Up Time Magazine- June/July 2008

What is the evidence for my argument? A typical scenario in my line of work is a bearing fault on a fan. Sometimes this will be a repeat failure because a post mortem was not completed when the previous failure occurred, or because the true root cause was not found and eliminated. I will come back to root cause analysis later, but for now let’s look at the typical causes of fan bearing failures. These include: • Poor fitting adapter sleeves • Having both bearings on a shaft fixed when one should be floating • Misalignment of bearings or their housings • Shaft under tolerance (typically due to repeated bearing changes) • Distorted bearings (e.g. misaligned bearing cap due to locating pin damage) • Poor lubrication (e.g. incorrect grease nipple location) • etc., etc., etc. And there is a similar list for every other type of fault. The point is that these failures are skills and practices issues, and they need not occur. You would not need your current level of resources if these avoidable problems were eliminated. There would be much less effort required in developing maintenance strategies, and carrying them out, as the failure modes would be eliminated. This would lead to an ongoing cost reduction right away. Rather than rely on further examples from my experience let’s have a look at the evidence from other sources. Pride and Wilson1 found that up to 2/3 of all machines have a failure probability versus time relationship as shown in Figure 1. This says that as many as 2/3 of all machines suffer infant mortality. We do not have to accept infant mortality! I contend that infant mortality is the result of skills and practices issues. If we think of infant mortality in terms of “failures after maintenance”, it is clear that they are self inflicted. Even the term infant mortality is misleading. It implies that there is something inherent in the equipment that causes the early life Failures Time Figure 1 - One study found that up to 2/3 of machines suffer infant mortality failures. This is simply not the case. But not all failures after maintenance fall into the infant mortality category. Some faults take time to develop, and, therefore, the association between the maintenance activity and the fault is not made. So the problem is even bigger than this infant mortality data suggests. Latino2 supports this conclusion. He says that only 10% of equipment reaches its design life. So 90% of premature failures are the result of what he calls “personnel avoidable events”. We have to exclude from this those things that I consider to be consumables. Those parts of the equipment that interact with the Our Students Call It CSI for Engineers. We call it “Practical Plant Failure Analysis” It’s a reasonably-priced practical seminar with hundreds of hands-on examples designed for plant people and engineers. You’ll work with gears, bearings, shafts, belts, fasteners, seals, and a selection of corrosion examples. With the hands-on analyses, you’ll learn how and why they fail – and how to diagnose the multiple causes. The public session is three days long and is held in Syracuse,NY in November. Private sessions range from two to four days and can be held at your site. From the Reliability Professionals who “wrote the book on logical and practical plant failure analysis”. For more details contact Dale Gamba at 315-487-4390 or email us at reliable@twcny.rr.com 10 june/july 2008 http://www.sengenuity.com http://www.sengenuity.com

Table of Contents Feed for the Digital Edition of Up Time Magazine- June/July 2008

Up Time Magazine- June/July 2008 Contents Upfront Upclose: The True Cause of Most Machine Failures Information Technology: Finding Future Perfect Lubrication: Building an Oil Analysis Winner Maintenance Management: Hidden Ways to Increase Safety Infrared: Air Barriers Lower Energy Costs, Increase Profits Motor Testing: Efficient Motor Repair, Part Deux Precision Maintenance: Good Maintenance, Smart Maintenance Reliability: Common Errors in Root Cause Analysis Ultrasound: Avoid the Baghouse Blues Vibration: Don't Let Your Sensors Get Eaten Alive Upgrade: Lining It All Up

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.