84 S. Johnson et al. two control degrees of freedom at a time. Rohe et al. used an aggregate technique that adds all controlled autopower spectral density (ASD) responses and compares them to the sum of all control specifications [2]. Adding ASDs together results in much less data to compare but loses the information about how each channel responds. Also, there is no exploration of how the loss in resolution might affect the overall assessment of the test. There are many difficulties with multi-axis testing, from implementation to assessment. This study aims to explore how changing the control location on a test article changes the response at monitored locations on the test article while utilizing a multi-axis test approach. In addition, this study will focus on data presentation can affect a test engineer’s ability to quickly and accurately determine if a test was successful. We will assess the GRMS, dB error (edB), and the frequency response function similarity metric (FRFSM) on aggregate response as well as individual response to determine what ”level” of fidelity is needed in analysis to assess the global behavior of the system compared to what it sees in the field. Testing Field test In this study, the Generic OBject for Laboratory Environmental Testing (GOBLET) is used as the test article. The GOBLET is an aluminum, three-piece assembly that is used to assess complex dynamics of cylindrical components. The GOBLET’s three parts are an adapter plate, a cylindrical mass, and a bell, which are shown in Figure 1. The bell is thin-walled and is relatively compliant, especially when attached to the more rigid cylinder. The bell bolts onto the cylinder at 3 legs. The GOBLET adapter plate and cylinder primarily exhibit rigid body motion, with their axial motion being nearly identical to the input at the base of the assembly. Fig. 1 Configuration of GOBLET during field testing. To have a set of reference data, we field tested the GOBLET in a transportation environment. To perform this testing, six tri-axial accelerometers and three uni-axial accelerometers were attached across the GOBLET assembly. The bell piece had four tri-axial accelerometers, with two on one of the legs and two at different heights along the bell. The assembly was bolted to a metal pallet that secured in the bed of a truck, as shown in Figure 1. We then drove the truck around several paved and dirt roads to record different types of excitation at many levels.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==