Reciprocating Compressor Frame Distortion During a Cold Start

Abstract

This paper presents measurements of vertical and horizontal distortions of a compressor frame during the days following a cold start. The distortions are measured using a laser beam directed along the back of the compressor at the level of the bearings. Laser targets attached to the frame generate voltages proportional to the vertical and horizontal displacement of the targets relative to the laser beam. Data are acquired by a microcomputer and high-speed A-to-D converter. Coherent averaging techniques reduce the influence of vibration on these readings to a tolerable level. Frame distortion readings are supplemented by temperatures measured at a number of points on the compressor frame and foundation block. Averaged signals from laser targets and thermocouples are logged by the computer every 2 min. Two sets of results are presented: one for a compressor mounted on a concrete block with a sand–cement full bed grout, the other for a compressor mounted on epoxy chocks. Data for the full bed grout unit cover a 53-h period. Data for the chock-mounted unit cover a 96-h period. Both units show some unexpected distortion transients during the first 24 h, and the magnitude of these is significantly greater for the full bed grout unit. Both units exhibit pronounced swings in frame distortion as ambient temperature changes from day to night. This daily cycling with ambient temperature, in particular, shows the need for careful interpretation of isolated readings of frame or crankshaft bearing alignment. The frame distortion and temperature data are supplemented by predictions of nominal crankshaft bending stress resulting from the observed misalignment.