Shaft Alignment Questions

zipfactor · Dec 14, 2018

I'm attempting to align the shafts parallel to one another on this piece of machinery I'm putting together, and I'd like to be able to do so without purchasing expensive laser aligning tools, etc.

My first thoughts are to pick a reference shaft, hang plumb bobs for vertical reference lines at each end of the shaft, and measure from the shaft being aligned to those reference lines. I'm using a starrett protractor head with the rule set 90 degrees to the head in order to give two faces to butt up against the shaft, and then I use the built in level on the protractor head to level the rule up while taking a measurement at the reference lines.

Doing it this way is bit tedious and not very precise. I'm thinking there is a better way to do this. Any thoughts?

mmarquette · Dec 14, 2018

inside micrometer comes to mind.

zipfactor · Dec 14, 2018

mmarquette said:
inside micrometer comes to mind.

I agree, except the shafts are in different planes. The only way I can think to make this happen is to have multiple large(ish) right angle reference to butt up against the shafts.

HuFlungDung · Dec 14, 2018

I think you first need to establish a reference plane: use a straightedge and level to align two shafts level. Then, you could clamp a combination square to one end of your straight edge, with the scale hanging down far enough to catch one shaft. Then, use a second combination square in a similar manner to locate the other shaft and clamp the square to the straight edge. Then just move the straight edge/ double squares assembly from one end to the other and adjust the shaft position to get the bearings spaced equally.

While using an inside mic will work, it is very tedious to find the 'minimum swing through point' and won't actually be valid without both shafts lying in the same plane.

Once you have two shafts aligned, then continue on from there with the same rig, always working from one proven shaft to the next unknown one.

gbent · Dec 14, 2018

I can't see in the picture very well. Do you have to align two degrees of freedom at one time, or can the degrees of freedom be aligned separately?

If the alignments can be set separately, level all the shafts in the horizontal plane. Stand the unit upright and again level all the shafts. You could also stand the machine so the shafts are vertical and plumb the shafts, setting two degrees of freedom at once.

Otherwise use an inside micrometer. An inside micrometer could be off 30 degrees to the horizontal plane and still be quite useful to align horizontal shafts.

tdmidget · Dec 14, 2018

Inside mike. They do not all have to be coplaner. You are only concerned with the plane that the two being aligned are in. The way you are trying to do it you will only be accurate to the thickness of the line , at best.

HuFlungDung · Dec 14, 2018

tdmidget said:
Inside mike. They do not all have to be coplaner. You are only concerned with the plane that the two being aligned are in. The way you are trying to do it you will only be accurate to the thickness of the line , at best.

If they are not coplanar, then they are not parallel.

zipfactor · Dec 14, 2018

gbent said:
I can't see in the picture very well. Do you have to align two degrees of freedom at one time, or can the degrees of freedom be aligned separately?

If the alignments can be set separately, level all the shafts in the horizontal plane. Stand the unit upright and again level all the shafts. You could also stand the machine so the shafts are vertical and plumb the shafts, setting two degrees of freedom at once.

Otherwise use an inside micrometer. An inside micrometer could be off 30 degrees to the horizontal plane and still be quite useful to align horizontal shafts.

I can level the two axes independently, which makes it easier for sure. That's an interesting thought standing the unit upright, I may go that route as I don't have inside micrometers big enough to span between the shafts. The distance between shafts is on average around 18" apart.

tdmidget · Dec 14, 2018

HuFlungDung said:
If they are not coplanar, then they are not parallel.

Wrong. Each pair must be co planar. Each additional shaft creates a new plane that includes the previous shaft. They must all be parallel.
To start, shim each shaft to be level. Then with the inside mic adjust them to parallel. If you don't shim them to matching height end for end then the mic will not bring them to parallel. This appears to be for a conveyor of some kind. No matter how hard you work at this there may still be some tweaking required to get the belt to track properly due to irregularities of the belt it self. This method will give a good starting point so that you are not chasing your tail. I'm a millwright, we do this for a living.

zipfactor · Dec 14, 2018

tdmidget said:
Wrong. Each pair must be co planar. Each additional shaft creates a new plane that includes the previous shaft. They must all be parallel.
To start, shim each shaft to be level. Then with the inside mic adjust them to parallel. If you don't shim them to matching height end for end then the mic will not bring them to parallel. This appears to be for a conveyor of some kind. No matter how hard you work at this there may still be some tweaking required to get the belt to track properly due to irregularities of the belt it self. This method will give a good starting point so that you are not chasing your tail. I'm a millwright, we do this for a living.

Thanks for the feedback. It is a conveyor of sorts and I am fully expecting to still have to tweak it to get the system tracking properly. The last machine I put together was a royal pain as we didn't align the rollers to each other and it was a tweak and watch it run situation.

What do you use in instances where the shafts are further apart than a typical inside mic can measure? In this instance they are on average 18" apart. Do you use some type of gage block on the shaft?

tdmidget · Dec 14, 2018

I don't see a lot of them with that many rollers that close together. Most are pretty simple with longer distance. These are usually done to a tape measure where there is basically a head and tail sheave. Troughing or load rollers don't have as much effect. Your machine appears to change the belts direction several times and those rollers will have a major effect. At least this one does not go around a curve

.

awander · Dec 14, 2018

zipfactor said:
Thanks for the feedback. It is a conveyor of sorts and I am fully expecting to still have to tweak it to get the system tracking properly. The last machine I put together was a royal pain as we didn't align the rollers to each other and it was a tweak and watch it run situation.

What do you use in instances where the shafts are further apart than a typical inside mic can measure? In this instance they are on average 18" apart. Do you use some type of gage block on the shaft?

Easier to just get a inside mic that can do 18"...

Actually, since you don't need to know the actual distance, just an adjustable rod with rounded ends will allow you to compare left and right spacings...

HuFlungDung · Dec 14, 2018

tdmidget said:
Wrong. Each pair must be co planar. Each additional shaft creates a new plane that includes the previous shaft. They must all be parallel.
To start, shim each shaft to be level. Then with the inside mic adjust them to parallel. If you don't shim them to matching height end for end then the mic will not bring them to parallel. This appears to be for a conveyor of some kind. No matter how hard you work at this there may still be some tweaking required to get the belt to track properly due to irregularities of the belt it self. This method will give a good starting point so that you are not chasing your tail. I'm a millwright, we do this for a living.

That's what coplaner means: they must be parallel in pairs. Not that a half a dozen have to lie in the same plane.

tdmidget · Dec 14, 2018

Coplanar means in the same plane. The Cartesian x-y coordinate system is a plane. The X and Y axes are coplanar. The lines have 1 point in common but they are coplanar.

Erich · Dec 14, 2018

Make up a Rod with a V notch at one end and a dial indicator at the other.

Put the notch on one shaft and sweep the next shaft. Record minimum reading.
Go to the opposite end of the shaft and repeat. Adjust shaft until both ends give the same min reading.

Move on to the next shaft in line.

zipfactor · Dec 14, 2018

tdmidget said:
I don't see a lot of them with that many rollers that close together. Most are pretty simple with longer distance. These are usually done to a tape measure where there is basically a head and tail sheave. Troughing or load rollers don't have as much effect. Your machine appears to change the belts direction several times and those rollers will have a major effect. At least this one does not go around a curve.

Yeah, it's a specialized coating machine, so lots of support :/. Our last one was worse, believe it or not.

zipfactor · Dec 14, 2018

awander said:
Easier to just get a inside mic that can do 18"...

Actually, since you don't need to know the actual distance, just an adjustable rod with rounded ends will allow you to compare left and right spacings...

This is something I was starting to cook up, i figured a few shaft collars and tap it around a bit.

Illinoyance · Dec 14, 2018

With an inside mike or measuring rod it is neccesary to pick up the high point of both shafts simultaneously. I would much rather measure out-out on two shafts with vernier caliper.

John Garner · Dec 15, 2018

A couple of quick thoughts:

Think in terms of differences.

1) The frame will almost certainly change shape between its present location on the workbench and its finally-installed position, and the shaft alignment will tend to follow the frame.

2) It's often easier to use a good telescopic sight level and an improvised optical target to level rolls than it is to use a block level (Starrett Mod 98 and Mod 199z are block levels).

3) Improvised tram bars made from a broomstick, a plastic V -- which can be cut out of a plastic cutting board -- and a dial gage are a quick and easy way to quantify and correct differences in roll-end-to-roll-end distances.

4) You don't need to know the length of a spacer block or rod if you use the same spacer at both ends of a roll.

5) Double-stick tape makes a reasonable substitute for an aluminum magnet dial gage holder to measure adjustments as you make them.

Tyrone Shoelaces · Dec 15, 2018

zipfactor said:
Thanks for the feedback. It is a conveyor of sorts and I am fully expecting to still have to tweak it to get the system tracking properly. The last machine I put together was a royal pain as we didn't align the rollers to each other and it was a tweak and watch it run situation.

What do you use in instances where the shafts are further apart than a typical inside mic can measure? In this instance they are on average 18" apart. Do you use some type of gage block on the shaft?

We had screw together inside mic sets that went from 6" to 126".

Regards Tyrone.

Shaft Alignment Questions

Cast Iron

Cast Iron

Cast Iron

Diamond

Diamond

Diamond

Diamond

Cast Iron

Diamond

Cast Iron

Diamond

Stainless

Diamond

Diamond

Hot Rolled

Cast Iron

Cast Iron

Stainless

Titanium

Diamond

Similar threads