196 P. M. Vinze et al. 0 5 10 Analytical mode number 15 20 25 5 10 15 20 Experimental mode number Fig. 18.8 Cross-MAC between two experimental results for 2014 Circular Plate Dataset Another observation was that the frequency comparison for these values was such that the frequencies were far apart. Also, two sets (of two) of modes that had a very high spaceMAC sometimes gave a spaceMAC of about 0.6 when group as a quadruple. For example, the mode at 761.104 and 764.163 Hz had a 0.924 spaceMAC between them and both these modes showed a very low spaceMAC with the modes at 1223.011 and 1224.080 Hz while they showed a spaceMAC value of around 0.6 for the modes at 1328.055 and 1328.769 Hz. This can be clarified if the actual modeshape is observed (animated). Figures 18.9, 18.10, 18.11, 18.12, 18.13 and 18.14 show the modeshapes at these six frequencies. As can be seen from these figures, the spaceMAC values around 0.6 seems to be because of spatial aliasing phenomenon. But clearly the spatial aliasing was not strong enough to affect the cross-MAC values. SpaceMAC was observed to have higher sensitivity to spatial aliasing effects. Following an increasing order of selected multiplicity is an important factor in spaceMAC implementation. SpaceMAC calculations for multiplicity 3 were done. Some sample values are show in Table 18.3 below. As is already known this set of modal vectors does not contain any repeated modes of multiplicity 3. But some high values can be seen. Sets of 3 modes with spaceMAC3 values of around 0.6 usually result from spatial aliasing of a pair of repeated mode with another mode. Similar effects can be observed when calculating spaceMAC4. The MAC threshold was varied from 0.7 to 0.95 but the spaceMAC values remained unchanged. If MAC value threshold is set too low it could eliminate some repeated mode pairs so it is important to set the threshold depending on which mode have to be checked.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==