We have a rotary kiln driven by two 500 HP DC series motors. One motor has a tach generator for control loop. Our optimum running speed is 80% or 900 rpm. We recently started experiencing speed changes of 105 cpm, this can be heard in the motor control center and at the motors and gearboxes. It can also be seen with a strobe. It shows in the vibration spectrum as side bands of running speed and around some GMFs. The 105 frequency is constant at 900 rpm but has varying degrees of amplitude and is evidenced by enunciation in the gearboxes and arcing at the brushes.
Changing speeds to 70% or 790 rmp significantly reduces the amplitude and give us a new frequency of 77 cpm which is not proportional to 900 rpm / 105 cpm ratio. I think that rules out a mechanical problem. There is no corresponding mechanical frequency in the drive train that matches either 105 or 77 cpm.
We have a spare drive in the motor control center, we switched to it and amplitude and freq remain the same, at both 70 and 80% so we've ruled out drive, scr, card, fuses, etc
We can rule out system resonance as this system is 30 years old and has been run at 80% for years.
Oil and vibe analysis doesn't show any problems in the mechanical drive train.
We have been able to attenuate the amplitude, somewhat, by dampening control response.
We know that the tach gen has bad bearings but the voltage output, I'm told is constant.
????????

Could the frequency change may be a torsional vibration. What is the frequency of the changes? Control loop/torsional instability could be a possibility worth testing.

We are running falk steel flex couplings both hs and ls, I had the falk guy here, i dont believe we're running near any natural frequency of the couplings.
I forgot to mention that at 80%, the 21.6k frequency is excited.

Check armature for grounds and shorts. Have run into this several times before.

Howard

Gonzo,

Some time ago I worked on an aluminum rolling mill with speed instability in the DC drive motors. I could not find any unusual vibrations. I did find high and erratic ultrasound levels on a shaft speed encoder that provided speed feedback to the control system. The defective encoder was replaced, and the speed instability disappeared. I am not sure if this story fits your problem.

Walt

Gonzo
I has same years ago the same problem. The error was at motor control, same circuit board is defective. In 80% there is 360 Hz (21.6 kcpm)
verific if there are submultiples of 360 Hz, the line frequency family. If presents, the problem is at circuit board - not in mechanical parts.

Ricardo Góz - from Brazil

Look for post on this board in this section we are in called SCR Help. This may help.
Do you have any SCR firing frequency present in vibe data (360 or 180 hz)?
Has any IR scan found any loose connections in any of the circuits?
Short field windings can cause this type of speed change. A voltage drop test accross the field poles/coils will confirm if shorting is present.

As mentioned, we have a spare static drive, upon switching drives the problem persists. I think we're looking at either a motor or speed control or loop.

OZNOG <-?-> GONZO,

How about replacing tachometer and its shaft coupling? It has worked for me before.

Walt

I would follow up about the SCR firing frequency. Chances are you have 6 SCR on your control board and the power supplied to the motor is not being smoothed enough. Basicly you would have an AC voltage being imposed on your DC motor. Long term, this creates havoc on gearboxes and chain drives.

Take Care,

Is there a way to verify if these modulations are load related? Does current signature shows the same? Do I read it correctly that neither 105 RPM nor 77 RPM is not gear box output speed?

gonzo / oznog same person, sometimes cannot sign on from work.
Our kiln is down now for 18 days and due to gear replacement am limited to any dynamic electrical testing and crane availability for motor replacement.
We plan on replacing the tach generator and bringing in an outside contractor to perform a static analysis of motor and leads as recommended by "motor doc".
In response to Chris Olson, what do you recommend for follow up on SCR frequency, what specifically am I looking at or for? How is this AC voltage imposed on the motor and what is correction? I suspect in time it will reek havoc with the mechanical drives, but what is probability of electrical failure?
In response to Dave G, I know load issues exist, as the kiln rotates 1 rpm, the material rolls and "sluffs", but this condition existed from the beginning of time (29 year old plant)and I cannot correlate exactly with sluffing; however I've never seen the corresponding arcing at the brushes or severe changes (around 10-15 rpm changes / 105 times per minute)

Gonzo,

If I picture this power trane arrangement correctly, the kiln is crowned with teeth and driven by a pinion gear. If the kiln, which is probably resting on rolls, is rotating at 1 RPM, then it looks feasable for the driving pinion and rolls to spin at 77 or 105 RPM, isn't it? If the above is correct look for a pinion/rolls problem.

This message has been edited. Last edited by: David_G,

The 19 tooth pinion driving the kiln ring gear turns at 17.6 rpm with the motor at 900 rpm / 336 pinion / ring gmf. Falk 2195N3 gearbox, triple reduction internal gmf's: 40500, 9878, 1918. Input rpm: 900, Int shaft rpms: 1st: 493.9, 95.9, reducer output 17.6. Nothing close to the 105.

Monthly vibration and oil analysis, and semi annual gearbox filter ferrography don't indicate a gear box problem, nor does't temp profile or inspection of the pinion and ring gear.

This 105 cycling is also evidenced at the amp meters on the static drives by amp swings up to 50 amps.

We've had this problem for a couple of months, I assume if it were caused by some mechanical drag, it would've worn out or created a hot spot by now.

Gonzo,

Your geartooth combo's must be as follows to match the gmf's you state:

1st reduction 45/82
2nd reduction 20/103
3rd reduction 20/109

I don't have any info on a 2195n3 (should the n be a y?) but the high speed pinion of 45 teeth sounds out of the ordinary for Falk. Are you sure of the tooth counts?

It is a "Y" unit, 2195YN3 is info I have, and you are correct in gear counts.
This box was built in mid '70s, and these are the counts programmed into our database.

Gonzo,

I guess the gmf's that you stated have been verified by vibration data?

That just seems like a pretty big pinion but I guess the 70's were a pretty weird time for many of us.

Gonzo,

What about the rotational speed of kiln's support rollers and their bearings? I have seen problems with these on lift pumps at a waste water treatment plant. I'm just covering all the mechanical frequencies.

If you exhaust all possible mechanical and electrical sources for this apparent non-linear or load-related modulating frequency, then consider the material inside the kiln. No point in laughing at this possibility right now.

Walt

Kiln rides on 4 support rollers, 8 each 30" dia journals riding on brass bushings, both hydrodynamic and hydrostatic high pressure lubrication. Oil analysis and filtration monitoring show no problems. 3 thermocouples in each bearing are monitored 24 / 7 and my experience has been any mechanical issues with these bearings result in immediate increase in bearing temp.

Rollers are aligned so kiln axially just marginally rides on thrust rollers. Oil analysis shows both to be ok.

We know the material 'sluff' contributes, just not to this magnitude.

Gonzo,

You say kiln is down for gear replacement. What was the fault in gearbox, and were you able to detect it? Was gear fault directly or indirectly related to the 105 cpm vibrations or motor speed variations?

What is support roll speed compared to kilm speed, or what is kiln diameter at support roll location to ratio with 30" roll diameter?

I consider Falk steel-flex couplings to be a torsional nonlinear spring, so torsional natural frequencies can possibly change with load. I have measured torsional vibrations on variable load blowers with these couplings.

Walt