impact the strain prediction using proposed technique. However, the solution starts to degrade with higher levels of noise, especially for high strain regions as seen from Figure 12b. Figure 11: Overlay of strain response of box-beam at (a) tip and (b) interior point; comparison made between reference solution and predicted solution with 5% noise. Figure 12: Overlay of strain response of box-beam at (a) tip and (b) interior point; comparison made between reference solution and predicted solution with 10% noise. CASE 3 There were 14 adof simulated measurements used for expansion up to 434 nodes for all the cases discussed in this paper. In case 3, 10 redundant sensor measurements are simulated in addition to the 14 adof simulated measurements to study the effect of greater number of sensors on the strain prediction. A total of 24 sensors are used for generating simulated adof measurements. The 14 and 24 sensor set-ups are shown in Figure 13. Notice that the 24 sensor set-up has sensors located somewhat redundantly at locations that are considered to be critically important locations for expansion. Redundant 197
RkJQdWJsaXNoZXIy MTMzNzEzMQ==