Alaska Research CubeSat 1 Separation Switch Sub-Assembly Failure Analysis

Transcription

1 Alaska Research CubeSat 1 Separation Switch Sub-Assembly Failure Analysis Matt Pacheco Author/Engineer Signature of Approval Date Denise Thorsen Director Signature of Approval Date Alaska Space Grant Program University of Alaska Fairbanks Duckering Hall Tanana Loop Fairbanks, Alaska 99775

2 Revision History Rev Level Description Author Date Approval 1 Initial draft M. Pacheco 3/25/2016 1

3 TBD/TBR Item No. Location Summary Author Due Date 2

4 Table of Contents 1. Introduction Purpose Scope Document Organization Reference Documents Internal Documents Industrial Documents Pre-Launch Testing and Results /9/2015: Vibe Test /10/2015: Vibe Test /11/2015: Vibe Test /24/2015: Mission Readiness Review /12/2015: Vibe Test Post-Launch Testing and Results /28/2016: Vibe Test /5/2016: Vibe Test /25/2016: Vibe Test Analysis

5 Table of Figures Figure 1: ARC1 satellite with anodization removed from one rail due during vibration test Figure 2: Separation switch subassembly in the stowed position... 8 Figure 3: Separation switch subassembly in the deployed position... 8 Figure 4: Photo showing lit LED while in Test Pod Figure 5: Photo showing missing roller lever on separation switch... 7 Figure 6: Revised separation switch sub-assembly design Figure 7: Photos after vibration test showing spring plunger not released Figure 8: Photo showing LED switch indicator on during vibration testing Figure 9: Photo showing the released spring plunger after vibration test Figure 10: Photo and Solid Works drawing of spring plunger engaging the separation switch leaf spring Figure 11: Modeled spring plunger (a) depressed, (b) released, (c) exploded view

6 1. Introduction 1.1 Purpose The ARC1 satellite was launched October 8, 2015 and failed to transmit its beacon. This document analyzes one possible failure mode that could have prevented the ARC1 satellite from turning on. 1.2 Scope Included in this document is a description of the vibration tests required by the Launch Service Provider (LSP) that was performed prior to launch on the complete ARC1 satellite and the actions that resulted from those tests. Also included in this document is a description of the vibration tests performed on the structure subassembly after launch and an analysis of the failure that was observed. 1.3 Document Organization Section 1 Specifies the purpose and content of this document. Section 2 Lists reference documents. Section 3 Provides an overview of the testing that occurred prior to launch and the actions that resulted due to the results of that testing. Section 4 Provides a description of the testing that occurred after launch and the analysis of the failure that was observed. Section 5 Provides an analysis of the cause of the failure. 2. Reference Documents The following documents listed here were used as a reference for this document. Please refer to them for detailed information not included herein: 2.1 Internal Documents 1 ARC1-SMT-MNL- VibrationTestingProcedure-R0.pdf Required procedure for vibration qualification. 2 ARC1MRR-R1.pptx Mission Readiness Review 2/24/2015 PowerPoint slides 2.2 Industrial Documents 3 test_pod_user_guide.pdf Test Pod User s Guide version 6 4 1U_UsersGuide_Rev8.3.pdf 1U TestPOD User s Guide 5 VID_ _ mp4 Video of vibration test showing LED flicker indicating switch chatter. 5

7 3. Pre-Launch Testing and Results All vibration tests are performed according to test procedure [1] with the satellite installed in a Test Pod provided by the Launch Service Provider (LSP). The Test Pod User Guide [3, 4] describes the satellite integration procedure /9/2015: Vibe Test 1 Test Pod User s Guide version 6 [3] was used to integrate the satellite with the Test Pod. During vibration the satellite was observed to be rattling around pretty good in the Test Pod. The test was immediately halted. On closer inspection it was determined that the compression plate on the Test Pod was not properly set. When removed from the Test Pod, the satellite rails were seen to have been gouged in vibration, removing the anodization on one of the rails (see Figure 1). Figure 1: ARC1 satellite with anodization removed from one rail due during vibration test. Results and Actions: Test Pod compression plate not properly set Discussions with the LSP led to the discovery that an incorrect version of the Test Pod User s Guide [3] was being used. An updated version of the Test Pod User s Guide [4] was acquired. Damage to satellite rails Photos were provided to LSP for inspection and were given the approval to continue. 6

8 3.2 2/10/2015: Vibe Test 2 Test Pod User s Guide revision 8.3 [4] was used to integrate the satellite with the Test Pod. This test was halted after the X-axis vibration when a lit LED was observed through the USB port hole indicating that the satellite had turned on in the Test Pod (see Figure 2). Inspection of the satellite showed that the roller lever of the Omron J-V7 switch broke off (see Figure 3). Figure 2: Photo showing lit LED while in Test Pod. Figure 3: Photo showing missing roller lever on separation switch 7

9 The separation switch sub-assembly is shown in Figure 4 and Figure 5. This sub-assembly consists of a rod which runs down through the tab in the X+Y- rail and connects to an aluminum block. The aluminum block is angled on one corner and pushes against the Omron J-V7 switch roller lever when engaged. The aluminum block is held in place by a captive panel screw and is pushed upwards by a spring. Figure 4: Separation switch subassembly in the stowed position Rod Aluminum block Spring Captive Panel Screw OMRON switch Figure 5: Separation switch subassembly in the deployed position 8

10 Results and Actions: Broken Omron J-V7 separation switch roller lever Visual inspection of the switch showed that the roller lever was attached to the switch housing with a spot weld. This spot weld appears to be fragile in vibe. The broken switch was replaced and epoxy added at the point of connection to support the spot weld /11/2015: Vibe Test 3 Test Pod User s Guide revision 8.3 [4] was used in this test to integrate the satellite with the Test Pod. This test proceeded to completion with no further structural failures /24/2015: Mission Readiness Review Although ARC1 technically passed the required vibration testing on February 11, the team continued to be concerned about the reliability of the existing separation switch sub-assembly. Specifically, the epoxy used to support the roller lever spot weld also made the lever pivot stiff which caused the lever not to reliably release. The mechanical engineering team explored alternate designs that could be implemented quickly and proposed retrofit actions prior to the Mission Readiness Review. The retrofit was approved (2/20/2016) prior to the Mission Readiness Review with a required re-vibe of the satellite. The retrofit included replacing the current Omron switch with a wraparound leaf spring eliminating the spot weld. Additionally, the previous separation switch plunger mechanism was replaced with a purchased retractable spring plunger. The new plunger required that the switch also be moved. The team believed that with fewer parts, a shorter push rod, and eliminating the roller, would result in a more robust and reliable separation switch sub-assembly (see Figure 6). Figure 6: Revised separation switch sub-assembly design. 9

11 3.5 3/12/2015: Vibe Test 4 During the Mission Readiness Review, and for the two weeks after, the proposed retrofit was created. This required building a new mounting block for the switch, drilling and taping the rail to accept the new spring plunger, modifying the spring plunger to fit the rail, and installing all new components. On March 11 the Day In The Life test was accomplished and on March 12 the final vibration test was performed and passed. At this time, a failure was not suspected because the structural integrity of the satellite was intact. Upon removing the satellite from the Test Pod we noticed that the spring plunger did not immediately release. After a slight disturbance the spring plunger released. At the time it was believed that thermal cycling would release the spring plunger should it become stuck again during launch. This issue was not further investigated. On March 19 the delta MRR was held via teleconference and on March 25 the satellite was delivered. 4. Post-Launch Testing and Results The ARC1 satellite was delivered to the Launch Service Provider on March 25, 2015 and launched on October 8, 2015, 6.5 months later. During this time there was no access to the satellite. After the launch the satellite did not start to beacon. Immediately, investigations into possible failure scenarios were started. Given the difficulties with the separation switch subassembly during pre-launch testing, a second structure assembly with the separation switch subassembly was built in order to perform more detailed tests. Although our complete board stack was not included in these tests, a microcontroller board was attached to the separation switch to measure the degree of chatter on the switch /28/2016: Vibe Test 5 A modified version of the vibration tests, including only the random vibe tests, was performed. The Test Pod access ports were not covered which allowed monitoring a LED connected to a microcontroller board and the separation switch sub-assembly. The switch chatter test was inconclusive since there was enough switching noise with the table idle to make the microcontroller think that the switch was being switched. When the structure assembly was removed from the Test Pod it was observed that the separation switch plunger had not released (see Figure 7). However with a slight bump the separation plunger did release /5/2016: Vibe Test 6 The purpose of this test was to monitor when the spring plunger releases after being removed from the Test Pod. A microcontroller board, powered by a 9 V battery, was used to mark the time that the separation switch closes. The separation switch powers a MOSFET used to signal the MSP430 and a LED was used for visual indication. The intent of the LED was for additional visual verification when the switch pops up after vibration. The expectation was that it would be off during the vibration test. During the X-axis vibration test it was observed that the LED was turning on (see Figure 8). The LED was not visible during the Y-axis test because of the 10

12 mounting brackets. When running the Z-axis profile the LED did not turn on. Additionally, a video of the flickering LED was captured [5]. As with the previous test the spring plunger did not release after being removed from the Test Pod. The structure assembly was placed on a protected shelf for over a week. The spring plunger did not release during this time. Figure 7: Photos after vibration test showing spring plunger not released. Figure 8: Photo showing LED switch indicator on during vibration testing. 11

13 4.3 2/25/2016: Vibe Test 7 During this test the structure assembly without the separation switch leaf spring was vibrated. When the satellite was removed from the Test Pod it was observed that the spring plunger had released (see Figure 9). Figure 9: Photo showing the released spring plunger after vibration test. 5. Analysis The Post-Launch testing showed that the spring plunger did not reliably release after vibration when the separation switch was present. It did reliably release when the separation switch leaf spring was not present. Figure 6 shows the spring plunger in the released state. When the spring plunger is pushed down, it engages the leaf spring of the Omron switch, bending it down as shown in Figure 10. The bent leaf spring puts a horizontal force on the spring plunger causing it to tilt slightly inward at the top. A close examination of Figure 7 shows that the spring plunger appears to be jammed on the inside corner of the housing confirming that the force of the leaf spring causes the spring plunger to tilt inward at the top. It is believed that the force of the leaf spring on the spring plunger causes the spring plunger to strike the inside of the spring plunger housing while under vibration. This results in the spring plunger getting stuck. One suggestion is that the vibration creates a friction weld between the plunger and the housing. At this point the most reliable path forward is to redesign the separation switch assembly to remove the adverse lateral force of the leaf spring. 12

14 Figure 10: Photo and Solid Works drawing of spring plunger engaging the separation switch leaf spring. Figure 11: Modeled spring plunger (a) depressed, (b) released, (c) exploded view 13