1 Advanced LIGO UK G040060-00-K Blade transmissibility by FEA Justin Greenhalgh, Rutherford Appleton Lab LSC, March 2004 LIGO-G040058-00-K
Dec 26, 2015
1
Advanced LIGO UK
G040060-00-K
Blade transmissibility by FEA
Justin Greenhalgh,
Rutherford Appleton Lab
LSC, March 2004
LIGO-G040058-00-K
2
Advanced LIGO UK
G040060-00-K
Contents
• Report– Motivation– Background– Results– Next steps
• Discussion– Is this the right way to go?– Any caveats etc?
3
Advanced LIGO UK
G040060-00-K
Background
• Norna (ALUKGLA0010) showed that the isolation in the quads (and triples?) will be OK provided that the internal modes of the different blades are high enough and do not coincide.
• Ken went on to assert (ALUKGLA0007) that high enough meant about 40 Hz.
• Picture from Norna.• Will the blades give adequate isolation?• In particular, will the peaks in transmissibility caused
by the internal modes overlap?• And what is the effect of the mass of the wire clamps?
6
Advanced LIGO UK
G040060-00-K
• So – find the actual transmissibility of the blades and multiply by the curves above (updated). (Even better, reproduce the curves above PLUS blade internal modes).
7
Advanced LIGO UK
G040060-00-K
Run through T040024
• Modal analysis of reference blade• Modal analysis of blade from Conceptual design• Harmonic
– No damping (cf Norna’s plot)– With damping, various damping ratios
• Test effect of changing alpha on internal modes• Prestress• Add wire• Effect of wire clamp mass
8
Advanced LIGO UK
G040060-00-K
T040024
• Pretty simple stuff over Christmas• Uses ANSYS macro language for ease of “what if”
scenarios• Simple to write macros so that a command of the
form – bf1,.48,.0045,.096,.01,20,1,1000,.050,11
• Will find the eigenvalues of a blade 0.48m long, 0.0045 thick, .096 root width, 0.01m tip width, find up to 20 eigenmodes in the range 1 to 1000 Hz, with .05kg wire clamp and 11kg mass.
9
Advanced LIGO UK
G040060-00-K
T040024
• ANSYS will find– Eigenvalues and normal mode shapes– Transmissibility in a given frequency range and
linear frequency step
10
Advanced LIGO UK
G040060-00-K
What equations is the FE solving?
• Modal analysis
• Harmonic analysis
• Damping
11
Advanced LIGO UK
G040060-00-K
FE equations
• So, use 1/(2Q) as the damping value in the ANSYS DMPRAT command.
12
Advanced LIGO UK
G040060-00-K
Modal analysis of reference blade
• Modal analysis of “reference” blade• Frequency has been measured at 55 Hz
17
Advanced LIGO UK
G040060-00-K
Harmonic – with damping
• “Zoom in” on peak by specifying restricted frequency range
29
Advanced LIGO UK
G040060-00-K
Note T040025
• Revised dimensions• Easy changes allows look at a set of blades• NB resolution of peak• Results so far
31
Advanced LIGO UK
G040060-00-K
Resolving the peaks…
• All three blades are analysed at a background range plus an “zoom” range for each one:
33
Advanced LIGO UK
G040060-00-K
Next
• So far – simply testing out the methods + tools• Next – get more detailed/serious
– Reference to this group – am I doing anything silly?– Check this method can reproduce Husman results– Get latest blade/wire dimensions– Check transmissibility vs measurements
• Calum’s thesis?• Elsewhere?
– Cross check with more detailed FE models (RAJ, MPL, others?)
– Explore effect of clamp mass again– Model sloping wires– What other effects to bear in mind?
34
Advanced LIGO UK
G040060-00-K
Discussion points
• Reference to this group – am I doing anything silly?• Check transmissibility vs measurements
– Calum’s thesis?– Elsewhere?
• Cross check with more detailed FE models (RAJ, MPL, others?)
• Explore effect of clamp mass again• Does this include “Norna’s curves” or do those need
to be added in to the results?• What other effects to bear in mind?