M. Barton, J. Kissel, B. Shapiro
After losing M0 RT, and failing to decouple F1 from Yaw, we went for the next best thing to tell us what's wrong with the suspension -- transfer functions. I attach a semi-complete set of transfer functions for the main chain (M0), and a complete set of Reaction Chain (R0) transfer functions.
Main Chain (2011-11-19_H2SUSITMY_M0_ALL_TFs.pdf):
Because we've lost M0 RT, we were not able to get either Vertical or Roll transfer functions. However, of the four degrees of freedom that we could measure L, T, P, and Y -- L, T, and Y compare well with the model. P looks like a total disaster, but is surprising given the poor results from diagonalization. However, because L, T, and Y compare well, the poor pitch results may just be a result of F1 being totally misaligned, and not a fundamental problem with the dynamics of the suspension.
Some notes on the analysis:
- I've switched to using a monolithic model as the reference, whose parameters can be found here:
${SusSVN}/sus/trunk/QUAD/Common/MatlabTools/QuadModel_Production/quadopt20060509productionmainasbuiltoffdiag_fiber.m
- I've used a special version of plotquad_dtttfs.m to generate these sets of plots, because it took a good deal of finagling to get the script to work with less than all six degrees from freedom. It can be found here:
${SusSVN}/sus/trunk/QUAD/H2/ITMY/SAGM0/Scripts/plotquad_dtttfs_111119_0251_NoVR.m
Probably not a good idea to try and use it in the future, it was quick last minute hack up.
- Because the monolithic and metal configurations' dynamics are so different, I've elected to not compare this measured transfer function against the other measurements we have of main chains (since they've all been metal dummy masses up to this point).
Reaction Chain: (allquads_111120_H2SUSITMY_R0_ALL_ZOOMED_TFs.pdf)
Here, this is also the first time we're looking at a glass reaction mass, but I don't expect the dynamics to change that much between metal and glass. However, expectations may not be well grounded, because the degrees of freedom where were expect very little change (because they are relatively insensitive to physical parameters, or are defined by tightly control mechanical dimensions), V and Y, show changes. The attached plot shows this same suspension as its reaction chain evolved:
ORANGE (H2SUSITMY, "Metal, Before Lacing").
2011-08-10 measurement of H2 SUS ITMY
- Metal dummy masses in place.
- In the LVEA, on the mechanical test stand 2, suspended from a solid stack.
- Protected from air currents with a C3 parachute.
- No lacing cables in this measurement.
- No cage stiffening elements are in place
BLACK (H2SUSITMY, "Metal, After Lacing 1").
2011-08-15 / 2011-08-12 measurement of H2 SUS ITMY,
- Metal dummy masses in place.
- In the LVEA on the mechanical test stand 2, suspended from a solid stack.
- Protected from air currents with a C3 parachute.
- All lacing cables installed but for the ESD cabling (i.e. UIM, and PUM cables).
- No cage stiffening elements are in place
MAGENTA (H2SUSITMY, "Metal, After Mating").
2011-08-25 measurement of H2 SUS ITMY,
- Metal dummy masses in place.
- In the LVEA on the mechanical test stand 1, suspended from a locked BSC-ISI.
- Without the protection of a C3 'chute.
- It has all lacing cables installed but for the ESD cabling (i.e. UIM, and PUM cables).
- All stiffening elements are in place, including stiffening sleeve, lower structure side panels, and upper structure cross-braces.
CYAN (H2SUSITMY, "Glass, Fully Laced").
2011-11-19 measurement of H2 SUS ITMY,
- Glass TCP masses in place.
- In the LVEA on the mechanical test stand 1, suspended from a locked BSC-ISI.
- Without the protection of a C3 'chute.
- It has all lacing cables installed, including connection to TCOP
- No cage stiffening elements are in place
We see that L, T, and R remain constant. Awesome.
Both the Highest (visible) V Mode (pg 3) has increased from the expected 3.6 Hz to 3.9 Hz, and the Second Yaw Mode has increased from 1.3 Hz to 1.358 Hz. I'll need more information from the modeling experts before I can make a claim as to what causes such change (mostly because I've never had to think about these DOFs before!).
As usual, pitch remains as the most dynamic (haha, get it?) degree of freedom. It appears as though with this configuration of lacing cables, the second pitch mode remains at 1.8 Hz between Metal and Glass TCP, but the lowest mode is now clearly resolved and split between 0.71 Hz and 0.81 Hz. Further, it appears as though we have found a configuration that effect the highest modes, specifically noting the frequency and shape of the response at 3.4 Hz and 3.9 Hz.
So, in summary -- we've got lots of [thinking / investigating / discussing / adjustments] to do on this suspension before we can move on.
Welcome to the land of glass, H2SUSITMY!
