Last Tuesday (September 23) afternoon we performed a test of speeding up the INJ HVAC supply fan. This is complementary to the slow-down test performed some time ago (https://logbook.virgo-gw.eu/virgo/?r=60550). The return fan was kept at constant speed reduced mode setting, 20.4 Hz. At each step the air speed out of the supply duct is measured with a hot-wire speed meter, in the first section of the supply duct  after the AHU. 

The air speed is the average of measurements taken in 6 positions in a grid across the duct section. Air flux is computed as: flux = speed * A * 3600. A is the area of the duct: 0.55 x 0.65 m2.

Inside the Laser Lab the microphone (ENV_LLR_MIC) was placed at 10-20cm distance from one air inlet (ceiling): Picture 1

We realized afterwards that the HVAC_INJ* monitors were off, as reported in https://logbook.virgo-gw.eu/virgo/?r=67784

However, the INF*PRES sensors measured an overpressure, indicating an increased air flux (Figure 2).

Microphones indicate that substantially the broadband acoustic noise inside the Laser Lab did not change, despite the measured flux into the lab increased by 60%. The only noticeable change in acoustic spectra is a narrow peak moving corresponding to the frequency of the supply fan. This is in Figure 2.

This result is a bit puzzling. Indeed this is in contrast with the observation made many times that when in "nominal" mode (supply at 37Hz, return fan at (30?) Hz)

We make two hypothesis:

• The supply fan is silent and the acoustic noise is mainly associated to the return fan.
• the HVAC is actually running in a "crazy" state, as noted in  https://logbook.virgo-gw.eu/virgo/?r=67740. This onset since the replacement of the HVAC UPS.

Conclusion: The INJ HVAC operation status should be checked. Further tests are needed to better address the noise produced by the INJ HVAC. Similar test to be done with the DET HVAC.