Problems And Solutions

Every wear test, whether for bulk materials or coatings can be complicated by problems with equipment, procedures, specimen preparation, inconsistencies in abrasive media, and in interpreting the test results. Rather than enumerating all the known problems with each wear test described above, several specific examples of some general problems in testing of coatings will be described along with some solutions.

Penetration of Coatings

Thick coatings, such as those produced by thermal spray techniques and weld overlay hardfacings, seldom experience penetration during wear tests. However, thin coatings applied by CVD, PVD, and r.f.v. sputtering techniques, may be penetrated if care is not taken during the test procedure. This problem also exists for materials that are treated by such surface-modification processes as ion-implantation, diffusion alloying and ion-nitriding. Special care must be taken to use less than normal stresses and shorter stresses and shorter test durations when testing thin coatings and shallow surface treated materials. A recently formed subcommittee of the ASTM G2 Committee on Wear and Erosion has begun studying standard wear test procedures for coatings. Preliminary progress from this committee indicates that many existing standard test procedures, with minor modifications can be used for coatings.

A modification to the jet impingement erosion test for thin coatings is being studied by Wear Technology Inc. Rather than producing a deep crater-like wear scar on a stationary specimen by the impinging gas driven abrasive particles, a linear motion device moves the specimen under the jet. The resulting wear scar will resemble a ditch rather than a crater. The linear speed of the specimen can be adjusted to allow material removal in the coating without penetration of the coating.

Wear Scar Problems—What to Measure

All wear tests have one thing in common. They are designed to produce material removal in a prescribed and controlled manner. However, not all types of wear tests remove material from a specimen in the same way, that is, by the same mechanism. The removed material must be characterized in some way to provide quantitative analysis of the particular wear process or wear rate. Several types of measurements are available to quantify the amount of material removed in a wear scar. Probably the simplest measurement is weight-loss. The difference in the specimen weight measured before and after a wear test is a common means to quantify wear. This method is successful if all materials being considered in a study are very nearly alike; i.e., they are homogeneous, and their densities and compositions are similar. Another measurement that is frequently used is volume-loss. Generally, the volume-loss is calculated rather than measured. The weight-loss divided by the sample density, if it is known, produces the volume-loss value. Obviously, this method only works if the density is known and if the density is homogeneous throughout the coating. Many coatings, however, are made up of layers of differing material with different compositions and densities. Other types of coatings have a composition (and density) gradient from the outer surface toward the base material. Examples of these include diffusion alloyed surfaces, variable composition plasma coatings, and ion-implanted surfaces. In all the variable density situations, it would be incorrect to calculate volume-loss by dividing the weight-loss by the density. Some researchers get around this problem by measuring the maximum depth of the wear scar. In some cases this method produces reliable wear data. But if a coating contains several layers or phases, or a continuous density gradient none of the above methods is exact. George and Radcliffe (14) of the U.K. have developed a method and equipment for directly measuring the volume of a wear scar. The equipment consists of a RTH Talysurf 4 profilometer modified to both scan and traverse a wear scar, a computer and interface to control the motion of the profilometer, and appropriate software to operate the motion control components and log the resulting data. Isometric plotting of the wear scar, and computed scar volume are produced automatically. At present, automated profilometry for wear scar volumes is expensive, but as more of this type of equipment becomes commercialized, it will become more economical.

One particular wear scar problem frequently found in jet impingement erosion testing can be seen graphically in Figure 15, a profile of an erosion crater. The crater was produced with an impingement angle of 90 degrees;that is, the direction of travel of the particles was 90 degrees from the surface plane of the specimen. If the wear scar (crater) is allowed to become large and deep, as depicted in the profile, the impingement angle at impact for all particles is no longer 90 degrees. A

small fraction of the particles will strike the very bottom of the crater at 90 degrees, but it can be seen that particles striking elsewhere in the crater will have impingement angles less than 90 degrees. Also, particles impinging into a deep crater tend to interact with each other more. Their energy is reduced by collisions with each other, thereby reducing the particle energy transmitted to the specimen. Wear rate is thus slowed. Therefore, it is recommended that erosion tests be designed to avoid large deep craters.

Density and Composition Gradients

Problems associated with composition and density gradients were mentioned earlier. This type of problem requires further discussion. In order to provide better understanding of the problem, specific examples will be examined. The subjects of discussion are two samples of Co-WC materials diffusion alloyed with boron in one case, and boron with titanium in the other case. The boron system, designated 601, developed a layer of mixed borides on the surface of the Co-WC material. The boron/titanium system, called 745B, developed two layers; a layer of titanium diboride on the outer surface, a layer of mixed borides and other intermetal1ic compounds sandwiched between the titanium diboride and the Co-WC substrate. In both cases, the wear resistance of the cobalt-tungsten carbide was significantly increased. When erosion tests were conducted on these materials, it was not clear how to properly calculate erosion loss. The bar graph in Figure 16 shows how different the results can be when different methods are used to calculate the loss. The amount of material loss on the treated and untreated specimens was calculated by weighing the specimens before and after the test. The weight loss, normalized in the bar graph, is shown in the pair of bars on the left side of the figure. The center pair of bars shows the volume loss data when the same weight losses as before were converted by dividing by a density value for the bulk Co-WC material. The relative loss ratios for the treated and untreated materials are very nearly the same. However, it may not be appropriate to use a bulk density factor for volume calculations. It was found by x-ray diffraction examination that the surface material on the titanium/boron diffusion treated Co-WC specimen was titanium diboride. If one calculates the volume loss based on the density of titanium diboride, the results would appear as in the pair of bars on the right of the figure. The ratio of loss in this case is much different than in the other two cases. Deciding which way to treat the data is not straightforward. But the point to be made is that for a complex, multilayered system, or one that has a continuous density gradient, great care must be taken in performing the wear test, analyzing the wear scars and the data, and in choosing the proper loss calculation method.

Perhaps an appropriate way to study the wear resistance of a complex coating system is to purposely measure the wear in stages, being careful to monitor the wear rate as a function of depth into the coating. An example, again using the diffusion alloyed Co-WC material, is shown in Figure 17. In this case, the alumina slurry high stress abrasion test (ASTM-B611) was used to abrade through the surface layers of the diffusion treated Co-WC. The curve on the left of the figure shows the wear rate of untreated Co-WC (scar depth as a function of revolutions). The curve is linear, as one would expect, since the material is homogeneous, and has a constant density with depth into the sample. The center curve represents scar depth vs. revolutions data for a specimen of Co-WC diffusion treated with titanium and boron. Three different wear rates are seen. Starting from the origin, the first section of the curve represents the wear rate of the outermost layer (titanium diboride) of the alloyed material. It is clearly seen to be more wear resistant than the untreated material. The second section of the curve represents the wear rate of the innermost layer of diffusion treated material. It too is more wear resistant than untreated Co-WC. The third section of the curve shows the wear rate of the parent material (untreated Co-WC) where the diffusion layers have been penetrated by the test. The slopes of both curves are seen to be the same in the untreated material. Similar information can be taken from the curve on the right; wear data for Co-WC diffusion treated with boron. In this case, the first portion of the curve shows the wear rate of the diffusion zone. The second portion of the curve again shows the wear rate of the untreated material where the coating was penetrated. It is suggested that the interrupted wear, test on complex coatings and surface modifications can be very useful for accurately understanding their true wear behavior. One could predict expected wear life of the coatings and even the individual layers using this method. And the question of what density to use for volume loss calculations is eliminated.

Ranking and Test Parameter Confusion

The rank order of a group of materials for wear resistance is frequently desired for purposes of selection of materials for application. Several problems have been identified that add to the confusion of why rank order differs from one test to another, from one laboratory to another, and even from one day to another in a single laboratory. Some of these problems are discussed here.

Several wear tests require using abrasives for producing material loss at a controlled rate. Some test, such as the pin-on-drum test, use an abrasive coated paper. Others, such as the dry sand rubber wheel test, the jet impingement erosion test, and the alumina slurry test require a loose abrasive grain. Care must be taken to assure the use of a consistent abrasive in all like tests. A slight difference in the particle size, the particle shape, and even the microscopic surface texture of the particles can cause test results to vary. It is recommended that a sample of each lot or batch of abrasive be saved and catalogued for future examination. Also, characterization of each lot is wise. Most wear technicians do calibration tests frequently to assure repeatable results. These tests should be made on a carefully chosen standard material and with each lot of abrasive. Any inconsistencies should prompt an investigation into the cause before testing is continued.

Another cause of ranking confusion is attempted comparison of wear tests which measure different wear mechanisms. The wear resistance ranking of a set of materials is often different when test results from one type of wear test are compared to those of another type of test. It is clearly important to select a wear test that attempts to simulate conditions which the material will see in the real world. Ranking will also be different if parameters for the same test are varied. An example of this is shown in Figures 18 and 19. The data shown is for jet impingement erosion tests on untreated and diffusion alloyed Co-WC (the same materials discussed earlier). Erosion results shown in Figure 18 were produced with an angular .lumina abrasive. In this case, a large difference is seen for wear at 30 and 90 degree impingement angles in the untreated Co-WC. But the diffusion alloyed specimens show nearly the same erosion behavior at 30 and 90 degrees. Also, it is noted that the boron diffusion system and the boron/titanium system are nearly identical in erosion resistance. Both are, however, much more erosion resistant than the untreated material. An entirely different story is clearly seen in Figure 19. The same Co-WC materials were tested with rounded silica abrasive in this test series. The erosion resistance of the untreated material is nearly the same for impingement angles of 30 and 90 degrees.

But a significant difference in the erosion at the two angles is seen for both types of diffusion treatments. Also, the 90 degree erosion results of the two diffused materials are not very much different than that for the untreated material. From these results, it can be seen that it is very important to examine materials at several test conditions, and with more than one type of abrasive. Only then can realistic predictions of material behavior be made that will permit proper choices for engineering application.

0 0

Post a comment