Summary of modeling discussion with the Hytec design team  (reported by T. Thurston)

 

Tom Borden, Erik Swensen, Franz Biehl, Tim Thompson, and myself met throughout the day on August 6, 2002 to discuss the current status and progress in understanding the structural failure resulting from the tracker prototype vibration tests.

 

The development and evolution of the models were discussed.  The current model is in its third generation.  Models 1 and 2 were developed for the proposal and during the early design of the Tracker modules.  Model 3 has been developing over the past year to account for design changes and more specifics features related to the flexure supports and tower faceplates.

 

The primary interest of the discussions was to gain an understanding of the FEA model resolution and fidelity and to understand the correlation of the FEA analyses to the predicted failure.  A secondary interest was to build ART confidence in the analysis establish a path for resolving analysis issues.

 

The results of model 2 have been presented in HTN-102070-0005 with comparisons to the tracker tower ’01 prototype random vibration test in HTN-102070-0008.  These results were summarized also in PDR design report (January 2002 LAT Review) for the tracker tower design in report HTN-102070-0002. 

 

Model 3 is an upgrade modification of model 2 which added several design features related to the flexure implementation and element selections (Basically the model reduced the number of shell elements and increase the number of brick elements).  The results have been presented in the most resent review, July 2002.

 

The FEM evolution presented at the meeting was as follows:

·         Tower model

o        Model 1

§         Trays represented using shell elements

§         Closeouts represented using beam elements 

§         Fixed base boundary conditions on the bottom tray

§         Sidewall fasteners were not represented

§         Modified model to include fasteners

o        Model 2

§         Trays represented using solid elements for the closeouts, solid elements for the core, shell elements for the facesheets

§         Sidewalls represented using shell elements

§         Fixed base boundary conditions on the bottom tray

§         Payload stiffness effects were studied and represented as layers on the facesheet; studied stiffness effects with and without payload

§         Sidewalls were first attached all along the sides and modified to include specific fastener locations

§         Modifications were made to include the individual attachment points of the sidewall

§         Included flexures (1st concept) with fixed base boundary conditions at the bottom of the flezures

§         Included V-therm in the model and studied the stiffness effects of the gasket

o        Model 3

§         Model 2 with new flexure concept

§         Top flexure bracket rigidly attached t the bottom tray

§         Modified flexure attachment to include fasteners

·                                 Prototype Tower Model                    

§         Representative model of the prototype test tower including aluminum trays

·                                 Other Models

§         Tray models representing each of the five tray types using solid elements for the closeout and core, and shell elements for the facesheets and payload

§         Boundary conditions represent fastener locations to the sidewalls

 

Models were illustrated and analysis discussed

 

Comments and notes:

·                     Models 1 and 2 primarily focused on displacements.  Model 3 focusing on dynamics.

·                     They believed load spectrum is too high which drives margin to be high – “need coupled load analysis

·                     Grid stiffness not included in Hytec modeling

·                     The modeling and testing has not taken in to account CLA and any impact spectrum impact that may occur

·                     Mode shape and frequencies observed in the FEM modeling

o        Mode 1&2 (tipping)  - near 90 Hz

o        Mode 3 (plunge) – near 235 Hz

o        Mode 4 (twist) – near 260 hz

o        Mode 5 (c.g. rock) – near 320 Hz

·                     Work on getting sample to GSFC – 1”cut from 16”bar

·                     Design rules – The design rules came from NASA guidelines

·                     Model conversion – Keep it going as planned

 

Actions of Meeting:

§         Add Hytec reports on FEA modeling to our website.

§         Plan to continue:

o        SLAC:    Get grid stiffness to Hytec ASAP

§         Get grid stiffness to Hytec ASAP

§         Get CLA to Tracker interface to Hytec ASAP

§         Fund Hytec to include partial grid in their FEA model.

o        Hytec:

B.       Compare model 3 (current version) results with 6/19 results.

C.       Investigate –1 & -2 corner bolts to see redistribution at strain/stress/loads.

D.      Determine stress/strain levels in bottom tray corners from results at model 3.

E.       Hand calc or detail corner from model.

F.       Apply proposed solutions to model 3.

G.       Repeat C&D for proposed solutions

H.      Show additional margins.

 

Conclusions:

·         The model appears to be assembled correctly, test results appear to correlate well with model 2.

·         The mode shapes and deflection patterns of the models appears match expectations and reinforce the need to look closer at the stress levels in the area of the failure.

·         Analysis is not as mature as necessary to instill a sense of understanding of the failure and associated margins in and around the area of the failure.  Either a refined FEM model or an extraction of FEA results for use in hand calculations needs to be performed to demonstrate the stress and strain relationships associated with the failure.

·         There should be review points established at completion of points C, E, and H of the actions plan presented above.