Putting Plastic Bearings to the Test

Over the past 20 years, plastic bearings have come a long way from their dime-a-dozen, injection-molded nylon ancestors. Today’s bearings, often made from engineered plastic compounds, can meet a wide range of speed, load, temperature and chemical-resistance requirements. Compared to metal bearings, engineered polymer bearings can also reduce cost, weight and vibration. If self-lubricated, modern bearings can also eliminate some messy maintenance chores.

To reap all of these benefits, however, you have to pick the right polymer bearing for the job by matching the bearing’s predicted service life to the application’s actual service life. This is ensured by using realistic methods to characterize the tribological properties of polymer bearings. In many cases, the tribological data gained from the standard pin-on-disc and block-on-ring tests cannot be accurately applied to plastic plain bearing applications.

Rather than relying on standard tests, igus instead calculates the lifetime of a polymer plain bearing using bearing-on-rotary-shaft tests (see Figure 1). Based on years of testing, we have compiled a database of friction and wear values for various bearings running against a variety of shafts. We measure factors like load, sliding speed, type of motion and temperature when generating the data. Here’s a look at why this rotary shaft data is so important when it comes to making accurate predictions of bearing performance: Standard Test Problems

The block-on-ring and pin-on-disc (see Figure 2) tests are widely used by researchers as reference tests for applications involving rotary motion. As opposed to plain-bearing-on-rotary-shaft tests, however, they are very limited in their capability to represent thermal conditions for radial bearings. This is demonstrated in the ratio between the contact surfaces of the tribological partners e. e is defined in Equation 1 as the quotient between the tribological partners’ contact surfaces. For radial plain bearings, the projected area of plain bearing A_RGL and the area A_W swept around the shaft always yields a ratio e_RGL=0.32 as shown in Equation (2).

In the case of washer disks, both tribological partners have full surface contact — that is, the ratio e_AGL=1. In reference tests involving a block-on-ring or pin-on-disc, the ratio depends heavily on the test configuration. The block or pin frequently has a contact area between 4 × 4 mm² and 10 × 6 mm². The track’s circumference depends on the employed test rig and metallic counterparts. The track diameter typically lies between 30 and 120 mm. The ratio is shown in Equation 3. The block-on-ring contact area is assumed to be approximately flat, while the pin-on-disc contact area is assumed to be approximately rectangular.

Table 1 (see page 68) lists engagement ratios yielded by various tests, together with associated influencing variables.

The ratio e in the plain bearing applications is distinctly higher than in the reference tests. For a given load generated by sliding speed and surface pressure, this causes the plain bearings to generate much more heat as compared to the reference tests. The engagement ratio is independent of the tribological partners’ geometries in the radial bearing/shaft tests as well as in the rotary bearing applications. By contrast, the engagement ratios in the block-on-ring and pin-on-disc tests depend notably on the employed test rig. This can even restrict the mutual comparability of the reference tests, if they are conducted on test benches of varying dimensions.

Friction Values

Another important variable describing tribological systems is the coefficient of friction µ. Equation (4) defines this coefficient as the ratio between the frictional force F_R and normal force F_N. The frictional force vector always opposes the relative motion. The friction values of polymer bearings running against metallic counterparts were investigated using the various measuring techniques described above. The measurement results are summarized in Figure 3; the test conditions are listed in Table 2.

The test parameters comprising a load p = 0.7 MPa and sliding speed v = 0.15 m/s have been selected by igus as standard conditions for measuring friction coefficients on plain bearing/shaft combinations. Measurements with a load p = 1 MPa and sliding speed v = 1 m/s were performed to facilitate a comparison of the friction coefficients with the results of the block-on-ring measurements. This collective load exceeds the permissible p-v values of iglide plain bearings in some applications. It was nevertheless possible to determine friction coefficients for iglide G, H, J and L280 plain bearings, although significant wear was found in some cases. No friction coefficients could be determined for the iglide GLW plain bearings.

One trend exhibited by all bearing-shaft tests showed that the friction coefficients at a p-v value of 1 MPa*m/s are higher than those at a p-v value of 0.1 MPa*m/s. The friction coefficients determined with the block-on-ring method at a p-v value of 1 MPa*m/s exhibit notable deviations with respect to the friction coefficients at the same p-v value in the bearing/shaft tests. The friction coefficients of the iglide J and L280 plain bearings operating against their respective counterparts differ by 30-50 percent. Their trends also oppose each other in some ranges. The coefficient of friction of the iglide H plain bearings in the block-on-ring test is 20 percent lower as compared to the bearing/shaft test. By contrast, the coefficient of friction of the iglide J plain bearings is 50 percent higher in the block-on-ring test compared with the bearing/shaft test.

The coefficients of friction in the pin-on-disc tests are more than twice those in the bearing/shaft tests. In other words, none of the reference methods supplies comparable results for radial plain bearings.

Loads

Besides the selected material, the collective load generated by pressure and sliding speed is another primary factor influencing the properties of a tribological system. This is represented in Figure 4. The p-v value is the product of the surface pressure (p) exerted by the normal force and the sliding speed (v) of the tribological partners’ contact surfaces. Specific to individual tribological systems, the p-v diagrams allow estimates of these systems’ application limits at particular engagement ratios. If the collective load passes outside the linear p-v range, the wear rate starts to increase dramatically. At the static load limit one or more tribological partners reach their deformation limit; exceeding the critical sliding speed results in thermal failure of one or more tribological partners.

Whereas the static load limit is influenced only moderately by the ratio e, the dynamic load limit is related very closely to this ratio. At least one of the tribological partners experiences thermal overload near the dynamic limit. The ability to dissipate frictional heat depends strongly on the ratio e. The higher this value, the lower the proportion of the metallic tribological partner’s surface area capable of releasing heat. Consequently, an engagement ratio e_RGL=0.32 for radial plain bearings permits just 68 percent of the shaft’s surface to release heat. During a block-on-ring test with eB=0.03, however, 97 percent of the metallic counterpart’s surface is able to release heat.

Metal’s Important Role

Another important difference between the reference tests and bearing-shaft tests are the metallic counterparts employed in the tests. The block-on-ring tests frequently make use of needle-bearing inner rings made of 100Cr6 (USA AISI 52100 = 100Cr6). However, this material is used primarily for roller bearings, hardly ever for shafts running against plain bearings. At the same time, the friction coefficients in the bearing/shaft tests differ notably in some cases, depending on the employed shafts.

The coefficients of friction of iglide plain bearings running against hard anodized aluminum shafts tend to be lower as compared to hard chromium-plated CF53 (USA AISI 1050 = CF53) shafts as counterparts. This tendency is reversed in the case of iglide GLW plain bearings. The corresponding friction values are shown in Figure 5. The shafts differ in terms of heat conductivity as well as surface texture. Accordingly, the friction values are not always predictable, not even by the same measuring technique. In igus’ tests, the coefficients of friction of iglide plain bearings are therefore determined against a variety of shafts.

Another example of the variation of coefficient of friction within a single measurement technique is shown in Figure 6, which represents the coefficients of friction of iglide P and iglide R plain bearings against CF53 shafts as a function of the surface pressure. The test conditions are summarized in Table 3. The coefficient of friction drops notably as the load increases.

To obtain results which are meaningful in the real world, it is extremely important for test conditions to agree as closely as possible with actual operating conditions.