18 Developing a Correlation Criterion (SpaceMAC) for Repeated and Pseudo-repeated Modes 195 Extract modes from both analysis sets Are there any rigid body modes present in the two sets Geometry Mapping - FEM to Experimental Downsizing modal vectors Remove vectors with high cross MAC Create all combina ons of remaining modes Calculate space MAC Remove the rigid body modes Fig. 18.7 Flowchart for implementing spaceMAC 18.3.3 Results When two different sets of experimental results were compared for both plates, one important observation was that the repeated modal vectors were numerically consistent to begin with. The cross-MAC is shown in Fig. 18.8. In such cases spaceMAC will not be required. One factor that might be affecting this is the real dominant normalization procedure that is applied to modal vector during residue estimation. SpaceMAC would still be useful for FEM to experimental comparison. The spaceMAC method correlated one modal vector space of a selected multiplicity and another modal vector space of the same multiplicity. This is important as the drawback of the correlation method described in Sect. 18.3.1 is that it correlates a space with a vector. This gives space-MAC an advantage over this method. For example, if the 12th, 13th and 14th modes for 2001 Circular Plate Dataset are taken as in Sect. 18.3.1 it is easier to detect and correlate repeated modes with spaceMAC. But it has to be made sure that the search for repeated modes must be made in an increasing order of multiplicity of repeated modes. Table 18.3 shows the space-MAC values of multiplicity equal to 2. The cross-MAC threshold was selected as 0.9. The cross-MAC values can be seen in Fig. 18.9 for 2001 Circular Plate Dataset. Space MAC values above 0.8 can be considered a good enough correlation. It can be seen in the Table 18.2 that there are some space-MAC values around 0.6.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==