LASER OPTICS ACCURATELY MEASURE RUNNING SHAFT ALIGNMENT

Nov. 5, 1990
Heinz P. Bloch Consulting Engineer Montgomery, Tex. Laser-optic alignment systems can be employed effectively to detect and accurately measure rotary machinery shaft alignment while the machinery is running. These measurements can be used to maintain proper alignment at operating conditions. One of the many significant factors influencing machinery reliability in process plants is alignment accuracy. When driven process machines operate in misaligned condition with respect to their drivers,

Heinz P. Bloch
Consulting Engineer
Montgomery, Tex.

Laser-optic alignment systems can be employed effectively to detect and accurately measure rotary machinery shaft alignment while the machinery is running. These measurements can be used to maintain proper alignment at operating conditions.

One of the many significant factors influencing machinery reliability in process plants is alignment accuracy. When driven process machines operate in misaligned condition with respect to their drivers, equipment bearings are exposed to additional loads.

Vibration severity may increase, bearings will be more highly loaded, and equipment life expectancy will diminish as a result.

Although accurate initial or cold alignments are regularly done on process-plant compressors, pumps, blowers, and other rotary machinery, alignments can change significantly as the equipment reaches normal operating temperatures and conditions. Alignment changes because of movement caused by such things as thermal growth of the equipment, external piping forces, and movement in foundations and support members.

Initial or cold alignment specifications usually include values that are offset from zero misalignment to account for the movement of both driver and driven machine so that the running alignment is within specifications at normal operating conditions.

Offset values are derived from thermal growth calculations and empirical data from machinery tests and may not account for actual machine movements in a specific application. If alignment can be checked while a machine is running, corrections to initial alignment specifications can be made that will account for conditions present in the specific application and keep the machine in correct alignment.

To verify the accuracy and effectiveness of monitoring running alignment conditions, tests were conducted on a 2,000 hp process air compressor and driver. The tests showed that running alignment measurements are effective for maintaining specified alignment and for investigating the causes of unusual external forces and movements.

COLD ALIGNMENT INADEQUATE

There is hardly a major process plant in existence today which does not use one method or another to verify the initial, or standstill, alignment between driver and driven shafts on critically important machinery. Careful measurements use dial indicators, or for greater speed and precision, laser-optic devices. Both types of measurement allow the accurate initial alignment of machinery shafts in a nonrunning condition.

However, even the most knowledgeable operators have frequently experienced machine movement after start-up due to thermal, piping-related, and other mechanical or nonmechanical factors. This movement is extremely difficult, and sometimes impossible, to predict with purely analytical methods such as formulas that calculate thermal growth.

Consequently, alignment accuracy under operating conditions is often far from acceptable. Faced with inaccurate movement predictions, many plants are simply not able to achieve the degree of running alignment accuracy and hence equipment reliability that would otherwise be possible with sophisticated cold alignment procedures.

RUNNING ALIGNMENT MONITORS

Newly developed laser-optic, running-alignment verification systems are showing promise as an effective method to monitor and verify on-line shaft alignment. They are derived from their predecessor systems which use the laser optic technique for rapid and accurate initial (cold) shaft alignment (OGJ, Oct. 12, 1987, p. 71). Fig. 1 depicts one such device that has been in successful use in a number of major machinery installations worldwide.

The running alignment monitor uses the basic principles of the cold alignment system but detects and displays relative positional changes between two points in terms of direction and magnitude while a machine is running.

As an example of the principle of operation, when directed toward an ordinary mirror, a flashlight beam may be reflected straight back to the flashlight when the mirror is perpendicular the light beam. If the mirror is moved, the light beam is reflected in a different direction.

The on-line laser optic device operates the same way in that it detects the change in direction of an accurate laser beam reflected from either a 90 roof prism or triple prism, depending on the application. The laser device accurately measures and calculates the magnitude and direction of the change in the beam's reflection.

If the reflective device is mounted on one of the drive train components and the laser source is mounted on the other, the system provides an accurate way to detect and measure relative movement between the two components.

It is important to distinguish between the well established laser-optic devices used for cold or standstill alignments and the more recently developed running alignment systems. In this article, the terms "Locam," for laser-optic cold-alignment monitor, and "Loram," for laser-optic running-alignment monitor are used to distinguish the two systems.

The laser source and detector and prism of existing Locam systems used for cold alignment are usually mounted on the drive and driven shafts (Fig. 1). With the Loram system, however, the laser source and detector are mounted on a suitable stationary surface of one component and the prism on a stationary surface of the other component.

MISALIGNED MACHINES

At a major company, a Loram system was recently mounted on a 2,000-hp, high-speed centrifugal air compressor train. Fig. 2 shows a schematic view of this particular machine of which many hundreds are in service worldwide.

To begin with, the compressor was precisely aligned to its driver before the machine was started. Special attention was paid to the alignment using the Locam system to ensure the alignment values were in accordance with the compressor manufacturer's specified values. The specified values included predetermined offsets.

The compressor was started and the relative movements, thermal and otherwise, between the driver and compressor were measured in both the vertical and horizontal directions.

Not only were thermal movements found to be present, but there were also considerable mechanically induced changes in alignment. Air drafts on the machine and loading and unloading of the compressor also caused noticeable changes in alignment during operation.

The relative movements between the compressor and driver far exceeded the anticipated alignment changes of the manufacturer and plant operations people.

The alignment verification and monitoring task started out by installing the compressor and electric motor in accordance with the manufacturer's instructions. Shaft alignment and vertical offsets conformed to the prescribed values shown in Fig. 3.

Initial alignment verification was obtained with the Locam system shown in Fig. 4, and, for the sake of comparison, with the traditional reverse dial indicator method. Next, the Loram system was in stalled and activated (also shown in Fig. 4).

The compressor was started and the relative shaft movement between motor and compressor plotted. Fig. 5 illustrates the extent of the relative movement after several hours of operation.

The values show that the motor back end moved 0.1205 in., and the front end moved 0.066 in. relative to the compressor shaft centerline.

It could be said that the compressor's front feet shrank 0.0185 in. and the compressor's back feet shrank 0.041 in. relative to the motor shaft centerline.

Because both compressor and motor actually heated up, the perceived shrinkage simply means that both the motor and compressor bodies grew more at the front feet than they did at the back feet. Fig. 6 depicts the movement based on data collected on the machinery train during warm up.

The motor's back feet grew 0.005 in., and its front feet grew 0.011 in. The compressor's front feet grew 0.020 in., and its back feet grew 0.005 in.

One of the concerns sometimes voiced with Loram applications relates to the possibility of the laser-optic device being mounted on surfaces that can move relative to the component's shaft due to uneven thermal growth or nonlinear movement. Any movement of the mounting surfaces relative to the shaft would obviously cause errors in any running alignment measurement.

To check for the possibility of mounting surface movement, the Locam system was used to check the alignment after the Loram system was mounted.

Then, after the compressor was started and had reached a stable operating temperature, the machine was shut down, and hot alignment readings were taken immediately with the Locam system.

At the time of the hot alignment readings, the Loram units continued to record at 1-min intervals. All Locam readings fell between the Loram readings just before and after the Loram readings were taken.

This verified that the Locam and Loram readings were in agreement three times: cold, hot (16 hr later), and partial cool down. Also, the possibility of misalignment readings being influenced by bracket warpage or some other nonlinear movement could be discounted.

It was also noted that the compressor train, like many other major machinery trains, did not return to its initial alignment conditions upon shutdown and cooldown. Knowledgeable sources report having observed this nonrepeatability rather frequently, usually upon start-up after extensive rebuilding.

When good correlation is obtained between out of specification Locam and Loram readings, serious movement and, hence, misalignment are known to exist. Physical explanations should be sought. In one location, differential settling between the foundation under a steam turbine driver and the foundation under a large process compressor was found to be the source of misalignment.

In several installations, significant alignment changes were attributed to unexpected forces and moments exerted by piping on the suction and discharge nozzles of process pumps.

In any event, the running alignment of major process rotating machinery is sometimes surprisingly severe and can result in considerable reductions in the mean time before failure of these machines.

Laser-optic running-alignment monitoring must, therefore, be considered one of the more important machinery-related reliability improvement tools.

Already well proven in a number of installations, Loram can be termed a cost effective means of ensuring the long-term satisfactory operation of major process machines.

Copyright 1990 Oil & Gas Journal. All Rights Reserved.