Essais & Simulations 151

MESURES Figure 3 - MIMO

MESURES Figure 3 - MIMO Random control results in term of strain response PSDs. Solid Magenta, MIMO Random, two acceleration control channels (FuseFore:+Z, FuseBack:+Z); Solid blue, control on all the acceleration channels; Solid black, rectangular control on all the strain channels. Next step of the study was to improve the strain response in multiple strain channels adopting MIMO Control strategies. The results from the simple test case controlling all the strain responses without introducing a difference between the “pseudo-test” and “pseudo-operational” setup will not be shown. More interesting is however the case where a difference in the excitation mechanism is introduced by de-activating the inertial shaker located in correspondence of the DUT head and adopting the remaining two inertial shakers as inputs during the tests. Figure 3 illustrates the results of three different control strategies: two acceleration channels adopted as controls (FuseFore:+Z, FuseBack:+Z), all the acceleration channels adopted as controls and finally all the strain channels adopted as controls. The results illustrated in the Figure show that using MIMO control enables replication of multiple responses simultaneously. In this sense, increasing the number of control sensors on the structure improves the replication of multi-input operational responses. This propagates to uncontrolled channels too (measure channels). Figure 3 shows a drastic improvement of the measured strains when controlling just only two acceleration channels. Deviations are still observed but are much limited if compared with the strain responses of Figure 2, where only 1 acceleration was defined as control, even though the excitation mechanism was substantially different in this test-case (1 missing force input). The deviations further reduce when controlling all the acceleration channels and lead to a perfect replication when all the strain sensors are used as controls. Figure 1 (c) shows the configuration adopted for final test. The setup has been modified to introduce dramatic differences in the setup with respect to the pseudo-operational tests. The inertial shaker on the DUT head was de-activated. The shaker attached to the upper part of the DUT tail was moved to the lower part and the collocated lumped mass was completely removed. Figures 4 and 5 show the MIMO Random Control results in terms of acceleration and strain response PSDs, respectively. Although bigger deviations are observed with respect to the previous use-cases due to the dramatic difference between “pseudo-operational” and “pseudo-test” conditions, also for this use-case the results show that controlling directly on strain improves the replication of the strain field (at the price of the replication of acceleration responses). CONCLUSION Adopting Multi-Input Multi-Output (MIMO) Random Control enables the replication of multiple random responses simultaneously. MIMO Random Control therefore allows for a better replication of multi-input operational responses compared to legacy Single-Input Single-Output (SISO) Random Control. Figure 4 - MIMO Random Control acceleration response PSDs. Solid green: operational measurements. Dashed black: control on acceleration. Solid red: control on strain sensors. 22 I ESSAIS & SIMULATIONS • N°151 • Novembre - Décembre 2022 - Janvier 2023

MESURES Increasing the number of control channels (controls) allows for a better replication also for uncontrolled responses (measure channels). These considerations, well known in the MIMO testing communities, are shown in this work to apply also for control on strain responses. Because the differences in strain response of the Device Under Test (DUT) is correlated to differences in stress distribution, controlling directly on multiple strain sensors may lead to a closer replication of the fatigue failure mechanisms of the DUT during random vibration control tests ● Umberto Musella, Raphael Hallez et Bart Peeters Siemens Industry Software NV (Leuven, Belgium) Figure 5 – MIMO Random Control strain response PSDs. Solid green: operational measurements. Dashed black: control on acceleration. Solid red: control on strain sensors. PUBLI-COMMUNIQUÉ Un leader de l’instrumentation propose des solutions innovantes et compactes Trafag, fabricant suisse d’instrumentation depuis 80 ans, propose des produits dans les domaines de la détection et la mesure de pression, de température, de niveau et de densité de gaz. Dans le domaine de la pression, qui est son coeur de métier, Trafag propose à ses clients 3 technologies, ce qui permet d’offrir les solutions les plus adaptées, notamment aux fabricants de bancs de tests, avec des solutions spécifiques telles que : • Capteurs de haute précision, • Capteurs haute dynamique, jusqu’à 20kHz, • Capteurs avec afficheurs, • Capteurs paramétrables par les clients, au travers de logiciels gratuits sur smartphones, de faces avant des capteurs, ou de pockets, • Capteurs dédiés pour l’hydrogène. Il est intéressant de noter que Trafag se démarque de ses confrères par la compacité de ses produits, leur robustesse et leur précision, y compris dans les conditions les plus extrêmes, et une dérive dans le temps particulièrement faible. Pour ce qui est des applications hydrogène, la technologie s’affranchit à la fois des solutions piézo-résistive et céramique ainsi que de la dorure de la membrane, ce qui lui permet d’avoir un process de fabrication simple, particulièrement résistant, et beaucoup plus facile à compenser en température, donc beaucoup plus précis dans les applications où elle varie. EN SAVOIR PLUS : TRAFAG FRANCE 6-8 Rue René Cassin 91300 MASSY www.trafag.com/fr Max DUDIT - Tél. : +33 (6) 38 22 32 37 ESSAIS & SIMULATIONS • N°151 • Novembre - Décembre 2022 - Janvier 2023 I23