Original Date: 04/07/1997
Revision Date: 04/14/2003
Best Practice : Preventive/Predictive Machine Maintenance
(equipment: Entek IRD Fact Track FS and dataPAC 1500 Data Collectors; Entek IRD 246, 245 Balancers, and B1500 Balancing Stand; Entek IRD 350, 880, and 885 Analyzers; Entek IRD Emonitor for Windows SE and Eshape Software; Inframetrics Thermagram PM 250 IR Camera; Enframetrics Thermagram Software; Dycam Model 10-C Digital Camera)
Preventive maintenance (PM) at Rock Island Arsenal (RIA) involved an annual shutdown of the foundry for two weeks. The philosophy was "if it ain't broke don't fix it." Scheduling was done manually which resulted in a very costly PM program that did not prevent failures. Once a machine did fail, the work either was at a standstill or had to be moved. Either avenue was a cost to the installation and the customer.
Realizing this, management set about changing this process in 1985. A new program was established that incorporated predictive maintenance. The new program utilized the Arsenal's computer system to do scheduling. This included Occupational Safety and Health Administration (OSHA) inspections, quality inspections, and visual inspections. However, its cornerstone of success is the use of vibration analysis, infratography, and tribology (oil analysis).
All machines have vibrations. RIA’s personnel have found that all machinery troubles generate distinct vibration frequencies and, with proper analysis, problems can be detected and corrected prior to a failure occurring. Through the use of transducers mounted on the machinery, data is collected and fed to a computer. The computer then generates a graph (Figure 2-2) which, with proper analysis, helps determine which part is defective. The computer also produces trend data (Figure 2-3) which is used to predict the need for shutdown for repairs. The installation also uses vibration analysis to determine the proper levels of bearing lubrication. It has been noted that many bearing failures were due to over greasing. Through the analysis, RIA has been able to determine the proper grease levels for its machinery. Analysis of the oil contaminants is also used to help determine which parts are wearing faster and to project when failure will occur.
Thermography, or infrared imaging, is the other technique that is used to predict failure. By using an infrared camera, hot spots can be identified on the machinery indicating areas of heat buildup which could lead to failure.
The results of the Preventive/Predictive Maintenance program are dramatically obvious. The reduction in lost production time, from 180 hours using conventional troubleshooting to 29 hours using vibration analysis, is well documented. RIA has developed a high tech, low cost solution to lowering costs and being more productive.
Figure 2-2. 211 Maxi-Bar
Figure 2-3. Analyze/Trend Data
For more information see the
Point of Contact for this survey.