Optical Alignment, Cylinder to Crosshead Guide
Including perpendicular relationship between crosshead centerline and bed section and distance piece flange face mating surfaces.
Dresser-Clark 2-CLBA-1 Reciprocating Compressor
This machine is one of three similar units originally installed in 1982. They are operated as part of a natural gas reinjection system in a producing oil field. Alignment checks were performed as part of a system upgrade in 2002. New cylinders were being installed to handle larger gas volumes.
This unit has a chronic history of various failures including excessive packing wear, worn cylinder liners and cracked piston rods. Cylinder to crosshead guide alignment checks had previously been performed in 1988 and again on cylinder #1 in 1995, with the old cylinder configuration. Cylinder alignments were documented as exceeding manufacturer’s tolerances, but were not corrected at those times.
2002 Alignment: New cylinder installation was started in March 2002. Alignment data was obtained with standard optical tooling instruments. Cylinder to crosshead guide alignments were measured with an optical line scope. The telescope was mounted on a universal adjusting bracket attached to a custom shelf bracket, which was bolted to the cylinder head. Wire centering targets were installed in the cylinder liner and crosshead guide surfaces. Cylinder targets were held in the 7.500” diameter bore by custom made target holder plugs. These holders were fabricated by the customer using TERN specifications. Crosshead guide targets were mounted in custom built adjustable centering devices supplied by TERN.
The cylinder target plugs are machined to 7.498” diameters. This allows .002” clearance so that they may be slid into the 7.500” diameter cylinder liner. Thus, the center of the target rests .001” below the actual center of the cylinder liner. The optical data has been corrected to account for the cylinder targets being .001” low. Also, all readings are taken with the targets in a 0 degree and 180 degree orientation. This eliminates any error induced by eccentricity between the target plug inside diameter and outside diameter, or between the target wire center and wire target outside diameter. The average of these two readings, which are 180 degrees opposed, is reported as the measured value.
When the cylinders were first bolted into place the alignment was found to be totally unacceptable. Inspection revealed that the cylinder head spigot was coming into contact with the shoulder of the spigot bore. This situation was corrected by cutting a chamfer on the cylinder spigot. While this improved the alignment position of both cylinders, misalignment still exceed manufacturers’ specifications. Further inspection revealed that the cylinder flange face mating surface was not perpendicular to the cylinder liner. One cylinder had a run-out of .0065” and the others runout was .0045” over the bolting circumference. Maximum allowable runout is .002”. This was the first of three units being upgraded. Therefore, there were four other cylinders on-site. The run-out on these other cylinders was checked. Three were found to have runouts of .002’ or less. The best of the lot were installed. Cylinders with excessive runout were sent for re-machining.
After installing the cylinders with acceptable run-outs, misalignment was still outside the manufacturers’ maximum allowable limits. They were however, within the historical range of alignment readings found in 1988 and 1992 with the old style cylinders. Since the cylinders had now been thoroughly inspected the remaining misalignment pointed toward anomalies with the bed sections and/or distance pieces. However, due to the maintenance schedule and production considerations the unit was re-assembled and brought back on-line. It was decided that this machines alignment would be re-visited once the cylinder upgrades were completed on the other two machines. In the meantime, and in anticipation of possibly finding a mis-machined bed section or distance piece, a spare bed section and distance piece were located and expedited to the site.
The following photographs illustrate the optical line scope and target set-up used for this alignment check. The connecting rods were removed exposing the whole crosshead guide for measurement.
Picture One, Optical Line Scope Mounted on Cylinder
Picture Two, Adjustable target holders installed in crosshead guide
Alignment work commenced again in June 2002. Prior to cylinder installation the alignment relationship between the distance pieces and bed sections was verified. Data showed excessive misalignment. Next, the flange face mating surfaces of the bed sections were checked to verify their perpendicular relationship to the crosshead guide centerline. The #1 bed section was found to be out of square by .011” over the bolting diameter. The #2 bed section was found to be out of square by .006”.
The following photograph illustrates the optical set-up used to measure the perpendicularity of the flange face mating surfaces. In this application a special optical transit square with a focusable cross scope is used. This instrument is actually two telescopes built together such that they are at right angles to each other. With the transit adjusted to the centerline of the working piece, the cross scope is used to measure optical scales held on the flange face being measured. If the flange face is perpendicular to the crosshead guide centerline, all scale readings obtained would be equal. Unequal readings indicate the amount and direction which the flange face is out of square relative to the crosshead guide. The same adjustable target holders used to measure cylinder alignments are used to establish the center of the crosshead guide.
Picture Three, Cross Scope set-up on bed section
Note: Optical scales on upper right side of flange face and adjustable optical targets in background.A similar set-up was used to measure the distance pieces. Distance pieces were removed from the machine and measured in the shop. The distance pieces were found to be true with runouts at both ends within .002”.
While one spare bed section and distance piece combination was available for installation it was not used. Instead, the existing bed sections were field machined in place. A flange face cutting tool was mounted centered to the crosshead guide centerline and the flange face was skim cut to make it truly perpendicular. Before and after optical data is shown below.
THROW #1 THROW #2
As-Found After Machining As-Found After Machining
0 Degrees N/A N/A N/A N/A
45 Degrees 9.479” 9.4985” 9.508” 9.471”
90 Degrees 9.480” 9.4995” 9.511” 9.471”
135 Degrees 9.4825” 9.499” 9.514” 9.4725”
180 Degrees N/A N/A N/A N/A
225 Degrees 9.490” 9.499” 9.5125” 9.471”
270 Degrees 9.488” 9.4995” 9.511” 9.473”
315 Degrees 9.485” 9.499” 9.5095” 9.473
Maximum Deviation .011” .001” .006” .002”Picture Four, Flange face cutter on bed section
Cylinder Alignment Data: Data below expresses the cylinder centerline position relative to the crosshead guide centerline. Left (L) and right (R) is as viewed from the line scope location on the cylinder end looking toward the crosshead guide.
Reference Reference Vertical Horizontal
Crosshead Crosshead Cylinder Cylinder Cylinder Cylinder
Crank End Cyl. End Inner Outer
1988, Cylinder #1 .000” .000” -.0165” -.018” R .004” R .004”
1988, Cylinder #2 .000” .000” -.011” -.0145” R .0085” R.0085”
1995, Cylinder #1 .000” .000” +.0275” +.0305” L .004” L .006”
1995, Cylinder #2 . Data Not Recorded
2002, Cylinder #1 .000” .000” -.001” -.0035” R .002” R .0025”
2002, Cylinder #2 .000” .000” -.0095” -.0095” L .0075” L .010”The following alignment drawings show the 2002 final alignment position of the new cylinders.
Piping Movement Survey using Permalign® Laser Measurement System
Dresser-Clark Hot Gas Expander (Turbine), 60-inch Inlet Flange
Piping support modifications and piping replacements were being made as part of activities during an oil refinery turn-around. Similar modifications on a like unit at a “sister” refinery had resulted in major turbine casing rotor rubs. This caused major damage and prevented post-turnaround refinery start-up. The monitoring program was designed to insure that, in this instance, the inlet flange would not be moved so as to distort the turbine casing and cause a similar damage and shut-down. Movement was also monitored during piping dis-connect. This was simply to catalog what piping stresses the turbine casing was under prior to modifications. A “static” testing period was also added to show the normal amount of movement occurring that is unrelated to piping activity. Most of this movement is thought to be caused by ambient temperature fluctuations and changing sun position throughout the testing period.
A Permalign® monitoring system was used to collect movement data on the turbine inlet flange. A total of three laser monitors were used to measure movement in the vertical, horizontal and axial planes. Permalign® is a laser sourced measuring system. It senses relative movement of a prism and detector across distances of zero to thirty-three feet. By employing a reflected beam measuring principle, thermal and vibration stability are achieved. System resolution and repeatability are one micron with a maximum error of less than 2% of the displayed value. The high resolution, repeatability, and accuracy are due to the use of a linearized photo detector in the design.
Permalign® monitors may be fixed to the foundation or other positions of interest to reveal displacements or angles of those points along the horizontal (X) and vertical (Y) axis of the monitor. Standard Cartesian coordinate conventions apply to Permalign( monitors and are defined as if looking into the lens of the monitor. Hence, with the monitor oriented in a normal upright position, the X-axis is lateral data (horizontal or axial) and the Y-axis is vertical data.
For the purposes or this study, 90û roof prism reflectors were used. In other applications triple prisms may be used. Triple prisms respond to displacement movement along both axes and not to angular deviations. The 90û roof prism responds to displacement along the Y-axis and angular deviation along the X-axis.
The Permalign® laser system is used in conjunction with application software, Winperma®. Data is collected in spreadsheet and graphical formats. Winperma® allows the user to set the polling frequency for each monitor for measuring changes in displacement over time. Monitors were connected to a PC for the purposes of data collection. Winperma® was configured to poll data from the monitors every minute during piping disconnect and 15 seconds during piping re-connect activities. Following each poll, Winperma® averages and plots displacement information received from the biaxial detector.
Monitors were mounted to the turbine deck and I-beam supports. The movement recorded is relative to each monitor attachment point.
It is essential to understand machine and monitor orientation to properly interpret the Permalign® data. The machine centerline is said to run on a north/south axis, with the turbine inlet on the north end. The following photographs are provided to help illustrate the position and relationship of the various monitors and prisms. A total of three monitor/prism pairs were used for this study. Monitor #7 was mounted to a bracket support bolted to an I-Beam on the east side of the unit. Monitor #7 orientation is such that the Y-axis records horizontal displacement along the machine east-west axis. Angular deviation or rotation about the machine east/west axis is recorded on this monitors X-axis. Monitors #0 and #9 were mounted on a pipe stand support bolted to the west side of the turbine skid. Monitor #0 orientation was such that it recorded vertical displacement on the Y-axis. The X-axis recorded angular deviation or rotation about the vertical axis. Monitor #9 recorded axial displacement on the Y-axis. The monitor X-axis recorded angular deviation or rotation about the machine north/south axial axis.
Sample Winperma( graphs of displacement vs. time are shown below. All Y-axis data is shown in black (offset displacement). All X-axis data is graphed in “red” (orange) and is displacement representing angular rotation about the X-axis. Graphed Y-axis data is the actual displacement value recorded. The graphed X-axis value is the raw displacement value converted to an angular value expressed in mils per 60 inches (inlet piping flange diameter).
Monitor orientation is detailed below. Following this explanation photographs are provided to help further illustrate the laser installation and orientation.
Monitor #7, East Side, Horizontal and Axial Displacement
Monitor Orientation: The vertical Y-axis is oriented along the machine east/west axis such that it records horizontal offset values. Plus (+) is movement toward the east and minus (-) is toward the west. Therefore the horizontal
X-axis is oriented so that (+) is up and (-) is down. This axis shows displacement representing rotation about the east/west machine axis. Plus (+) values indicate rotation such that the bottom of the flange moves toward the north (toward the inlet) and the top moves to the south (toward the expander).
Monitors #0, West Side, Vertical and Axial Displacement
Monitor Orientation: The vertical Y-axis is oriented in the normal upright position such that (+) is up and (-) is down. It records vertical offset values on the west side of the flange. The horizontal X-axis is aligned along the north/south axial axis such that (+) is toward the north, or away from the expander, and (-) is toward the south, or toward the expander. The vertical Y-axis shows vertical offset. The horizontal X-axis shows horizontal angular change, or rotation about the vertical Y-axis. Plus (+) values indicate rotation with the west side of the flange moving toward the north, or away from the expander and the east side of the flange moving toward the expander, or south.
Monitors #9, East Side, Axial and Vertical Displacement
The vertical Y-axis is oriented along the machine north/south axial axis such that it records axial offset values on the west side for the flange. Plus (+) values represent movement toward the expander, or south, and minus (-) movement is away from the expander, or north. Therefore, the horizontal X-axis is oriented so that (+) is up and (-) is down. This axis shows vertical angular displacement or rotation about the machine north/south. Plus (+) values indicate rotation with the west side of the flange moving up relative to the east side of the flange. Minus (-) values show rotation in the opposite direction.
Prism #7, Monitors Horizontal Offset and Rotation about the east/west machine axis.
Monitor #7, Records Horizontal Offset and Rotation about the east/west machine axis.
Monitor/Prisms #0 and #9
Monitor #0 Records Vertical Offset and Rotation about the machine vertical axis.
Monitor #9 Records Axial Offset and Rotation about the north/south machine axis.
Note: Monitor #9 is labeled #F in this photograph
Permalign® Data Overview
A significant amount of movement was observed during piping disconnect. It was also observed that the inlet piping moved away from the expander inlet flange at unbolting. This movement was in excess of one inch, and indicated that there was a significant amount of cold pipe stress on the expander inlet flange. The largest movement observed is in the axial angle position of the flange. This value was .0563”/60” diameter. This may be because this is the easiest direction in which to influence flange position, or it may be that this was the direction of the greatest amount of existing pipe strain.
The most significant factor revealed during the static test on November 6, 2002 was the rapid progressive change in the horizontal offset reading at Monitor #7 between approximately 9:00 am and 11:30 am. Observed movement during this period was approximately .020”. Fluctuations in this position after 11:30 am were less than .010”. However, since pipe bolting and un-bolting activities occurred after 11:30 am the movement that occurred in the morning hours did not severely affect the interpretation of the movement caused by piping activity.
Very little expander flange movement was observed while attaching the new piping. Since this was the desired effect, the monitoring showed a successful outcome. Offset value changes, vertical, horizontal and axial, were each under .002”. They were -.0004”, -.0012” and -.0004” respectively. Angular value changes were each under .011” when expressed over a 60” diameter. This is approximately .002” per foot. Angle values were +.0016”/60”, -.0106”/60” and +.0036”/60” for axial, vertical and horizontal respectively. Also, the static test showed that as much as .0032” of the vertical angle rotation could be the result of normal fluctuation, and not the result of movement caused by piping activity. So the reported vertical angle of -.0106” can be interpreted as really being less, or -.0074”.
The following drawings illustrate the movement seen at:
- Piping dis-connect, and
- Attaching new piping