KAGRA Logbook

VIS (SR3)

enzo.tapia - 11:14 Monday 10 September 2018 (6073)

SR3 IP LVDT - Coil actuator diagonalization

With Eleonora and Lucia,

We have implemented a diagonalization matrix of the coil-magnets actuators of the IP. First we checked the position of the sensor (LVDTs) in the CAD model and we also measured this (Figure 1). We excited the 3 different coils with a sinusoidal signal at 3 Hz.

We include the TFs for L, T and Y of the IP before (black) and after (blue) the diagonalization (Figure 2).

There are still some issues to be investigated, like the phase shift and the coupling to IP Yaw at 0.125 [Hz].

Later we should add more mass on the IP to get lower frequencies for L and T and then calculate again a diagonalization matrix for this configuration.

Note: this Preisolator (PI) is different from the one of BS.

Images attached to this report

Comments to this report:

yoshinori.fujii - 20:59 Monday 10 September 2018 (6075)

Thanks for this work:) I have two quick comments just for a check:
- I think you can refer the transfer functions like "IP_DAMP_L / IP_COILOUTF_H1_OUT", instead of "IP_DAMP_L / IP_COILOUTF_H1_EXC". Otherwise we have to change the actuator polarities somewhere later.
- Can you please make plots which hold "IP_DAMP_L / TSET_L_EXC", "IP_DAMP_T / TSET_L_EXC" and "IP_DAMP_Y / TSET_L_EXC" in one figure, for example?

enzo.tapia - 15:52 Wednesday 12 September 2018 (6113)

Thanks for your comments!

About the phase plots: I will try to measure again using the OUT channel instead of EXC channel to solve the polarity issue (phase), but I think we should fix this issue by adding this '-1' to the new diagonalized actuation matrix and not at the OUTPUT filters.

And for the plots you were asking for: Here are the results of the IP (L,T and Y) before and after diagonalization. (figure)

PD: We may want to try this again from the beginning by injecting a signal in the actuation coils of a frequency of ~2 Hz. And also start thinking about the frequency dependent diagonalization.

Images attached to this comment