condition I normally use so it will be easier to compare to previous data. If it will live so long or survive full load. Olov

The really good ones just keep coming back.

Good Luck,

Danny

quote:

Extruder's screw was attached to GB output hollow shaft

Dragon T
We have a problem with our extruder gearbox and I learned that the screw shaft that is attached to the GB output hollow shaft will not break loose . Did you guys have any problems with this?
Mark

Mark,
Some years ago this one had a lock up of the coupling due to problems to distribute grease in the spline type coupling. It gave hi bearing values around that but no distinct frequencies. Expert software did however point it out correctly... I have heard about exactly the same from a milk carton plant where they put the plastic on the cardboard. In these designs the axial bearing is outside the gbx just in front of the extruder so there is no axial force from that in the gbx the coupling is between. I also have seen a axial extruder bearing go down due to no lube and running for 6 months, just keeping the flow lines clear from brass coming out as it cloggs the exit.... (when the lube was turned on again) and no significant change in any vibe parameter you can think of, but that is another machine. Olov

Olov,

Still a problem? I'd be glad to have a look if you haven't resolved the problem. Send me the data in imperial please and I'll reveiw and can post my opinion here if anyone is interested.

Hi Mark,
We haven't had a problem with screw shaft, but GB input shaft would bend. It is huge torque and input shaft is the weakest spot.
Have a great weekend.
Dragan

Still going like a train....
Will it run until January?
It was stopped and investigated again once after emergency repair and the middle shaft had moved axially as the spacer was not fitting. New axial spacer was made and fitted. Emergency gear fit using 2 expansion devices has not moved. Since then it has been grinding in to a new match and the train sound is still on. Operator say thumping came instantly one day, weeks after repair. Spectra sum and twf peak increase like 10% every 14 days. Iron and boron in oil increase, first a jump of 20ppm then also approx 10% increase every 14 days at least for iron. Since I have seen gbx´s with about twice the current levels still operating and a 10% increase per 14 days give a number of weeks to reach that (presume linear and it is not that) then very bad state... One more stop will or have been done to take measurements for the new gear and shaft set that will be fitted at as a late date as possible. Will it survive until spareparts are made? What will actually break? Gears are 15+ years and there are a lot of material to grind on. What can be the problem, middle shaft got bent as it has less diameter, to be able to fit emergency repair gadget? Gear fitting got excentric? Some gear damage can be seen on photo but nobody looked for it and I was not on site. Olov

gbx.doc (159 Kb, 29 downloads) Gbx