The general philosophy for a Multi-Exciter Test (MET) is essentially the same as that of the Single Exciter case; however, there are additional considerations that need to be addressed in the conduct of a MET. It is addressing the additional considerations associated with MESA and MEMA, and assessing the adequacy of a laboratory MET, i.e., comparing the reference time histories or spectral content with the results obtained in laboratory based tests, that are the concerns of this Annex. As of the inclusion of this new test method into MIL-STD-810G, the primary vibration control system vendors offer MET options for time waveform replication (TWR), sine, shock, and random. Options for combined environments such as narrowband-random-on-random and sine-on-random are generally implemented via TWR based techniques.
In the simplest terms for MESA and MEMA tests, multiple exciters are employed to excite one or more mechanical- degrees-of-freedom. For traditional SESA testing, the test reference is provided as either a single reference time trace as discussed in Method 525.2, or in terms of simple magnitude versus frequency plots such as an auto spectral density as discussed in Method 514.8. For a MET, multiple channels are required in the control process. For a MET defined in the time domain, multiple time traces will be required, and for a MET defined in the frequency domain, cross spectral densities are required in addition to auto-spectral parameters in defining the test reference. For either case, the system identification (transfer function) estimation process is now a matrix operation as opposed to a simple division as in the SESA case.
The additional complexities associated with MESA and MEMA testing require an increased level of technical skill from the test engineers in planning such tests, and from the test operators that will ultimately perform the tests. Test objectives must be clearly understood to ensure that, in addressing the inevitable test-specific obstacles associated with any MDOF test, the test objectives are still properly addressed.
527.2-1 TRANSFER-FUNCTION ESTIMATIONS
Exploiting the over-determined feedback technique as discussed in Annex A is also advantageous in providing a weighting technique, analogous to the common practice in traditional SDOF testing in which various multiple-channel averaging schemes are employed to address payload dynamics issues. In the conduct of an MDOF vibration test, if an over-determined feedback scheme consisting of properly placed linear accelerometers is employed, H xy is approximated in a Least-Squares sense, thereby providing a sound method of implementing a multi-channel control scheme. However, as is the case for the general 1-DOF case, one should always optimize the fixture design because no control scheme will force motion of a structure in an unnatural manner. The accuracy of the Least Squares approximation of H xy will be directly related to the degree of modal similarity between the field deployment platform and the laboratory test platform.
Based on the previous discussion of kinematic considerations for transducer placement, it is clear that great care must be taken to establish a central point to which all measurement locations could be referenced. Carefully measure and record the specific location and polarity of each transducer. In addition, this process requires forethought as to how the test item will be fixtured in the laboratory to ensure the “exact” measurement locations can be used.
527.2-1 PROCEDURE I MET (TIME DOMAIN REFERENCE CRITERIA)
527.2-1 PROCEDURE II MET (FREQUENCY DOMAIN REFERENCE CRITERIA)
527.2-1 SCOPE
This Annex presents considerations and techniques for developing Laboratory Vibration Test Schedules (LVTS) that can be utilized to simulate field vibration environments on a vibration table. Laboratory vibration tests are used extensively in lieu of more time-consuming and less cost effective field exposure tests. This Annex specifically addresses random vibration testing controlled to frequency-domain vibration spectra and is intended to address multiple “exciter” (also referred to as “shaker” or “actuator”) scenarios with the emphasis on mechanical multiple degree-of-freedom (MDOF) scenarios. There is a significant increase in complexity between single-exciter/single-axis (SESA) and multiple-exciter/multiple-axis (MEMA) testing in terms of both mechanics and control. MEMA specific issues ranging from definitions and nomenclature consistency, to data analysis techniques, will be addressed.
1.1 Preprocessing
Since placement and orientation of transducers are paramount in the conduct of MDOF MET, performing a thorough pretest review is essential to overall test validity and efficiency. Misalignment of one transducer will adversely affect the transfer function matrix as a whole. To address these types of issues, take detailed measurements and photographs of the actual field setup (i.e., how and where the materiel was mounted) to aid in proper laboratory setup (since the laboratory configuration should mimic the field setup as accurately as possible). In addition, once the test item and associated measurement and control instrumentation are configured in the laboratory, examine phase and coherence measurements between drive channels and control channels to make sure that input points and their resultant responses are logical (e.g., a vertical input should largely affect vertical responses at low frequencies). Also, ensure the spectral characteristics of the control accelerometers and associated signal conditioning equipment have the same frequency response characteristics as that of the instrumentation used to make the original reference measurements, or properly pre-condition data as required to ensure proper phase relationships between channels.
527.2-1.1 Preprocessing
Since placement and orientation of transducers are paramount in the conduct of MDOF MET, performing a thorough pretest review is essential to overall test validity and efficiency. Misalignment of one transducer will adversely affect the transfer function matrix as a whole. To address these types of issues, take detailed measurements and photographs of the actual setup (i.e., how and where the item was mounted) to aid in proper laboratory setup (since it should mimic the field setup as accurately as possible). In addition, once the test item and associated measurement and control instrumentation are configured in the laboratory, examine phase and coherence measurements between drive channels and control channels to make sure that input points and their resultant responses are logical (e.g., a vertical input should largely affect vertical responses at low frequencies). Ensure the spectral characteristics of the control accelerometers and associated signal conditioning equipment have the same spectral characteristics of the instrumentation used to make the original reference measurements, or properly pre-condition data as required, to ensure proper phase relationships between channels. Also, it is highly recommended that an FEM model of the MET configuration be developed. A prior knowledge of the modal characteristics of a laboratory-based MET system often proves to be of great value in addressing undesired modal response through implementation of additional feedback to be considered in the control scheme.