posted 06-03-2000 10:51 AM               

I monitor two underwater pelletizers that have been going through gear couplings nearly every 6 months. The driver is an 8 pole, 100 HP motor powered through a VFD (I suspect PWM). The pelletizer consists of a 20 to 24 knive hub that cuts polymer into pellets as the material is extruded through a die and quenched by a continuous water flow. The motor and pelletizer are coupled with a curved tooth gear coupling (with considerable axial play). The following is a picture of the UG-1000 pelletizer. The two units I am discussing are similar but smaller (UG-500s).

Recently, the VFDs were reconfigured such that the top end of the motor RPM was increased from 900 to 1100 RPM nominal. Since the change, the units have been operating at speeds higher than in previous years. Also, the coupling gear teeth have been "wearing out" such that coupling replacement is required on a 6 month or less interval. The damage starts within a week of placing the coupling in service and appears as tooth pitting due to adhesive wear. Then as time progresses, the hub teeth will wear down to a sharp edge and the spool teeth will have a considerable indented imprint of the hub teeth. The grease appears caked and burnt.

The team working on this problem has identified problems with the lubrication. The grease supply was old (nearly 9 years old), too little grease was being used, and the grease was not formulated for gear couplings (centrifuging of base oil from the thickening soap).

On Monday, one of the couplings will be inspected to determine if it has suffered any damage. The coupling was placed in service approximately 3 weeks ago, greased with FALK High Speed coupling grease, filled with the proper quantity of grease, and not operated above 900 RPM.

Now that I have given a "brief" description of the problem, I am fishing for some advice.

1) Does anyone think shaft current could be an issue?

2) Does anyone have a good method of checking for shaft current?

3) Does anyone have any suggestions of things to look for or possible remedies? (We have a good grasp on alignment issues and have deemed misalignment as a non-variable.)

posted 06-03-2000 02:19 PM               

Since both shafts rotate together you could make a nice jumper and attach it from shaft to shaft, or shaft to spool and spool to shaft.

Wouldn't expect a coupling to last long with bad lube.

Got a non-contact Temp probe? Might be nice to check spool temp from time to time. Also Monitor and record motor current or some other power related thing. Friction, Heat and wear are related, but not the same thing. It is possible to have a lot of friction, most of which goes directly to heat, and relatively little wear.

Here is a second-hand hearsay misalignment theory I heard once that made, and still Makes a lot of sense to me. It was suggested (reportedly by an un-named Falk engineer, who undoubtedly disappeared mysteriously and suddenly after uttering this radical theory), that good gear coupling life requires lots of misalignment. The reason was because, in a gear coupling, to establish a hydrodynamic film, it would be necessary to have some sliding velocity. Angular misalignment (the only misalignment possible when 2 meshes and a spool are used) accomplishes this. If alignment were perfect,and there were no axial thermal expansion, each tiny piece of tooth contact area would remain in contact with the exact same opposing piece of tooth contact. In a 24 hour a day constant load application, there would be no load reversal,and once the original assembly lube squeezed out, there would be little or no chance for any more to ever get in again. I would expect fretting and wear.

In my mind this is a very logical extension of the well accepted notion that Good lubrication requires separation of surfaces by an oil film. Yes, yes, if we can walk the line of boundary lubrication the friction is the lowest of all, but the penalty for being being on the hydrodynamic side is extra heat from lubricant shearing. The penalty for stepping on the OTHER side of boundary lubrication in metal to metal contact.

Lots of machinery uses either plain journal bearing or ball or roller bearings. All types rely on a sliding velocity to generate an oil film to withstand the pressure from loading, and separate the metallic surfaces. Even the contact ellipse in a ball bearing has large zones of sliding. These zones need a minimum velocity and viscosity to have Kappa larger than 2, which is the point hydrodynamic lubrication "lift off" is achieved. The exception is a hydrostatic bearing, that uses pressure applied over a large enough surface to lift the shaft, and balances peripheral clearance leakage against oil flowing in through orifices to maintain a fairly constant peripheral gap and pressure.

posted 06-03-2000 11:03 PM               

But.....we have seen a number of cases (and they are increasing in frequency) where VFDs particularly of the PWM/vector variety are the root culprit.....and so far their have been no easy solutions.

posted 06-04-2000 02:52 PM               

The pump moving the quench water stream across the die face is putting up a flow of approximately 1000 gpm. The polymer is extruded through the die at rates of 35,000 to 50,000 pph. Frame grounds for all of the equipment involved are very questionable. Also, the overwhelming majority of the damage seems to be happening on the end of the coupling nearest to the pelletizer. The side of the coupling nearest the motor usually shows very minimal to no damage.

I suspect the bearings within the machine are well insulated but can not say for sure.

posted 06-04-2000 05:03 PM               

Do you think I should take a reading from shaft to ground on both sides of the coupling?

posted 06-04-2000 06:49 PM               

I walked by a moving roll (unwind) of poly film and got quite a zap (fried the LCD display of the contact tachometer that happened to be hanging on that side of my belt).

Their rather simplistic but effective solution (usually) is to drape a piece of copper material (looks like a scouring pad, comes in a roll, from ....????) over the shaft and clip the ends together with a ground clamp such as that used to ground barrels of flammable material.

posted 06-05-2000 03:18 PM               

Testing is then performed by taking voltage traces across different shunt resistances in the decade box. Different traces can then be interpreted to indicate the general nature of the voltage source.

Basically you have a portable voltage current monitor that most big turbines have permanently installed. MPS out of New Jersey is planning on marketing this device in the future as a portable voltage current monitor.

So, yes, you can monitor shaft voltages on the fly without taking the machine down to install a brush.

I placed the portable brush myself on the shaft and we did a pre job safety review to make sure all precautions were taken. The brush just has to be barely touching the shaft to close the circuit.

posted 06-05-2000 10:06 PM               

The grease specified by the OEM was not suitable for gear couplings. The thickener would centrifuge from the base oil and leave little to know lubrication. Then the coupling would lock-up. Prior to my involvement with the problem, "others" "determined" the coupling lock-up was because the mechanics had used an excessive amount of grease. The mechanics were instructed to lightly coat the hub teeth (approximately 1/10 the amount required by the coupling OEM).

The current team investigating the problem determined the grease was not well suited for gear couplings and the supply in the lube cabinet was well past the usable shelf life. We missed the part about putting the proper amount of grease in the coupling. It only took us about 3 couplings (at ~$3000 each) to figure out the grease was not leaving the coupling because it was never in the coupling to begin with.

Live and learn!

Sorry Ron. I don't know how but we started talking about VFD's / Bearings and ended up on gear couplings.

[This message has been edited by Greg Klein (edited 06-05-2000).]

posted 06-08-2000 04:17 AM            

One more very important issue which has a strong effect on gear coupling life: the gear tooth surface hardness. For most applications, the coupling gear teeth should be hardened by the manufacturer to a Rockwell-C hardness of at least 45.

Many gear coupling manufacturers sell the non-hardened gear couplings(say HRC 35 and below) for a lower price,due to the lower manufacturing cost.These couplings are suitable for low load applications only.

posted 06-09-2000 03:23 PM               

[This message has been edited by Don Rainey (edited 06-09-2000).]