Wear testing is clearly a complex, perhaps bewildering, field. Hasty or inexperienced engineers will almost certainly draw improper conclusions, perform tests which don't test the properties they are interested in, or compare tests which do not test the same properties of a material. In reviewing briefly this complexity in the testing and interpretation of wear properties, our purpose was not to overwhelm the reader, but to emphasize the need for caution and informed discussion of precisely how to test the materials. The same approach is necessary when comparing literature data on materials which seem the same as the reader's but are usually not characterized sufficiently to interpret how similar they are. Ceramic materials and complex, multicomponent coatings are prime examples of this problem. There seems to be no substitute at present for simply doing the proper test(s) on materials of interest.

Having made this point, how does one proceed to decide what test, specimen configuration, test conditions, etc. are needed to develop the data necessary to answer the design or performance question at hand? If the end item will be a bulk material or a thick coating (e.g. weld overlay, thick diffusion layer, etc.) there are probably no restrictions on test type. One must simply try to estimate the level of stress to be encountered, decide whether an abrasive will be present, intentionally or unintentionally—as with an oxide or corrosion layer being removed from the surface, of the mating material(s) and decide whether the contact is rolling, sliding, or free particle erosion. These considerations indicate what type(s) of testing must be done. Of course, if a thin coating is involved, choices will be limited and tests more difficult to perform.

After one identifies the type of wear behavior to be modeled, choosing the best or most appropriate test is more difficult because specific parameters for the test environment must be chosen: temperature, particle size and distribution, eroding or abrading medium, velocity of particles or of specimens. It is possible that one or more of the candidate tests cannot reproduce the environment needed. The test environment should reproduce the actual or expected environment as closely as possible. For example, the desire should be resisted to "accelerate" an erosion test by substituting aluminum oxide for silica. Performance of materials changes drastically and erroneous conclusions will be drawn as was shown above for the example of aluminum oxide vs. silica erosion of uncoated and diffusion coated carbides.

Finally, try to characterize the specimen and test parameters as completely as possible. Some important parameters to determine are forces on components being worn, velocities of components and streams of particles, the metal 1urgy/chemi stry/morphology of particles or surfaces producing the wear and of surfaces being worn. The temperature and other physical properties of the environment (liquid or gaseous) are important, especially if electrical effects and/or chemical corrosion effects are present. This investment will be justified when complex and confusing data must be interpreted and different tests or materials must be compared. There is no alternative to this level of characterization for the individual or company who wants to solve wear problems and build a "corporate memory" in this area over a period of years.

0 0

Post a comment