Research Papers

Seeded Fault Testing and Classification of Dynamically Loaded Floating Ring Compressor Bearings

[+] Author and Article Information
Markus Holzenkamp

Department of Mechanical Engineering,
Rochester Institute of Technology,
76 Lomb Memorial Dr., Rochester, NY 14623
e-mail: markus.holzenkamp@gmail.com

Jason R. Kolodziej

Department of Mechanical Engineering,
Rochester Institute of Technology,
76 Lomb Memorial Dr., Rochester, NY 14623
e-mail: jrkeme@rit.edu

Stephen Boedo

Department of Mechanical Engineering,
Rochester Institute of Technology,
76 Lomb Memorial Dr., Rochester, NY 14623
e-mail: sxbeme@rit.edu

Scott Delmontte

Dresser-Rand Company,
Painted Post, NY 14870
e-mail: SDelmontte@Dresser-Rand.com

1Corresponding author.

Manuscript received June 22, 2015; final manuscript received September 3, 2015; published online January 4, 2016. Assoc. Editor: Ioannis Kougioumtzoglou.

ASME J. Risk Uncertainty Part B 2(2), 021003 (Jan 04, 2016) (17 pages) Paper No: RISK-15-1079; doi: 10.1115/1.4031566 History: Received June 22, 2015; Accepted September 03, 2015

This paper investigates a variety of signal-monitoring and data-driven processing techniques to classify seed faults imposed on floating ring main crankshaft compressor bearings. Simulated main bearing shaft motion using an adaptation of the mobility method is first applied to demonstrate the plausibility of the method. Condition monitoring for three different fault types is experimentally investigated through seeded fault testing. A novel method for feature extraction utilizes a fast Fourier frequency-domain transformation coupled with a binning method that uses information across the entire frequency range. A principal component transformation process is then applied to reduce the dimension of the frequency-based feature vector to a small set of generalized features. A Bayesian classifier on the generalized features designed through seeded fault training data is shown to have excellent classifier performance across all fault types. A single-axis position measurement of the crankshaft shows the most promising results compared to a traditional accelerometer on the bearing housing and a novel accelerometer on the crankshaft. The single-axis measurement provides a cost-efficient alternative method to the two-axis orbit measurement typically used for traditional journal bearings.

Copyright © 2016 by ASME
Your Session has timed out. Please sign back in to continue.


Yang, B., Han, T., An, J.-L., Kim, H.-C., and Ahn, B.-H., 2004, “Technical Note: A Condition Classification System for Reciprocating Compressors,” Struct. Health Monit., 3(3), pp. 277–284. 10.1177/1475921704045628
Ahmed, M., Gu, F., and Ball, A., 2011, “Feature Selection and Fault Classification of Reciprocating Compressors Using a Genetic Algorithm and Probabilistic Neural Network,” J. Phys.: Conf. Ser., 305(1), pp. 1–12.
Jiang, X. F., Qiu, Z. H., and Zong, Y., 2011, “Air Compressor Wear Condition Monitoring Based on Oil Analysis Technology,” Appl. Mech. Mater., 66–68, pp. 498–503.
Kaewkongka, T., Joe Au, Y. H., Rakowski, R. T., and Jones, B. E., 2003, “A Comparative Study of Short Time Fourier Transform and Continuous Wavelet Transform for Bearing Condition Monitoring,” Int. J. COMADEM, 6(1), pp. 41–48.
Parnell, V. L., Boedo, S., Kempski, M. H., Kochersberger, K. B., and Haselkorn, M. H., 2007, “Health Monitoring of LAV Planet Gear Bushings Using Vibration Signature Analysis Techniques,” SAE 2007 Commercial Vehicle Engineering Congress and Exhibition, Rosemont, IL, SAE International.
Yang, D.-M., Stronach, A. F., Macconnell, P., and Penman, J., 2002, “Third-Order Spectral Techniques for the Diagnosis of Motor Bearing Condition Using Artificial Neural Networks,” Mech. Syst. Sig. Process., 16(2–3), pp. 391–411. 10.1006/mssp.2001.1469
Luo, G. Y., Osypiw, D., and Irle, M., 2003, “On-Line Vibration Analysis with Fast Continuous Wavelet Algorithm for Condition Monitoring of Bearing,” J. Vib. Control, 9(8), pp. 931–947. 1077-5463 10.1177/10775463030098002
Tandon, N., and Parey, A., 2006, “Condition Monitoring of Rotary Machines,” Condition Monitoring and Control for Intelligent Manufacturing, L. Whang, and R. Gao, eds., Springer Series in Advanced Manufacturing, London, pp. 109–136.
Chen, Y., Du, R., and Qu, L., 1995, “Fault Features of Large Rotating Machinery and Diagnosis Using Sensor Fusion,” J. Sound Vib., 188(2), pp. 227–242. 0022-460X 10.1006/jsvi.1995.0588
Rohde, S. M., and Ezzat, H. A., 1980, “Analysis of Dynamically Loaded Floating-Ring Bearings for Automotive Applications,” ASME J. Lubr. Technol., 102(3), pp. 271–276. 0022-2305 10.1115/1.3251501
Kumar, A., and Booker, J. F., 1991, “A Finite Element Cavitation Algorithm,” ASME J. Tribol., 113(2), pp. 276–286. 10.1115/1.2920617
Booker, J. F., 1971, “Dynamically-Loaded Journal Bearings: Numerical Application of the Mobility Method,” ASME J. Lubr. Technol., 93(1), pp. 168–176. 10.1115/1.3451507
Goenka, P. K., 1984, “Analytical Curve Fits for Solution Parameters of Dynamically Loaded Journal Bearings,” ASME J. Tribol., 106(4), pp. 421–427. 10.1115/1.3260950
Holzenkamp, M., 2013, “Modeling and Condition Monitoring of Fully Floating Reciprocating Compressor Main Bearings Using Data Driven Classification,” Rochester Institute of Technology, Rochester, NY.
Chirico, A. J., and Kolodziej, J. R., 2012, “Fault Detection and Isolation for Electro-Mechanical Actuators Using a Data-Driven Bayesian Classification,” SAE Int. J. Aerosp., 5(2), pp. 494–502. 10.4271/2012-01-2215
De Boe, P., and Golinval, J.-C., 2003, “Principal Component Analysis of a Piezosensor Array for Damage Localization,” Struct. Health Monit., 2(2), pp. 137–144. 10.1177/1475921703002002005
van der Heijden, F., Duin, R., de Ridder, D., and Tax, D., 2004, Classification, Parameter Estimation and State Estimation an Engineering Approach Using Matlab, Wiley, Chichester, England.


Grahic Jump Location
Fig. 1

D-R ESH-1 reciprocating compressor at Rochester Institute of Technology (RIT) (3628 kg, 3  m×1.8  m)

Grahic Jump Location
Fig. 2

Floating ring main journal bearing geometry

Grahic Jump Location
Fig. 3

Main bearing load components (journal to sleeve)

Grahic Jump Location
Fig. 4

Effect of viscosity on predicted shaft motion: Cj=35  μm and Cs=31  μm

Grahic Jump Location
Fig. 5

Effect of radial clearance on predicted shaft motion, SAE 30 oil

Grahic Jump Location
Fig. 6

Effect of grooving on predicted shaft motion, SAE 30 oil

Grahic Jump Location
Fig. 7

Proposed classification methodology

Grahic Jump Location
Fig. 8

Raw x-direction shaft data from the simulation

Grahic Jump Location
Fig. 9

Hamming windowed FFT’s of the shaft motion data for the three health classes with a frequency resolution of 0.53 Hz

Grahic Jump Location
Fig. 10

Binned FFT data used to demonstrate the methodology

Grahic Jump Location
Fig. 11

Basic classification process

Grahic Jump Location
Fig. 12

Final classifier output for simulated example (top) (0% misclassification), percentage of eigenvalue representation, first two components are 87% and 10%, respectively (bottom)

Grahic Jump Location
Fig. 13

Schematic showing the location and orientation of the grooved bearing in the compressor

Grahic Jump Location
Fig. 14

Bearing with grooves: solid model (left); implementation: two grooves (center); implementation: one groove (right)

Grahic Jump Location
Fig. 15

Bearing with feedhole obstruction: solid model (left); implementation: 75% (center), 50% (right)

Grahic Jump Location
Fig. 16

Bearing housing accelerometer mounted inside crankcase (left) and wireless accelerometer mounted to flywheel (right)

Grahic Jump Location
Fig. 17

LVDT mounting bracket and mounted on the outside of the crankcase (flywheel removed)

Grahic Jump Location
Fig. 18

Main bearing temperature from start to steady-state

Grahic Jump Location
Fig. 19

Clearance variation—100% load: bearing housing accelerometer raw single (left) and binned FFT (right)

Grahic Jump Location
Fig. 20

Clearance variation—100% load: classification (1.1% miss classification)

Grahic Jump Location
Fig. 21

Classification performance for validation data sets for the clearance variation. Top: bearing housing accelerometer (13.3%, 23.7%, 7.7%); middle: LVDT (0%, 0%, 0%); and bottom: wireless crankshaft accelerometer (50.3%, 55.7%, 59.3%) (misclassification is shown in parenthesis)

Grahic Jump Location
Fig. 22

Grooved bearing—50% load: LVDT raw signal (left) and binned FFT (right)

Grahic Jump Location
Fig. 23

Grooved bearing—50% load: classification (0% misclassification)

Grahic Jump Location
Fig. 24

Classification performance for validation data sets for the grooved bearing. Top: bearing housing accelerometer (20.7%, 19.0%, 20.3%); middle: LVDT (0%, 0%, 0%); and bottom: wireless crankshaft accelerometer (23.3%, 28.7%, 24.0%) (misclassification is shown in parenthesis)

Grahic Jump Location
Fig. 25

Enlarged view of the healthy and double-grooved bearing classification from the bearing housing accelerometer for zero load

Grahic Jump Location
Fig. 26

Classification performance for validation data sets for the oil feedhole obstruction. Top: bearing housing accelerometer (33.0%, 35.3%, 17.7%); middle: LVDT (0.3%, 0%, 0%); and bottom: wireless crankshaft accelerometer (34.3%, 36.0%, 26.7%) (misclassification is shown in parenthesis)

Grahic Jump Location
Fig. 27

Classification performance for validation data sets for the oil viscosity. Top: bearing housing accelerometer (14.3%, 19.0%, 2.0%); middle: LVDT (12.7%, 17.3%, 6.0%); and bottom: wireless crankshaft accelerometer (52.7%, 65.0%, 48.0%) (misclassification is shown in parenthesis)




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In