It all started with a screwdriver
By Peter Phillips
This month, I'd like to focus on a tool that has been in use since the 1950s. In its early days, it was compared to the common screwdriver. In fact, the screwdriver helped to develop the parameters we use with this tool to determine the...
This month, I’d like to focus on a tool that has been in use since the 1950s. In its early days, it was compared to the common screwdriver. In fact, the screwdriver helped to develop the parameters we use with this tool to determine the condition of rotating equipment in our facilities.
The tool is vibration analysis (VA) and today a predictive maintenance program couldn’t function without it. VA in the early days wasn’t available to use right off the shelf and usually had to be invented and produced on the spot, where it was needed. Then in 1954, recognizing the demand for this type of equipment, Jim Chadwick and Jim Helmuth started a company that successfully produced vibration measurement systems.
Some of their first systems were applied to the manufacturing of Hughes helicopter engines. Over the years, these field-ready units were used by every maintenance facility that serviced aircraft engines.
In Canada, the application of VA began in 1974 when the Canadian Armed Forces started using it on its Sea King helicopters.
To get the story of these early days of VA in the Canadian Navy, I spoke with Jolson Sanford of Windsor, NS. A retired Navy Master Chief, Sanford helped to develop VA equipment and the early standards for measuring vibration in turbine aircraft engines.
I asked him what it was like in the beginning, when vibration analysis was a brand new tool.
“We used a vibration kit called Vibrex, made by Chadwick Helmuth, which was the only one available at the time. Before VA was introduced into the navy, we used a screwdriver held against the bearing housings of the Sea King engine. With the screwdriver grip against our ear, we determined if bearings were good or bad.
“This was the only way at the time to troubleshoot faulty bearings. As you got experienced listening to the sounds you felt and heard, you could determine which bearings were close to failing.
“When VA was first introduced, there were no baseline or benchmark readings to determine the condition of the Sea King engine bearings. In the beginning, we used our screwdrivers in conjunction with VA pickups (sensors) attached to the bearing housings. Back then, we even had to make the stand-offs (brackets) in the machine shop to mount the vibration pickups to the bearing housings.
“Once we had them mounted, we would compare the reading on the graph that the VA gear would print out, with what we heard with our screwdriver. If we found a bad bearing with our screwdriver, we would see if the graph corresponded. The graph would have high peaks if the bearing was failing. Bearings with the highest readings would score a ‘4’ on the graph. Other questionable bearings would score lower, usually a ‘1’ or ‘2’.
“Then we would dismantle the engine and check and measure the bearing. We paid special attention to the ones we found bad with our screwdriver and VA. This is the way we developed the first benchmark readings.
“Over a period of time we would chart every engine. Back then, a helicopter turbine engine had to be torn down and rebuilt every 200 hours of run or flight time. We’d put the engine in the test cell and take VA measurements on every bearing. Then we’d tear it down and examine the bearings, comparing their condition to their corresponding VA readings.
“We would replace all the bearings during the rebuild and put the unit back into the test cell again for multiple starts and run-ups, while taking more VA measurements. These measurements and repairs were recorded for each engine on recipe cards. Along with the VA graph, the card would be posted on the walls of the test cell office. One of the test offices we had was 12 x 14 ft. We had hundreds of these cards covering every wall in the room. This was the history of every repair that was done to each engine.”
Sanford added that these early years were interesting. “The first VA test carts we built were mobile. The engine would be bolted on to the cart and wheeled behind an earth berm and tied down with one-inch cables. We stood behind an earth berm in case the engine blew up.
“The Sea King T58 GE engine absolutely screamed when it was turned up. The compressor rotated at 26,000 rpm and created 4,200 hp. We really took our lives into our own hands testing those engines in the open air without any protection. Once, in a lead-lined indoor test cell, an engine blew up and the turbine blades were stuck in the walls around the room. Imagine what would have happened if we had tested it outdoors just with a test cart.
“In the 1980s, we built test cells with thick concrete walls in Shearwater and Greenwood, NS. By 1984, most Canadian Forces bases had engine test cells and vibration analysis equipment. When we went to sea, we’d pick up a VA kit to take with us. We’d need it to do repairs and services to the helicopters on the ship.
“As we refined our use of VA on our Sea King helicopters and Aurora aircraft, our efforts began to be noticed by our military colleagues. In 1985, the American Air Force contacted us to go to Memphis to help them build a test cart for their Sea Stallion helicopter. By then we had developed the benchmark readings and it was easy to determine the condition of the engine bearings. They were quite impressed with our methods.
“As time went on, engine reliability improved and the 200-hour teardown and rebuild intervals were lengthened. This had a lot to do with the historical data we had collected on our recipe cards, and our ability to predict failures and test engines.”
Sanford said that many of their bearing problems occurred due to lubrication issues. Faulty lubrication pumps and plugged lines caused premature failures. Whether it’s in the 1970s or now, the reasons for such failures are the same. Lack of lubrication is still a major cause of bearing failure.
Interestingly, the methods of measurement really have not changed much these days either, as experienced maintenance practitioners still use screwdrivers to listen to bearings. The big difference today is the availability of the new technology that has been incorporated in modern vibration analysis equipment.
New VA equipment comes in handheld units that can be quickly used to measure a bearing’s condition; it simply displays a good or bad reading. More sophisticated units measure and analyze the collected data, and though analytic software provide users with useful graphs and reports of the results.
Today, many companies have purchased vibration analysis units, and many of those who haven’t done so yet continue to purchase services from vibration analysis specialists.
Although VA is more commonly referred to now as condition monitoring, the latest equipment allows vibration readings to be recorded over time using trend analysis software. Gone are the paper records and recipe cards. Today, everything’s right at your fingertips.
Back in Sanford’s day, VA was only capable of measuring high rpm readings. Today, full spectrum measurements can be taken on critically low-speed rotating equipment to ultra-high-frequency centrifugal compressors that rotate up to 80,000 Hz.
Variable-speed equipment measurements, which once were difficult if not impossible to record and analyze, now can be taken as modern analyzers automatically adapt their diagnostic tools to variable turning speeds during routine data collection.
Even more advanced instruments use dual channels to correlate machinery vibration with process variables. This is accomplished by inputting the process information as a volt signal into one channel, while monitoring vibration on the other.
Of course, vibration readings can be recorded in a CMMS system and you can let it do the data crunching. It can then inform you that readings are out of spec and maintenance on the equipment needs to be scheduled.
This year, 2010, marks the 100th anniversary of the Royal Canadian Navy. I am pleased to dedicate this article to all those retired and active service people who have developed many tools and measurements that we now use
in everyday industrial life.
Peter Phillips of Trailwalk Holdings, a CMMS consulting and training company based in Nova Scotia, can be reached at 902-798-3601 or by e-mail at email@example.com.