This preview shows page 1. Sign up to view the full content.
Unformatted text preview: elements with the same parametric area. The point chosen for each element is randomly selected within the element. Figure 3 shows samples and the evaluated minimum deviation zone, which is 0.06766 mm. There is, however, no guarantee that the extreme deviation in these samples is the same extreme deviation in simulated geometry. Figure 4 shows the estimated density function using the Parzen windows method. The evaluated deviation by least square method is 0.0912 mm, which confirms the idea of overestimation of fitting by least square method. Figure 3; Inspection based on the 16 stratified samples. Figure 4; Density function estimated based on 16 stratified samples. 4.2. Iterative sampling Although 154 points are captured during the automated sampling process, only 24 points are selected as the critical points to evaluate the minimum deviation zone. Figure Evaluation of Geometric Deviations in Sculptured Surfaces 143 5 shows how the density function changes by iterative sampling. It can be seen that gradually over each iteration, the shape of the density function is extended, it takes a more symmetric shape and the density of samples within the average deviations...
View Full Document
- Spring '10
- The Land