...

This was a ground test of the Hermes Pathfinder rocket. It occurred on Field A at 3:45 PM on November 10, 2017. It was the first use of the Pyrotechnic Piston to separate the rocket. The piston contained 1g of 4F black powder. The tender descender contained .2g of black powder. The shear pins used were not documented but were likely 2x 6-32 nylon pins and 1x 4-40 nylon pin. The rocket successfully separated, and the tender descender released the main parachute appropriately. After de-integration it was found that the Avionics Bulkhead was seriously deformed. This prevented the planned flight on November 11, 2017 from occurring.

Image Added

Failure Investigation

Black Powder Sizing Investigation

...

FEA Investigation

Analysis Summary

Solidworks FEA confirmed that the deformation is consistent with a 2-bolt load application.

...

• In actuality, this would be better modeled by an impact.
• Since significant deformation was observed (see the image belowabove), it is clear that a nonlinear analysis would be required to accurately model the deformation. The static load analysis was performed to get a basic idea of expected deformation shape and to prove material exceedance at the expected loads.
• The Solidworks part is modeled as one continuous part instead of two separate steps welded together (the weld was not meant to be structural, so this is not the most representative model)

Steps

The steps performed are as follows:

1. Aluminum 6061-O was chosen as the material because welding the inner step probably eliminated its temper.
2. No connections were set.
3. The outer step was chosen as the fixed geometry. Technically, the Avionics Bay Coupler only constrains the outer step in one direction (the vertical direction), but the bulkhead is also constrained radially within the Avionics Bay Coupler by the inner step. The bulkhead is not strictly constrained circumferentially except by friction after putting the Avionics Bay Coupler under compression (in an ideal execution of the design, which was not necessarily achieved for this test): this is irrelevant to the load applied.
4. 1750 pounds were applied to each of the two bolt holes used during the test. Technically, a more accurate representation would be to include the bolts in the model. Using just the bolt holes isn't perfectly accurate because the load was actually applied to the outer side of the bulkhead by the heads of the bolts rather than in the bolt holes (which were not threaded).
5. The finest possible mesh was generated. An image is featured below, which also summarizes the previous steps.
   
6. The simulation was then run.

Image Removed

Results

As the following figure shows (in particular, pay attention to the red arrows indicating yield stress vs. stress experienced by the material), the bulkhead saw serious material exceedance.

Image Added

Additionally, the d

 Additional Analysis is underway.

Media