NCR 00040
Root Cause:
The joint
was analyzed with the shoulder of the bolt acting as a pin carrying all the
shear load from the tower. However,
alignment tolerances required up to 4 mils of clearance between the shoulder and
the flexure and up to 2 mils between the shoulder and the hole in the
Grid. Friction from the bolt preload
proved to be not nearly enough to prevent movement within this tolerance during
random vibration. The movement caused
the bolt to loosen.
Contributing Factors:
Contributing factors were a low
torque specification and the fact that no secondary locking mechanism was
used. The root cause of the bad torque
specification was the lack of any reviewed assembly procedure documentation in
Italy and lack of an assembly drawing.
The root cause of the lack of secondary locking was a deliberate
position within the Tracker group that the test phase should be passed without
such locking.
Impact:
Redesign of
the flexure-Grid interface impacts the Grid design and the flexure design, plus
tooling for drilling the Grid and flexures.
It impacts the Tracker schedule by delaying completion of the EM
vibration test, delaying the EM thermal vacuum test, and possibly delaying
production of the bottom trays.
Corrective Actions
- Redesign
joint to act as true shear pins, with no slop. This has been done, including a complete re-evaluation of
the tolerance in the interface, as well as installation and verification
methods. The new design uses
adjustable cones to take out any misalignment between Tracker and Grid
while maintaining tight joint. See
drawings LAT-DS-50-1, 417-4, 422-4, 851-1, 1803-3, 1921-3, 2786-1, 2787-1,
2788-1, 2789-1, 2790-1, 2791-1, 2792-1, 2793-1, 2794-1.
- Complete
the assembly drawing and procedure for the Tracker tower, including
mounting on the vibration fixture.
A draft, LAT-DS-00722, now exists in LATDOCs. It will be reviewed and finalized in
preparation for the tower assembly PRR, to be held in mid April. The assembly procedure, LAT-PS-01854,
is being drafted and also must be finalized in preparation for the PRR.
- Include
torque specifications.
- Include
dry lubricant on the bolts to assure a reliable preload.
- Introduce
a secondary locking mechanism for the bolts, even during any future test
phase. The Grid inserts in the new
design include a locking feature, and the nut on the Tracker side will be
staked with epoxy.
- The
new joint design will be verified by a set of static tests and a new
vibration test, either with an engineering model or during Tower A protoqual
tests. The plan for verification will be posted by 31 January 2004
on the Tracker website. The decision whether to use an EM or Tower A
for vibration verification is still under discussion with Project
Management and Tom Venator of GSFC.