COMPONENT AGE VS. BEARING PERFORMANCE
The life of a journal roller bearing can be measured and predicted by the fatigue cycles and loading it experiences during its service life. When any metal component accumulates sufficient “stress cycles”, it becomes subject to fatigue failure. Normal metal fatigue occurs when the bearing load-carrying components reach the end of their useful fatigue life. This can be the result of premature fatigue induced by loading the bearing beyond design limits (286-290k loads on class F bearings), but typically occurs as normal result of “old age” of the bearing components after many millions of revolutions under design load (Technical Forum 94-1). These fatigue stresses cannot be readily duplicated in lab testing so the effects of these must be deduced from service life data.
HOT BOX DATA
North American Railroads experience an average of more than three thousand hot bearing removals per year. Each of these removals requires at least one train stoppage and many require more than one. The ultimate cost of such delays can be staggering and railroads are devoting considerable resources to the elimination of these undesirable stoppages.
Figure 1 shows data for two years of hot bearings due to spalling (either water, impact or age induced) gathered from service inspections in North America. It depicts all manufacturers inspected and original equipment manufactured as well as reconditioned bearings and the number of times each was reconditioned. Original equipment manufactured (OEM) bearings make up over 42% of the population. These bearings averaged over 14 years service before spalling caused a service failure. Bearings that had been reconditioned 1 time (Recond. 1x) made up 38% of the population and failed with an average service life of only 9.9 years.
The number of bearings surviving each successive reconditioning becomes progressively smaller. After the fourth reconditioning cycle, failed bearings averaged 4 years service life. This is not economically realistic.
Figure 2 represents the 1999 hot bearing setouts for one Railroad in the U.S. for all reasons. This shows comparative service life of failed journal bearings for several different types of railcars. It is significant that for all of the car types none of the reconditioned bearings achieved a longer service life than the OEM bearings on the same car type.
It is commonly understood that box and tank cars accrue fewer miles per year than any other cars in the rail fleet. On the average, OEM bearings that caused a hot bearing set out on these two types of cars ran for over twenty years, while reconditioned bearings ran only about half as long.
The higher mileage cars at the bottom of the chart show decreasing bearing life. Many flat cars, hoppers, and gondolas will travel well over 100,000 miles per year, carrying a gross load of up to 286,000 lbs., and experiencing these loads over fifty percent of their service time. Bearings suffer metal fatigue at an earlier age under these conditions. Reconditioned bearings offer only about 50 to 60% of the life of the OEM bearings in these applications. Is an “often reconditioned” bearing an acceptable risk on a high mileage car? Is it an acceptable risk on any car?
RECONDITIONED BEARINGS
Bearings are removed at various stages of their useful bearing life. Many wheelsets are removed prematurely for wheel shelling; others are removed because the wheels have simply worn out. As a result, the bearings that come out of service and go into the reconditioning process are highly varied. Some are old, some new, some with recent repairs, and still others that have operated well for twenty years and never been removed from the wheel set.
The inspection results of nearly 50,000 bearing cups inspected in a Quality Bearing Service (QBS) Reconditioning Facility in the year 2000 are shown in Figure 3. Components returned from normal service are categorized as: Premium Cups (no defects or blemishes); AAR Good Cups (acceptable or repaired defects {spalls, water etch, brinells, indentations}); and Regrind Cups (shallow defects in the raceways such as etch or corrosion, indentations, impact brinells, or small pin-head spalls). Regrind cups are reground and reused later as a remanufactured component.
About one third of these bearings are OEM and the average length of service is about 14 years. Another third are Reconditioned bearings that have been reconditioned one time and average 19 years of age. If we assume the initial service life was 14 years, the second service life for the once reconditioned bearings is only about five years. The last third of these bearings have been reconditioned two or more times. Their serviceability can be expected to be even less.
Nearly 80 percent of the OEM bearing cup scrap had at least 20 years of service, as did more than 68 % of the once reconditioned cup scrap. Two, three and four times reconditioned cups had scrap rates of 80.5%, 88.2% and 93.8% (respectively) at the twenty-year plus mark. Twenty year old cups have a severely high scrap rate and shorter remaining service life due to defects and fatigue stress accumulated in prolonged or repeated service applications.
Figure 4 shows the rejection and scrap percentages for cone assemblies in the reconditioning process by date of manufacture. Once again, reject and scrap rates for cone assemblies older than twenty years is markedly higher than those of more recent manufacture.
Nearly half of all cones are rejected for defects during reconditioning. About one fourth of those are true scrap and about three fourths are candidates for remanufacture. Remanufacture means “to restore to like-new condition”, and these are anything but “like new.”
The cone data reveals the same results as the cup data. Twenty-year old cone assemblies suffer very high scrap and reject rates as they pass through the various reconditioning cycles.
CAR REPAIR BILLING DATA
1999 car repair billing (CRB) data indicates about 3,000 roller bearings removed as overheated or fused. Ten years earlier, the same removals totaled over 5,800 roller bearings. This decrease is due, at least in part, to changes made to bearing reconditioning standards by the AAR. Those changes have helped improved the performance of the roller bearing population measurably. Cone bore dimensions have been tightened and o’rings have been removed from the applications to promote greater bearing retention. The number and size of allowable and repaired defects has been reduced and wear on the cone assembly back face has been limited.
1999 CRB data shows over 500,000 bearings applied to wheel sets in North America in that year. Only 6.29 percent of these, slightly over 32,000 bearings, were new bearings going into service as replacement bearings. Bearing usage per billion ton miles was 545 in 1992 and has fallen to 396 as of 1999. We are running bearings longer and increasing their stress loading.
CONCLUSIONS
North American Railroads are experiencing heavier loads and more intensive car utilization than at any time in their history. Railroads and Regulators are looking for greater serviceability, and rail customers are demanding it. New expectations have come into play and they cannot be met by employing old standards.
The data presented here dictates removal of bearings and components after 20 years of age. The data also points to the necessity of removing bearing components that have been in service through two or more reconditioning cycles. The service life to be realized by reinstating these components is not at all realistic or acceptable to the industry and they should not be returned to service.
Bearing cups over 20 years of age comprise over half of the scrap in each reconditioning category. Twenty years is a significant hurdle in bearing cup serviceable life and cups should not be returned to service after that time.
The fact that cone assemblies are rejected and then repaired in various ways (bore plating, regrind, re-cage,) makes it highly likely that fatigue stressed races are often reused with restored bore dimension and new or replacement rollers. Limiting the age of cone assemblies to twenty years would insure greater serviceability when they are returned to service. In-service cone failures can be catastrophic and the possibility of their occurring must be reduced.
Since 1989 the number of hot box removals has dropped virtually in half. Is it possible to cut that rate in half again in the next ten years? The steps taken in the past ten years have made a significant impact on performance. Is another significant impact possible? We’ve entered a new millennium, and it requires a new way of thinking. Greater serviceability and better risk management is a necessity!
Older and often reconditioned components should be eliminated from the reconditioned bearing! The new standard should be two recondition cycles and twenty years of age.
The Technical Forum is an information resource for the rail industry and is provided as a courtesy of Amsted Rail Group. Suggestions, inquiries or comments are welcomed and should be directed to:
Editor, Technical Forum
BRENCO, Incorporated
P.O. Box 389
Petersburg, Virginia 23804
804-863-1713
Additional copies provided upon request.
BACK TO TECH SHEETS
|
