Engineering Concern for Long Spacer Coupling Or Drive Shaft Driven Machines

E-mail:

Ph: PCS 386-916-6700

Engineering Concern for Long Spacer Coupling or Drive Shaft Driven Machines

Drive shafts, although good for power transmission are not understood very well regardless of the application. I wish to address some basic fundamental issues regarding their use and set up. I did a five-month research and development project in MI for the automotive industry specifically for drive shafts. It isn’t so much that they are a problem; it’s the enigma of balance and factors affecting operation and setup. Drive shafts are engineered and designed basically to provide 100,000 miles of good service. So far they haven’t been able to make them better than that and won’t until some fundamental basic thinking is changed. Usually drive shaft thinking is based on automotive application, not industrial. So, how long is 100,000? At 1200 RPM, running continuously, one could expect to reach that many revolutions around one hundred days basically. When you’re putting a drive shaft in a racecar it undergoes some extra detail. Two-piece drive shafts are more trouble - Balance is a key factor; but, you can balance an out of round tire — it will never run smooth.

Now to the heart of the issue (our application) — the use of drive shafts in industry. These drive shafts transmit power from usually a motor (fixed) to a gearbox (fixed) or some driven machine (pump) that is mounted rigidly or fixed. The automobile application is flexible and requires misalignment so that the torque applied and motion in the flexible arrangement will never allow the angle to pass through zero. By adding slack and taking up slack under load, it will snap the U-joint. But, in an industrial application this doesn’t apply. The angle is fixed and one doesn’t have to worry about passing through zero.

In our industrial application we usually utilize pillow block bearings in conjunction with U-joints. Some have gone to Thomas couplings as they are dry (don’t require lubrication) and work very well carrying large loads and not having the centrality problem of the U-joint, and, usually work better in an industrial application. However, we’re addressing the U-joint design for the present.

When you rebuild a drive shaft; just replacing the pillow block bearings isn’t good enough. When installing a new unit; just assuming that all is OK isn’t good enough. One must check run outs and centrality. What is the condition of the bearings in the U-joint itself if rebuilding an old unit? The yoke that those bearing cups are housed in - do they fit right? What is the axial clearance?

Next, centrality must be maintained by shimming the bearing caps to reduce run-out to less than 0.002” while keeping the axial float to 0.002” or less. Bow must be less than 0.015” per 8’ and balance less than 0.08 inches per second. (I’m using a universal language here for the for the field technician or vibration analyst. Gram centimeters are available.) Drive shafts appear simple, but all of a sudden they aren’t.

The above-mentioned terms are straightforward and easily understood, so, what’s the big issue? ALIGNMENT! Our industrial application doesn’t conform to automotive standards of the flexible arrangement, which requires gross misalignment to maintain a positive angle. In the rigid application precision alignment is required; not just desirable. One hundred days of service isn’t acceptable. Thomas couplings require a greater precision or accuracy of alignment than do U-joints. U-joints require lubrication, so some movement of the rollers within the bearing cups is required. This movement doesn’t allow the needles to break through their film of lubrication. However, this angle ‘must’ be held at an absolute minimum. What is that absolute minimum? This is the biggest astigmatism surrounding the enigma of misalignment in drive shaft applications. I’ve been working on them since I was fourteen years old but only have scrutinized them to this large degree in the past eight years.

First I’d like to say, go for the Thomas coupling design or that type design whenever you can and shoot for precision alignment. One doesn’t have to contend with the centrality problem and lubrication. What is precision alignment? Thirty seconds! Thirty seconds! Thirty seconds! Yes, thirty seconds or less! So, 30 seconds over 12 feet would be 0.042” (under 1/16”). Are there any raised eyebrows?

Now for the U-joint application: five minutes or less, but close. That’s 1/12 of one degree, not 3 degrees per joint (I’ve seen some specifications at 4 ½o). When you want to know, ‘how much’ misalignment can exist in thousandths of inches — multiply the tangent of the angle times the length of interest in inches. Proof is in experience and field proven situations. We have a lot field proof against the gross misalignment approach (3 degrees per joint or anything close to that). Again, I’d like to reiterate — rebuilding a drive shaft — check yoke gap, axial float (maintain less than 0.002”), bow (less than 0.015”), run out within 6” of the joint (less than 0.005” TIR), clearance within the bearing (manufactures specification), balance (0.08 inches per second), pillow block bearing to shaft alignment (square on the shaft in both planes), pillow block bearing mounting, inspect (must be rigid so as not to set up a vibration pattern). I haven’t distinguished between drive shafts and torque tubes. But, torque tubes could and can have water in them some times. I’ve seen it happen. So, don’t discount it. Are these items part of your checklist? They should be.

I’ve addressed this topic this month because I’ve felt a need for this issue to be brought to the surface. This should clear the water and make for better relations between engineering and maintenance personnel as understanding and communications usually do.

I’ve been teaching machinery alignment and set up of machinery for over twenty years. Setting up machinery for thirty-six years and have specialized in alignment, balancing and vibration analysis for over twenty-five years. Basically, this paper was generated by a request from a client that is facing drive shafts as a new application in a new plant. He has much concern as they are installing these units at ~3 degrees misalignment per joint. Others that I’ve seen that have approximately this much misalignment in the U-joint design do have problems. (I went into one plant and correct their situation – they were experiencing failure between 3 & 6 months). But, they are experiencing these problems well after the fact of new installation. I don’t know how they were at day one. But, unless extreme attention is paid to minute detail, smooth operation shouldn’t be expected, just hoped for.

I hope this paper has been of benefit to you and helps to avoid future problems in your facility.

TOLERANCE CHART

Speed/Category
/
Rigid
/
Semi-Flexible
/
Flexible
Tolerance
/
Mils | HMS
/
Mils | HMS
/
Mils | HMS
0- 300 RPM / 2 | 0o 0’ 25” / 6 | 0o 1’ 30” / 8 | 0o 2’ 0”
301-715 RPM / 2 | 0o 0’ 25” / 5 | 0o 1’ 15” / 7 | 0o 1’ 30”
716-901 RPM / 2 | 0o 0’ 25” / 4 | 0o 1’ 0” / 6 | 0o 1’ 0”
902-1201 RPM / 2 | 0o 0’ 15” / 4 | 0o 0’ 45” / 5 | 0o 0’ 50”
1202-1801 RPM / 2 | 0o 0’ 15” / 3 | 0o 0’ 30” / 4 | 0o 0’ 40”
1802-2401 RPM / 1 | 0o 0’ 10” / 2 | 0o 0’ 20” / 3 | 0o 0’ 30”
2402-3601 RPM / 1 | 0o 0’ 5” / 1 | 0o 0’ 10” / 2 | 0o 0’ 25”
3602-7000 RPM / 0 | 0o 0’ 3” / 0.5 | 0o 0’ 5” / 1 | 0o 0’ 15”
7001- nnnnn RPM / 0 | 0o 0’ 1” / 1 | 0o 0’ 3” / 0.5 | 0o 0’ 5”

Mils = Thousandths of an inch; 1 mil = 0.001 inch or 1/1000 inch and 0.001”.

HMS = Hours Minutes Seconds [(degrees, minutes, seconds) 1/60th degree = 1 minute].

NOTE: When engineering or OEM’s engineering has specified a tolerance from plotted or calculated data then that should supercede all other tolerance specification. The chart above will usually meet or exceed all OEM specifications. This is especially true when warrantee by the OEM is mandated to meet their specific number/s.

Sam Pickens

386-916-9318

PdM Engineers economics into maintenance