ComputerAided_Design_Engineering_amp_Manufactur.pdf
ComputerAided_Design_Engineering_amp_Manufactur.pdf
ComputerAided_Design_Engineering_amp_Manufactur.pdf
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
The calculated minimum form error is the profile tolerance which is unique to the sculptured surface.<br />
The optimal transformation matrix giving the minimum form error then becomes the optimal setup.<br />
In this section, an iterative algorithm, MINIMAXSURF 11 , is explained using the Chebyshev norm<br />
between the measured points and the corresponding closest points. Again, let P i be the measured points<br />
data, Q i � R i be the calculated closest points on the offset surface defined by the CAD geometry, and T be<br />
the transformation matrix between the two data. The Chebyshev norm of power p, L p, can be defined as<br />
(4.30)<br />
The L p can be minimized with respect to the transformation matrix T and the optimum transformation<br />
matrix T 0 can be found. The profile tolerance E is then<br />
The algorithm for the profile tolerance evaluation is implemented as<br />
(4.31)<br />
1. Set the iteration index k as 0, and the initial transformation matrix as the multiplication of<br />
T1 and T2 in case of the rough-phase and fine-phase alignment.<br />
2. Input the measured point data Pi for i � 1, 2, …, n, where n is the total number of measured<br />
points on the surface.<br />
3. Assign the initial closest point corresponding to the Pi as the measurement target<br />
point.<br />
4. Calculate the initial profile error for i � 1, …, n.<br />
5. Increase the iteration index k by 1.<br />
6. Find the closest point corresponding to the transformed<br />
7. Calculate the kth iterative transformation matrix, such that the Chebyshev norm,<br />
be minimum using the universal alignment algorithm. 11<br />
( Qi � Ri) 8. Calculate the kth iterative profile error, for i � 1, 2, …, n.<br />
9. Evaluate the incremental maximum deviation, .<br />
10. If the incremental maximum deviation is less than the assigned tolerance TOL then proceed to<br />
the end; otherwise increase k by 1.<br />
11. Repeat steps 5 through 10 until the criterion is met.<br />
0 ( )<br />
Eo max | T 0 ( )<br />
[ ] 1 � Pi ( Qi � Ri) 0 ( ) �<br />
�<br />
|<br />
( Qi � Ri) k ( )<br />
T k�1 ( )<br />
[ ] 1 � Pi T k ( ) , �| T k ( )<br />
[ ] 1 �<br />
[<br />
Pi ( Qi � Ri) k ( )<br />
� | p ] 1�p<br />
E k ( ) � max | T k ( )<br />
[ ] 1 � Pi ( Qi � Ri) k ( ) �<br />
|<br />
DEk |E k ( )<br />
E k�1 ( ) �<br />
� |<br />
The flowchart of the implemented algorithm is shown in Figure 4.10.<br />
Error Evaluation for Basic Features<br />
L p<br />
�<br />
�<br />
|T 1 � Pi � Qi � Ri �1 Pi Qi Ri ( )|<br />
E � max |T0 � ( � ) | for i � 1, 2… , , N<br />
Flatness is the deviation of measured points with respect to the mathematically determined reference<br />
plane; the flatness error is evaluated as the distance between the maximum and minimum deviation.<br />
Squareness tolerance is the tolerance zone limited by two perpendicular planes to the given reference<br />
plane, where the two perpendicular planes contain the whole measured feature. The reference plane can<br />
be determined as the best fit plane by using the least squares technique. Parallelism tolerance is defined<br />
as the tolerance zone limited by two parallel planes which are parallel to the reference plane; the reference<br />
plane can be constructed as the best fit plane by using the least squares technique. The straightness<br />
tolerance is defined by the minimum distance between two parallel lines containing the measured datum.<br />
The slope of two parallel lines can be mathematically determined as the best fit line by using the least<br />
squares technique. The roundness tolerance is defined as the tolerance zone limited by two concentric<br />
circles; the circles can be calculated by the least squares circle (LSC), the minimum zone circle (MZC),<br />
the minimum circumscribed circle (MCC), and the maximum inscribed circle (MIC).<br />
p<br />
1�p<br />
T 0 ( )