technology space 5 nonlinear analysis of the st5 ......may 22 - 23, 2002 space 5technology femci...
TRANSCRIPT
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 1
Nonlinear Analysis of the ST5 Magnetometer Boom
Wayne Chen/Code 542 NASA/GSFC
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 2
Agenda
? Problem Description? Linear Versus Nonlinear Analysis? Typical Model? Trial and Error, Part 1? Trial and Error, Part 2? Trial and Error, Part 3? Refining the Design? Latest Model? Latest Results? Current Status? Conclusions / Lessons Learned
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 3
Problem Description
? Boom mounted magnetometer is one of the primary instruments on ST5? Mission consists of a three S/C constellation to test nanosatellite technologies? Overall S/C dimensions ~ 18 in wide and ~ 10 in high? Overall S/C weight ~ 50 lb
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 4
Problem Description (continued)
? Because of postbuckling behavior, regular linear statics solution sequences not adequate
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 5
Linear Versus Nonlinear Analysis
? Linear analysis? Comprises bulk of the work done at GSFC? Useful for analysis of deployed boom (normal modes, thermal distortion, etc)
? Nonlinear analysis? Minimal GSFC heritage, though capability has existed in various analysis codes? Only recently has nonlinear analysis been used for thin membranes, MEMS,
and postbuckling
? Analysis of the boom has been marked by steady progress through a lot of trial and error
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 6
Typical Model
? Main items of interest are torque capability of joint and tape stresses? Variables include material, radius of tapes, and size of windows
Tube Tapes
One [45] ply of 0.005” T300
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 7
Trial and Error, Part 1
? Early runs did not take into account contact between the tapes and used rigid elements (RBE2) at each end to enforce a rotation
? Problem due to lack of contact between tapes is obvious? Behavior of end moment after snap-thru does not seem correct? Run time of 2.40 hrs
0.0
2.0
4.0
6.0
8.0
10.0
12.0
0.0 20.0 40.0 60.0 80.0 100.0 120.0
Angle (deg)
Mom
ent (
in-lb
)
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 8
Trial and Error, Part 2
? Contact between the tapes added and used rigid elements (RBE2) at each end to enforce a rotation
? Contact between tapes more correctly modeled? Behavior of end moment after snap-thru still does not seem correct? Run time of 9.50 hrs
0.0
2.0
4.0
6.0
8.0
10.0
12.0
0.0 20.0 40.0 60.0 80.0 100.0 120.0
Angle (deg)
Mom
ent (
in-lb
)
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 9
Trial and Error, Part 3
? Contact between the tapes retained and rigid elements (RBE2) at each end replaced with massless aluminum plate elements to enforce a rotation
? Contact between tapes and behavior of end moment after snap-thru both seem correct and clean
? Run time of 3.35 hrs? Important result was that the steady-state torque was too low
0.0
2.0
4.0
6.0
8.0
10.0
12.0
0.0 20.0 40.0 60.0 80.0 100.0 120.0
Angle (deg)
Mom
ent (
in-lb
)
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 10
Refining the Design
? Because torques from integral boom designs using one to several plies of composite were not high enough, investigated other alternatives
? Went from integral boom design to assembled boom design? Tube sections still made of composite? Tape sections made of Be Cu strips bolted to tube sections
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 11
Latest Model
One 0.006” Be Cu tape
Ti 6Al-4V shim
Four [0] plies of 0.005” T300
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 12
Latest Results
Moment Vs. Angle, 3.00 in Window
0
10
20
30
40
50
60
70
80
0 10 20 30 40 50 60 70 80 90 100 110 120 130
Angle (deg)
Mom
ent (
in-l
b)
Snap-thru Snap-back
0
10
20
30
40
50
60
70
80
0 1 2 3 4 5
Angle (deg)
Mom
ent
(in-
lb)
Snap-thru Snap-back
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 13
Latest Results (continued)
? Run time of 6.64 hrs? Important result was that the steady-state torque was increased by quite a
bit (up to ~ 1.4 in-lb)? Be Cu tapes stacked to nominally double steady-state torque (because of
additional complexity and excessive CPU time, did not attempt to run; verifying by testing)
Two stacked 0.006” Be Cu tapes
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 14
Current Status
? Individual boom joints as well as full-length boom in fabrication
? Torque testing of individual joints to begin shortly followed by G-negated deployment tests of full-up boom mounted to a S/C mock-up
May 22 - 23, 2002
5Space Technology
FEMCI Workshop 2002 15
Conclusions / Lessons Learned
? Significant progress made in performing and understanding the nonlinear analyses of the boom since the last FEMCI workshop
? Doing trade studies of the different variables in the problem not very efficient because of large CPU times needed for each run
? Future analysis to support test program as needed