a Five Basic Errors in Shaft Alignment
  By L. Robert Pyle, Owner
Systemaitec – Systematic Maintenance Technologies.
For more products related to this
topic visit our
Product Showcase
For other articles related to this topic visit our Reference Library.
Need more training?
Search Database of training courses and conferences on this subject in our Tradeshow/Seminar Search Section

Properly aligned shafts will do more to increase bearing, seal, and rotor life than any other single thing you can do after lubrication. Unfortunately, many maintenance departments in smaller plants still think that alignment is only needed for large, high-speed shafts on somebody else’s equipment. Many have no idea how to align two shafts beyond using a straight edge to get them close. Besides, the guy who sells the couplings says that the coupling can take up to one degree of misalignment and not hurt anything.

That is a pretty gross figure. They are correct, though. They design couplings that will not wear out with that much misalignment. The life of the coupling, though, is not controlling here. Badly aligned shafts, and by that I mean much less than the one degree of misalignment the coupling manufacturers use, will ruin the bearings on the equipment in short order. All shafts, even low speed ones, must be aligned to within a few thousands of an inch TIR (Total Indicator Runout) if the bearings are to last for their full expected life.

Typical Types of Misalignment in Shafts

There are some tremendous systems on the market for alignment. They use lasers, computers, and proximity sensors. They will practically move the equipment and install the shims. They do no good at all, though, if they are not used by trained, qualified technicians who understand what the systems are telling them.

This paper does not purport to explain shaft alignment. The presumption is that the reader already has some idea of how to align shafts using a dial-indicator set. What we want to do in this paper is to point out five common errors made while aligning and what to do about them.

Couplings ain’t round.

Suppose we have two shafts that are perfectly aligned but one has a coupling improperly mounted on a shaft so that there is some error in colinearity of the axes. As we approach the maching the error is only in the horizontal direction. Let’s set up the dial indicator so that it reads the outside of the coupling then rotate only the shaft on which the indicator is mounted, leaving the coupling still. When we get to position 1 the dial indicator will be extended to, say, -0.010 inches due to the error. Turning to position 2 will bring the dial indicator back to zero. At position 3 the indicator is compressed due the error and reads +0.010. If we believe that the coupling is square on the shaft and the axes are collinear we will believe that there is a TIR or 0.020 inches in the horizontal direction. Acting on that information will cause us to misalign the shafts by 0.010 inches – introducing vibration and potential bearing damage.

How can we eliminate that error? By rotating both shafts together. The roundness of the coupling, the roughness of its surface, and the poor mounting will be eliminated from the readings (actually, all of those things will cause compensating errors so that only the actual misalignment of the two shafts will be indicated by the dials.

A better way is to eliminate reading the coupling at all by using a target for your indicator that is securely mounted on the shaft. This will necessitate rotating both shafts but eliminates the need to break the coupling, even to do rim-face readings.

Move the Driver

This seems almost intuitive – at least if you have been in the field for any time. Yet, I repeatedly have had clients who inexplicably believed it was easier to move the driven machine. I will not argue that there is a time when the driven should be aligned to the driver. In all of industry, there are undoubtedly situations where this is better. As a general rule, though, it is better to move the driver that is not connected to your process than to move the driven that is.

For example, suppose we are aligning a pump and motor. The pump is connected to the process by an inlet pipe and an outlet pipe. There are some applications where the piping is connected through flex-joints but I haven’t found very many of them. Most of the time, the pump is hard-piped into the system. I will make an assumption that you have corrected any pipe strain in the system. If not, then do that before trying to align the pump. You are just kidding yourself otherwise.

Once the pump has been connected to the system, any attempt to move the pump to align it to the motor will induce pipe strain; whereas, moving the motor strains nothing but the flextite on the electrical connection – and it was made to be strained. If you move the pump and introduce pipe strain, you will end up with a misaligned pump as soon as you start the system up. Hot fluid, cold fluid or just the movement of the fluid will begin to flex the piping system which will, in turn, move your pump, which will in turn ruin that careful alignment you just accomplished.

Threads don’t increase Strength

At one client’s plant I gave a four-hour class on shaft alignment. At the end of the class, we went to the field to practice on a 50 Hp pump. This was a pump that had been “aligned” by that crew the week before. The plant had been in service for about two years and the bearings had been failing on this pump every six months. In the middle of the class, there had been some sheepish looks and, when I questioned them, they admitted that they had never heard of angular misalignment and had only been correcting offset since startup.

With information from the first pass of readings, the computer required a movement of the back end of the motor which was 0.010 to 0.015 more than the holes in the motor mount would allow. After some cursing, the crew began to disassemble the coupling and pull all the mounting bolts. I stopped them and asked what they were planning. “We have to take the motor to the shop and enlarge the holes in the feet,” was the answer.

There are times when the two devices may be so misaligned that that kind of action may be necessary. Needing less than 0.015 inches is not that kind of misalignment. When you run across this situation, do this: Pull the offending bolt, take it to the shop and turn or grind the threads off in the area that is in contact with the motor foot. You can grind all the threads down to the root diameter and not effect the strength in tension at all. On a 5/8 –11 bolt the OD at the threads is 0.614 inches. The root is 0.515 inches. So, you can pick up 0.050 inches of movement with no decrease in strength by grinding the threads.

Besides, I have seen many alignments where the computer moved the offending foot right back after the second pass. Enlarging the hole in the foot would have been an extreme waste of time.

2 Planes are Better Than 1

I have seen alignment techniques that suggest that the technician align first the horizontal then the vertical. I disagree with that. I believe you should be making complete passes – all four positions, and calculating both your vertical and horizontal correction at the same time. Here’s why:

If the alignment is far out, it is better to correct both planes (horizontal and vertical) at the same time. Otherwise the requirement for positioning the indicators becomes too precise. Error will be introduced and you will end up bouncing around the solution for any one plane.

Suppose you are nearly perfect in the vertical direction yet still grossly misaligned in the horizontal as illustrated in the above figure. If you do not place your dial indicator exactly vertically, part of the gross horizontal misalignment will be read as vertical misalignment. This will cause you to bounce around the solutions with each pass.

It is better to remove the error from both planes together so that this situation does not arise. With the availability of computers, laser systems, etc. there is not reason not correct both planes simultaneously.

Reverse Indicator Works

Most experienced mechanics and engineers will shrug at that statement. Of course, reverse indicator systems work. My experience as a teacher and as a consultant belies that complacency. I have found experienced crews who knew nothing but rim-face techniques or who knew of RI but did not believe it worked. I had one very experienced technician tell me that RI only worked on very large shafts.

To correct this thinking, I utilize a very old technique popular and needed prior to the advent of hand held calculators and computers: the plotting board technique.

In it, the distance between in the indicators is plotted as X. The distance from the B indicator and the front foot of the moveable machine is plotted as Y. The distance between the feet of the machine is plotted as Z. So, the horizontal axis is normally in inches. The vertical axis, though, is not in inches but in thousands of an inch.

On the A line is plotted the TIR of the A indicator in the plane being corrected (I know, I just told you to correct both at the same time – only, we didn’t have the calculating technology back then.) The B indicator TIR is plotted on the B line. A straight line is drawn through them and extended all the way to the Rear line. The needed correction at each foot is read off the vertical axis where the plotted line intersects the “foot” lines.

In the example shown the A TIR was +0.003 inches. The B TIR was +0.002 inches. The plotted line crossed the Centerline very near the Front line and crossed the Rear line at approximately –0.0015 inches. If this was the vertical plane being measured, then the front legs would need no correction and the rear legs would need a 0.0015 shim added to bring it up to the centerline.

Normally, when this technique is shown to the technicians, a light bulb goes on and they are able to grasp and appreciate the RI technique.

RI is not right in every situation. It is just another tool the alignment technician has for getting the job done. Its biggest advantage is ease of set up and a balanced rotor compared to Rim-Face (RF). The RI set up is balanced and will not rotate unless made to. The RF is unbalanced and will rotate to the bottom position unless held in place.

These five errors, although simple and basic to the experienced mechanic, are ones that I have found many clients making – and making without even realizing there was an error involved. Teaching these things to an inexperienced crew or to your new craftspeople will improve your alignments and make them more time efficient.

L. Robert Pyle
Systemaitec First North American Serial Rights only
12505 Rolling Hills Copyright 2002 L. R. Pyle
Wichita, KS 67235 316-722-6433
If you have any comments about the article you have just read and you would like to share them with us at Maintenance Resources, please feel free to email us by clicking on the email button below.
Current Issue
E-mail Us
© Copyright 2002 Maintenance Resources, Inc.
Phone: 812.877.7119  -  Fax: 812.877.7116  -  E-Mail: info@maintenanceresources.com
Address: 1983 North Hunt Street  -  Terre Haute, IN 47805