Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
trapper

Sr. Member
Registered:
Posts: 75
Reply with quote  #1 

We have a pin-barrel extruder driven by a triple-reduction gearbox with single helical gears.The output shaft of the gearbox is hollow with the extruder's screw stub shaft inside coupled with 2 large keys, 108 degrees apart. Also on this stub shaft is a spur gear, also keyed to the shaft, which drives a gear-driven feed roll that pulls rubber sheeting into the screw. This feed roll is mounted on an access door and the driver and driven gears for the feed roll mate when the door is closed.

  10in Extruder Internal.jpg  10in Extruder Feed Roll Gear.jpg
  10in Extruder Feed Roll Screw Gear.jpg 

The tooth count on the gearbox output shaft is 51.
The tooth count on the stub shaft mounted gear is 48.
The tooth count mounted on the feed roll is 54.

During the last shutdown over the holiday, the feed portion of the casing and the feed roll/door were replaced to fix some leaking of rubber into the feed roll gear area. Upon startup, I let it run for half a day for both the feed roll gears to seat and temperatures to stabilize throughout the entire extruder.

In addition to taking readings on all the gearbox bearings, I take readings on the access door where the feed roll supports and bearings are mounted. The first set of readings on the gear end of the door is below:
E1H Latest Comparison.png
As can be seen, prior readings had a clear gear mesh and harmonics on the feed roll gear and also had peaks at the gear mesh from the output shaft of the gearbox (51x). On this latest set, I now had gear mesh peaks at 1x, 2x and 3x but also half order harmonics. Many sidebands are also around each which I had never seen before.

A more detailed view:
E1H After.png 

I also notice that now on the gearbox axial reading we're seeing amplitude modulation at once per rev.
G7A Modulation.png

My impression is that the two feed rolls are not meshing properly, either misaligned or not fully engaged with each other (or both). I've written a work order to inspect upon opening and then checking the mesh with Prussian blue or something similar. Any thoughts on what else to check or my reasoning?


Another question I have is regarding possible assembly phase issues with the feed roll spur gears. I don't know why they chose tooth counts that do not include a prime-numbered gear. All the documentation I have around here make it easy to determine the assembly phase factor. In fact, many of the articles use the same example that TAC has had on their diagnostic wall chart for years. In this case we have:

48 teeth = 2 x 2 x 2 x 2 x 3
54 teeth = 2 x 3 x 3 x 3

So, the common factors are 2 and 3. Do we use 2 and 3 in the GAPF formula or do we use 6?

 


OLi

Sr. Member
Registered:
Posts: 1,874
Reply with quote  #2 
I think it is the best you can do, also check if possible that the keys and shafts that connect to the extruders are not stuck it may be giving trouble to the centering of the gears.
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Walt Strong

Sr. Member
Registered:
Posts: 876
Reply with quote  #3 
Perhaps it my eyes that deceive, but the 48-tooth gear looks to have uneven wear with some teeth sharp, some rounded and some flat at gear OD. Gears that mate when door is closed suggest possibility for changes in gear vibration from door closure tightness and alignment and not from gear condition. This might affect repeatable measurements. What happened in 2018 to cause the high vibrations (trend plot), and what was the corrective actions taken in 2019 that produced lower vibrations?

Walt
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,394
Reply with quote  #4 
Are there gaskets on the door? It sounds like replacement of crushed gaskets with new ones would effect gearmesh.
fburgos

Sr. Member
Registered:
Posts: 659
Reply with quote  #5 
so 48/2 is 24 or 0.5xgmf


Any one uses the FRQCAL? 



trapper

Sr. Member
Registered:
Posts: 75
Reply with quote  #6 
Quote:
Originally Posted by Walt Strong
Perhaps it my eyes that deceive, but the 48-tooth gear looks to have uneven wear with some teeth sharp, some rounded and some flat at gear OD. Gears that mate when door is closed suggest possibility for changes in gear vibration from door closure tightness and alignment and not from gear condition. This might affect repeatable measurements. What happened in 2018 to cause the high vibrations (trend plot), and what was the corrective actions taken in 2019 that produced lower vibrations?

Walt


Walt, it's really hard to tell. I caught that picture on my way out for the day. It just happened to be there. 20 minutes before that picture, I was on-site and it hadn't been taken out yet (they were working on it). With the few minutes I had to inspect it, nothing stood out to me. With all the oil and ground rubber/carbon black on it, I wasn't going to mess up my "walking out" clothes trying to clean the gear. I did try looking at the tooth profiles and it didn't look like there was a noticeable variation. That is my focus as I try to find suggestions.

The high readings in the first half of 2018 were because of the leaking rubber which was packed into this drive gear cavity. They cleaned all the rubber out and changed the oil in the oil bath. In the photo below, some of the rubber has already been cut out.

Z-Cal 10in Ext Power Feed Roll Shaft Gear.jpg

The 2019 bump was when I suspect (never got a straight answer from anyone) the rubber had again started to leak into the gear cavity and they replaced the entire door assembly. That was the source of the 54-toothed gear picture above.

Quote:
Originally Posted by Danny Harvey
Are there gaskets on the door? It sounds like replacement of crushed gaskets with new ones would effect gearmesh.


Danny, there is no gasket on the door. It is a metal-to-metal joint. There is a vent on the door that is supposed to let any rubber shavings drop out as they build up but, as the rubber air-cures (and the heat from the screw and feed roll) in this cavity, it turns rock hard and, eventually I am assuming, the pressure builds up. Hence the rubber forcing its way into the drive gear cavity.

OLi, no one knows anything about the condition of the keys. All the work was done by contractors and they are long-gone.

As a follow up, they told us that they "supposedly" inspected the gears on Monday and everything was fine. No mention of using any indicator paint (Prussian blue-type) or even if they cleaned the oil from the gears and checked the wear pattern or surface defects.

I took another data set yesterday and the indications were very close to those I already posted. I compared another gearbox that is identical in another part of the plant (not one of my routes) and it appeared just like my "prior" spectrums and waveforms with very little sidebanding and very small half-order GMF peaks.

As of now, I suspect the drive gear is a big part of the problem. Our lead has mentioned that we should have a meeting with the affected parties to discuss but no word on that yet. If anything, I doubt much more is going to be done before the next shutdown (6 months from now) or even later unless the wear accelerates so much that they can't keep the calender fed at the appropriate rate.

Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,394
Reply with quote  #7 
Trapper,

I had a similar situation with an similar extruder/strainer except in the gearbox.  

Immediately after disassembly and replacement of the input shaft bearings, we found high gmf with sidebands and fractional harmonics in a gear set with a common factor.

I called it and found that they had not marked the gears.  

They asked what would happen, and I told them that it would have high vibration in the same spectral shape and would eventually develop a new wear pattern.  They asked if the shaft would still turn.  I said yes.  They asked who cares? It's been about 5 years and I still don't have an answer for him.

I would track vibes and wear particles to see if it gets worse.  How often do you collect data on this?


Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,394
Reply with quote  #8 
Is there one lead on the screw?
fburgos

Sr. Member
Registered:
Posts: 659
Reply with quote  #9 
Quote:
Originally Posted by Danny Harvey
Trapper,

I had a similar situation with an similar extruder/strainer except in the gearbox.  

Immediately after disassembly and replacement of the input shaft bearings, we found high gmf with sidebands and fractional harmonics in a gear set with a common factor.

I called it and found that they had not marked the gears.  

They asked what would happen, and I told them that it would have high vibration in the same spectral shape and would eventually develop a new wear pattern.  They asked if the shaft would still turn.  I said yes.  They asked who cares? It's been about 5 years and I still don't have an answer for him.

I would track vibes and wear particles to see if it gets worse.  How often do you collect data on this?




trend is our friend, right
trapper

Sr. Member
Registered:
Posts: 75
Reply with quote  #10 

Quote:
Originally Posted by Danny Harvey
<snip> ... They asked what would happen, and I told them that it would have high vibration in the same spectral shape and would eventually develop a new wear pattern.  They asked if the shaft would still turn.  I said yes.  They asked who cares? It's been about 5 years and I still don't have an answer for him.

I would track vibes and wear particles to see if it gets worse.  How often do you collect data on this?


Danny, thanks for the info. Yes, I suspect I'll be doing the same thing. My big concern, and Walt mentioned it above, was either the old gear is damaged or that somehow the new door and casing were manufactured without enough "reach" into the cavity to engage the driver gear fully. 

I collect vibration data monthly and there is no oil analysis on this particular gear.

And to your other question ... yes, there is only a single lead on the screw.

fburgos, yes, the trend is your friend. The outsized pattern change caught my eye and now I'll just have to watch how that progresses from here.

Thank you all for your thoughts about this.

Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,394
Reply with quote  #11 
I hadn't caught the part about the door and casing being new. Since it sets the pitch line, I would think that's definitely a point of interest. Is there a procedure for measuring the door and casing relative to the pitch lines? If the pitch line isn't right, there should be evidence of sliding on the top edges of the teeth.  If the pitch lines are set right, the theory is that there is only rolling on the teeth.

The reason I ask about the single lead screw is the running speed pulsation in the axial twf. A fixed point on the barrel end could be generating an axial thrust with every rev. That could drive up gmf and generate running speed sidebands without initially having any gear damage or wear evident.
John from PA

Sr. Member
Registered:
Posts: 941
Reply with quote  #12 
Quote:
Originally Posted by Danny Harvey
If the pitch lines are set right, the theory is that there is only rolling on the teeth.


In gearing using the involute tooth form rolling only occurs at the pitch line.  In the areas above and below the pitch line, sliding to one degree or another will occur.  If the sliding is excessive or lubrication conditions are not optimum, wear often occurs above and below the pitch line resulting in a ridge along the tooth length.  You should also strive for parallel rotors or you get end loading.
Danny Harvey

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,394
Reply with quote  #13 
John,

Thanks for the clarification.
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.