Cs is the wheel cost, cs, divided by the number of parts produced per wheel, Nw The number of parts per wheel is given by
(19.31)
 |
 |
where dsmax is max wheel diameter, dsmin is min wheel diameter, rs is radial grinding wear per dress, ad is dressing depth, nd is number of dressing passes, and Nd is the number of parts per dress. The wheel cost per part is, therefore,
Figure 19.25 shows typical wheel cost/part varying with number of parts/dress for alumina and CBN wheels. Redress life has a strong effect using CBN due to initial cost but cost/part may be offset by long redress life. A high wheel speed increases redress life by reducing chip thickness. Wheel cost/part becomes negligible with long redress life.
Initial wheel costs in centerless grinding are higher than in grinding between centers due to the larger diameter and width of centerless wheels. However, redress life is correspondingly increased so that wheel costs are spread over many more workpieces. The wheel costs in Figure 19.25 could be more than doubled in centerless grinding although the wheel cost per part is likely to be reduced.
19.10.4 Labor Cost/Part
 |
 |
Labor cost per part, C, is the product of labor rate, c, and total cycle time, 1t. Labor rate includes general overhead costs other than the cost of the grinding machine. The total cycle time is given by
|
TABLE 19.2 An L827 Experimental Plan Level
|
 |
|||
 |
 |
 |
 |
The total cost/part for wheels, labor, and machine is, therefore,
The number of parts/dress is a factor in all three costs contributing to the total cost. This explains the importance of redress life. Inclusion of machine cost and redress life allows realistic evaluation of high-speed grinding and application of superabrasives.
19.10.7 Experiment Design
To evaluate costs, it is first necessary to determine the effect of grinding conditions on redress life. The effect of long redress life was studied grinding AISI 52100 and the effect of low redress life with Inconel 718. Redress life was defined as the number of parts ground before either roughness of 0.25 pm Ra or roundness error of 1 pm was exceeded. By this means, the specified quality levels are built in to the test procedure. The following procedure allows a comprehensive analysis to be performed with a minimum number of grinding and measuring trials.
An example of an experimental design for a two-level investigation of seven variables using eight trials is given in Table 19.2. In the example shown, four trials are conducted for each value of a parameter. The result for each parameter is the mean value for the combinations of the other parameters shown. In each case, Level 1 represents the low value of a parameter and Level 2 represents the high value. A large number of workpieces are ground for each trial, thus allowing redress life and G-ratio to be assessed.
The investigation was conducted in three stages.