Large machine resonance

[testtech]

A demonstrated resonant condition exists in a gearbox. Vibration of 1"/sec is measured on the structure at 330 Hz. The gearbox is driven by a 1000 HP DC motor and runs a rolling mill. The gearbox is about 6' by 6' by 4'. It sits on a concrete foundation that reportedly extends 8 feet underground. An operation deflection shape animation indicates that the foundation is rocking back and forth at the resonant frequency. The resonant frequency is gearmesh frequency. Here is the question: is it reasonable to have a resonance on such as masssive structure at 330 Hz. or should I be seeking another source?

 

[electricpete]

I don't think the resonant frequency can possibly be that high.

I would question the ods analysis.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Perhaps I shouldn't be so quick to judge. Clearly the first resonant frequency would be much lower. I guess there can be very complex mode shapes with much higher frequencies although I have never encountered that type of thing. It would be interesting to hear a description of the mode shape that is theorized.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[EnglishMuffin]

1. Is it possible to have "a resonance" at 330 Hz on or within such a massive structure ? Yes.
2. Is it possible that the foundation rocks back and forth at 1 inch/sec at the same frequency at which something else is resonating ? Yes
3. Is it reasonable to suppose that an 8 x 6 x 4 ft thick concrete foundation supported by soil has a resonant rocking mode at 330 Hz ? No.

But more information is needed before one can come to any definitive conclusions. You presumably did, or could do, a modal analysis of the structure and foundation. What are all the significant indicated resonant frequencies, and where exactly were the accelerometers located ?

 

[testtech]

Muffin:

The resonance is well demonstrated by waterfall plots. At 333 Hz. the vibration is approximately 1"/sec. It drops to .1 when the machine turns at full speed.

I have done an ODS analysis on the gearbox and steel pad. The animation is based on 91 DOF's. Since the base is poured underground, I have no access to the the actual base.

I added a base to the animation to demonstrate interpolated movement.

The reason I suspect the base is because the ODS movement shows machine components, including all exterior shaft bearings (one set of bearings is completely inside the case) move in phase with the other casing movements.

The resonant frequency seems somehow tied to loading. The mill operates in two directions. The resonances is present in both but severe only in one. This would seem to eliminate modal analysis as a reliable tool.

Of course, I posted this thread because I have a hard time believing base instability at this frequency......

 

[testtech]

Pete:

Such a high resonant frequency on this massive structure is new to me. In my experience, they have typically occured at shaft speed, not a gearmesh frequency. Of course, I am not proposing a mode shape since I have conducted ODS, not modal testing.

See my post above, for concerns with modal testing on this machine. However, it is possible to undertake a modal test on weekends if it appears necessary.

 

[EnglishMuffin]

Have you no access to any part of the base ? Even where the gearbox mounts to it? Given that you could access even just the top surface of the foundation, I'm not sure why you would say that modal analysis would not be a reliable tool in this case. If there truly is a foundation rocking mode at 330 Hz, modal analysis should be able to detect it. You probably would need a decent sized instrumented hammer of course, with the impacts appled somewhere on the gearbox perhaps. The fact that the response differs depending on rotation direction merely suggests to me that the excitation force depends somehow on the rotation direction, which is not an unknown phenomemon in the case of gearboxes.

 

[testtech]

I can investigate this further. However, if the base is rocking, is there a solution without pouring a new base?

The resonance occurs below some operating mill speeds and above other operating mill speeds. It appears they can operate outside the resonant range. If so, the question may be whether constantly passing through the resonance will cause premature mechanical failure? If the answer to this question is no, that suggests an easy solution. If the answer is yes, then a solution to the resonance must be found.

During run up and cost down, dwell in the resonant range is perhaps 20 seconds.

Time to go to work.

 

[magicme]

hello
is this a new installation? or was the machine running smoothly and then developed this problem that you were asked to resolve? of course my point is that if it ever ran smoothly the problem cant be the base unless it has cracked or crumbled below the surface and its stiffness has changed.

like everyone here, i cant imagine a large machine having a fundamental rocking mode at this ampltude and frequency....how much does the gearbox and foundation weigh?(assuming a solid foundation)

also....the fact that the problem exists mostly in one direction of rotation might indicate a chipped geartooth on one side......so it might not be a resonance, but a forced response your looking at.

daveleo

 

[GregLocock]

5g at 300 Hz is probably tolerable if you don't dwell there - 5 g is a typical vibration level on an IC engine, in use, and 300 Hz is a typical crankshaft resonance.

Now, if this huge concrete block is jumping around at 5g then I've got to say that Mr Newton is going to get involved, and I have a hard time believing that the worst gear set in the world is going to create forces of the order of hmm, 18 tons *50/Q, say 90 tons, if Q is 10



Cheers

Greg Locock

 

[testtech]

Daveleo:

This is an existing mill that was moved from an old plant and newly installed in its present location. Some of the gears in the gearbox have been replaced.

There is no information on the performance of the gearbox in its prior location. The gearbox is about 25 years old.

As stated above, the vibration at gearmesh rises to 1"/sec at part speed and then drops to .1"/sec at full speed. I do not see how this cannot be a resonance.

 

[testtech]

On Wednesday I will obtain vertical ODS readings on the four bolts that secure the base plate to the block. If movement is present here, I guess I will conclude that the block is moving. If no movement occurs, I will obtain additional readings on the base plate. At this point, it appears that the most likely source of the resonance is the base plate. It may need to be secured to the block in more than four points.

If any of you have additional suggests, I look forward to receiving them.

I appreciate all of your insights.

I will post the results of the follow up tests.

 

[magicme]

hello again

you say the baseplate is bolted to the foundation at four points......how thick is the baseplate? and is it a plate supported below on a continuous concrete foundation or is it supported only around its perimeter....if the plate flexibility (rather than the concrete foundation) is controlling the resonant frequency you can bolt it down everywhere you can (especially inboard of the perimeter) and drive the mode up (hopefully) out of the operating range.

too bad we cant post sketches here.

daveleo

 

[testtech]

Daveleo:

I will find out the answers to these questions in the next few days.

 

[testtech]

Update:

I obtained additional ODS data on the bolts that secure the baseplate to the concrete block. Minimal movement occurs at the bolts, demonstrating that the concrete block is not rocking, confirming the suspicions of several of you. The ODS animation clearly shows that the machine baseplate is deflecting. There is likely a deficiency in the grouting installed under the base plate that is permitting the observed resonance. Inspection of the exposed grouting shows cracks in the area of highest deflection.

I will post the final resolution of this problem when it is completed.

Thanks for all of our thoughts.

 

[steve383]

I have a primary gearbox with about the same dimensions at a steel mill drving a 10 stand rod block. It also has a resonant frequency from the input gearmesh at around 380HZ of 1 IPS and drops to .1 IPS at other speeds. This unit only runs at that speed a few time a year so it is tolerable. I would be very interested in your solution if you find one. Please post it. I believe it is a gearbox design issue and repairing the baseplate will slightly reduce your levels and change the resonant frequency a few hz. Please let us know what you find. Good Luck!

 

[testtech]

steve383:

I have based my conclusions on observing the ODS animation of the gearbox movement. However, these things are rarely conclusive. A fix for the baseplate is being investigated. However, just to ensure that all issues are covered, can you state why you feel you have a design issue, rather than an installation issue. What component do you believe produces the resonant behavior on your machine?

 

[steve383]

My best guess is that the issue resonance is from the input shaft. I have not tried to calculate the shaft natual frequency or harmonics. I base my decission on the fact that the highest amplitudes come from the input shaft sensors. the amplitudes are lower at the other positions on the gearbox. I do not believe it is a gearbox resonace do to its mass and lowering amplitudes further away from the input shaft. The resonant frequency is also gearmesh of the input shaft. If you believe it is a gearbox/base resonace it would be interesting to mount a shaker unit tuned to 330 hz and see what amplitudes are generated and where.

 

[unclesyd]

testtech,
We had a similar condition in a large gearbox of 1200 HP AC Motor input. Output less than 150 RPM. I don't have the particulars but the experiences were the same as yours except that the amplitude varied more with the load. They analytical group fought this thing for several weeks trying to tie down the source. We had two of the boxes, the only two in the world, so there was nothing to compare except the other box. The other box was uncooperative. The results were that an intermediate pinion shaft and bearings were free to float as the thrust bearings had no restraint. The only thing holding the bearings were the cover plates. There was some wear on the gears but once locked into the proper position there was only a little spike at gear mesh. I never heard a full explanation of the problem and resulting vibration.
The other machine was built the same way but was corrected before any trouble.

 

[testtech]

steve383:

The relationship you find is just what I was seeking in the ODS. The ODS showed the the shafts were moving generally in unison with the general structure. This suggested that the resonance was not in the shaft or bearings. Your shaker experiment would be interesting. I predict the resonance would not appear. Remember, the mill turns in two directions but the resonance occurs in only one direction. In that direction, I believe the torque reduces the loading on the foundation, permitting the resonance to occur. I suspect that we would not see this resonant frequency if the machine were not in operation. Looking toward the mill, the large machine movement occurs on the right side of the base when the steel is pulled to the left. Thus, I believe the left side of the gearbox is loaded and the right side is unloaded. During reverse operation, the left side is loaded and resonance does not occur.