Alignment of vertical shaft relative to pulley and plates

Dear All,

To begin, I apologise for any really stupid comments below. I've tried to watch a variety of videos covering shaft alignment, and read several articles/guides, but my measurements continue to confuse me (quite probably because I'm doing it wrong!)

I have a vertical shaft passing through in the simplest terms two pairs of ball bearings, one at the bottom (which is fixed) and one at the top, which I can translate horizontally using what I believe are termed leveling screws.

At the top of the shaft, a toothed pulley is placed, which lies between two other toothed pulleys, one of which is below a servo-motor, coupled with a flexible coupling. The other can be moved horizontally to loosen or tighten the belt, and there are additional tightening pulleys (without teeth) for the outsides of the belt which allow some additional tensioning after the toothed pulley is fixed.

Above the shaft pulley, there is a fixed coupling to a split ring. This is basically just a piece of brass pipe with some setscrews on the sides at the top and bottom to affix to the shaft and to the split ring rotor.

Given the fixed coupling, I'm trying to adjust the shaft angle/position (at the top) using the leveling screws, and the co-axiality(?) of the coupling to the shaft using pieces of aluminium tape as shims, such that the rotor of the split ring isn't experiencing large movement in the horizontal plane (excluding of course the rotation).

I'm currently using a dial indicator mounted to the frame with the dial against the coupling or alternatively the shaft (when I remove the pulley to allow some minimal access and turn by hand). My understanding is that readings at 9 and 3 on a clock should sum to the same value as 12 and 6, which is not what I observe. I gather this means I'm probably doing something wrong, but I can't see what it is.

My readings are moderately repeatable - for example when measuring on the shaft directly (not the coupling) I get going in one direction 4, 3.5, 1, -3 and then 5 at the starting position, followed by (the same) 5, 5, 1, -5 and 3 back at the starting position. One increment is I believe 0.01mm. The second set is fairly similar to the first, but the sums of opposite points is not even remotely similar. When measuring the coupling a similar pattern is visible, but magnified by a factor of about 4. This might be due to measuring further from the ball bearing which constrains the top of the axis?

How can I proceed to try and reduce the noise produced by the slip ring thudding back and forth as the shaft rotates? I can't visualize what is wrong to allow these measurements to be made. The shaft runs at speeds between 0 and around 2000rpm, in either direction.

If you have a shaft with two pairs of bearings there will be two at one end and it will be over constrained. You also talk about a pulley coupled to a servo motor by a flexible coupling - not sure how you can do that directly.

You need to do a drawing or give us a picture.

Attached is a collected PNG of several parts of the schematics. Notes are added in Paint (sorry!). Some parts are not completely accurate - the pulleys no longer have flanges, the collar is much longer than shown in the schematic and there are two pairs of pulleys on the outside of the belt - not two individual pulleys as shown. Hopefully none of these changes are particularly relevant beyond the longer collar with regards alignment.


Edit: That seems to have been downscaled a lot - a link to an alternative upload if it's completely unreadable for you (as it appears to be for me) is My Images for cbutcher - Discussion Forums - National Instruments

sounds like you are measuring the shaft run out. The indicator must be mounted to the fixed shaft so that the indicator reads the shaft to be moved.

muckalee said:

sounds like you are measuring the shaft run out. The indicator must be mounted to the fixed shaft so that the indicator reads the shaft to be moved.

Click to expand...

I expect you're right - rereading the wikipedia page on Run-out makes it sound like that's exactly what I'm measuring. Can I do anything with this measurement?

I'm not completely sure how to apply your instruction - I can keep the indicator reading the top of the shaft, but currently the fixed point is mounted to the frame, which is where the slip ring will be mounted. Does the indicator need to rotate? In the videos I've seen, often there are two dials in what I gather is a "reverse dial method". These remain in line with one another, but it seems like they're also rigidly attached to their respective ends of shafts. This makes me a bit confused as to why their measurements ever change - aren't they measuring the same point all the time?

I have multiple dials, but not a great deal of space to attach anything, and in particular I'm not certain I can mount anything to the slip ring (or if the collar is attached to the slip ring rather than shaft, the collar) because neither are magnetic and there isn't really room to use a clamp and rotate.

You want the indicator mounted to something that does not move with the shaft, and will not move when you try to center the shaft.

Turn the shaft until you get to one extreme of the readings on the indicator (say, -5) Now turn the indicator face so that is zero. This does not change anything, it just makes it easier to keep track of in your head.

Put a very small mark on the shaft that lines up with the indicator stylus at this extreme (sharpie mark, say)

Now turn the shaft as you have been - you should be able to get reasonable repetition from say 0 to 10. That range is Total Indicated Runout.
Repeat this a time or two, and you should have a decent sense of where the "high spot" and the "low spot" are.

You want to move the "high spot" "in" - that is, make it tighter, etc. = by ONE HALF of the total indicated runout (TIR)

So if your TIR is 10, try to move the shaft in about 5 from the high spot.
You adjust the adjustments you have available until the total indicated runout is as small as possible.

This task is very similar to centering a workpiece in a four jaw chuck - if you look at videos about how to center a workpiece in a four jaw chuck those may help you more than other things. Different task, very similar principles.

ALSO - how low you can get the runout will be limited by at least the following:
a. The slop in the bearings - you may need to adjust preload or some other aspect of the bearing setup, or the bearings may be worn out.
b. How true the shaft is. If the shaft is bent very much, you can have perfect runout at the bearing and still have lots of wiggling at the end.

FINALLY - while getting rid of runout is a fine thing to do, it is not necessarily the only or even main source of clunking and vibration in that rather complicated machine.

Reverse indicator alignment

cbutcher said:

I expect you're right - rereading the wikipedia page on Run-out makes it sound like that's exactly what I'm measuring. Can I do anything with this measurement?

I'm not completely sure how to apply your instruction - I can keep the indicator reading the top of the shaft, but currently the fixed point is mounted to the frame, which is where the slip ring will be mounted. Does the indicator need to rotate? In the videos I've seen, often there are two dials in what I gather is a "reverse dial method". These remain in line with one another, but it seems like they're also rigidly attached to their respective ends of shafts. This makes me a bit confused as to why their measurements ever change - aren't they measuring the same point all the time?

I have multiple dials, but not a great deal of space to attach anything, and in particular I'm not certain I can mount anything to the slip ring (or if the collar is attached to the slip ring rather than shaft, the collar) because neither are magnetic and there isn't really room to use a clamp and rotate.

Click to expand...

If you are lining up 2 shafts that are attached together by a flexible coupling between bearings proceed with checking alignment, if it is a solid coupling you will have to disconnect it. Then if you have room for 2 indicators and their brackets, attach 1st indicator to one shaft and read across gap to opposite shaft, attach 2nd indicator on other shaft and read across to other shaft. Turn the shafts together, staying near same spots. You can graph this to show what distance you have to move opposite ends of shaft/s If you want me to explain the graphing process, I will be happy to.

John