Hello…

I work as a balancer in a jobbing shop on a Schenck HB5 balance machine. We frequently get requests to balance couplings. Both used, new or combinations (new motor end for example to match an existing pump end)

There are a number of concerns / issues that I have wondered about over the last 15 years and invite any comments from those that may have experience in this field.

Typically a coupling will come in without a shaft(s) so we must manufacture a mandrel that is suitable to do the job. This entails some involved machining and I wonder how the manufacturer do the job? (OEM) Consider a coupling with a tapered fit on both ends and a spool piece in the middle. My design gets pretty expensive.

If…. I am fortunate enough to have the original shaft(s)… and use that as a mandrel to balance the coupling then “what” do I do with the spool piece?

One place that every balancer sees discrepancies in is the weight to add to the void of the key way. The API 610 (and other API segments) state that one should use a fully crowned half key. The equivalent of “as if there was no key seat / way at all”.
Given that the key seat is typically cut deeper than the minimum required to allow for clearance, how do others deal with this discrepancy?

Thanks for any comments / thoughts
Ralph

Most OEM's balance couplings in a vertical balancing machine prior to cutting the keyway. A vertical machine, obviously eliminates the need for a mandrel shaft. Even after the keyway has been cut, most modern balancing instruments will allow "keyway compensation" of a fitment, which is quite accurate and precludes the use of a 1/2 key. Be very careful when mandrel balancing any part, especially when trying to get to a tight target such as 4W/N API, since a very small fit-up problem on the mandrel can lead to several times the value set by the tolerance. I recently wrote a short paper on the fallacy of attempting this - let me know and I will email you a copy.

During a tour of Thomas Coupling a few years ago, had opportunity to observe the balancing process for spacer disc coupling to API spect. The coupling was rigidized using washers and supported on the OD of the coupling flanges in the balancing machine. No arbor. Obviously the OD and ID must be concentric.

Thanks Ken.
I have done this myself as well, including at times skimming the periphery of the flange to provide a true running surface. My concern with this method is the problem with "rigidizing" (good word )Hard to ensure 100% perpendicularity to the bore of the couplings, and with a tapered bore hard to check. Even a very small axial runout will translate into radial offset when subsequently installed

Hello Ralph. My experience primarily revolves around field balancing, although I have spent some time behind the balancing stand. My last attempt to balance a coupling failed miserably. The coupling in fact had been balanced by a motor shop. The "new" coupling was drilled on the outer perimeter of both driver and driven flanges. I had some concerns about making yet more corrections as this meant drilling more holes in locations other than the existing holes, my fear was that I did not wish to further weaken the coupling. I enquired about this with the manufacturer, the technical engineer nearly came unglued when I advised that the coupling had been drilled. I was advised that the coupling would be void of warranty. This left the motor shop being obliged to supply the end user a new "untouched coupling". When this new coupling was mounted and the keys were correctly sized, 1x vibration was greatly reduced. To trim balance, I simply added a single washer to the coupling bolts and achieved better than satisfactory vibration. The moral of my story, is to be consider what you might leave yourself open to when balancing couplings. I'm not suggesting that you don't do it, but consider that the machine may experience 1x vibration from another source of unbalance on the rotor.

Ralph, some thoughts... To balance such objects as coupling halves and pump wheels we use in horisontal machines like your HB5 or ours HB4 a set of shafts that have a kit of variable dual conical inserts. They are used a lot for setting up any cylinder in a surface grinding machine. The two cones have when put together an inner and an outer diam. well fit to the mandrel and to the objects inner dia. I use a threaded nut to press it up to be fixed during balancing.
An index test 0-180 degrees flip is easy to do to verify the mandrel does not add something, neither runout nor unbalance. We drive the mandrel both with a belt and cardan, pending the size.
On keys, we apply the ISO 8821 standard, which has the intention that one should use half keys. When there is time, I always mill off the protruding key part.

I also like new tolerances of 1 mil on run-out and all stack tolerances must not exceed 2 mils. So 1 mil run-out on each end 180 out = 2 mils....... etc... +

Ralph,
I have worked in the petrochemical field where the compressors and turbines are all critical. Machine speeds 9000 to 12000 rpm.Any one of them meant the entire plant was down.
For those machines we had entire balanced couplings as spares. No mixing of hubs were allowed or inhouse balancing.
Any repairs/rebalancing was shipped back to Rexnord or the OEM.
Then there were other less critical application where we would allow switching of hubs especially if they were the shim pack type as this design lends to have low residual unbalance even when you swap hubs.
I guess my advice is you is similar to what Yvan was saying ,know the application and the consequences before attempting any job.
The OEM understands the stresses carried by the coupling and how much material can be allowed to be removed before the safety factor is compromised.

Regards,
James