Vibe Shock Test

In order to ensure the satellite can withstand the harsh launch environment, the ELFIN team put the spacecraft through extensive shock and vibration testing. Spanning multiple test campaigns over several years, the shock and vibration testing helped to shape the structural design of the ELFIN CubeSat. 

EM 1 on the vibration table at Cal Poly SLO. Accelerometers can be seen mounted on the Test POD as well as the solar panels of the CubeSat.

EM 1 on the vibration table at Cal Poly SLO. Accelerometers can be seen mounted on the Test POD as well as the solar panels of the CubeSat.

Test Setup

The ride up to Low Earth Orbit (LEO) aboard a Delta II can be pretty rough. In order to ensure that the spacecraft could survive the trip, it was put through environmental testing. These tests consisted of two components: random vibration and shock. The random vibration simulates the CubeSat shaking around in the P-POD as the Delta II climbs to orbit, and the shock testing is representative of stage separation and deployment.

The vibration testing was conducted on a shaker table that sweeps through various frequencies over several minutes of vibration. Accelerometers are placed throughout the Test POD prior to testing to measure any acceleration seen by the spacecraft. A sine sweep is performed before and after the vibration test which allows the team to recognize any shifts in the peaks, indicative of relative motion inside the spacecraft due to vibration. The spacecraft is inspected after each vibration test for any damage or loose fasteners. After inspection, the orientation of the spacecraft is changed, and a different axis is evaluated.

Shock testing was not implemented until the first Engineering Model was tested. Cal Poly SLO developed a shock testing apparatus consisting of a large aluminum plate, a Test POD fitting, and a massive pendulum. The pendulum is raised and dropped, striking the plate and providing the required shock. Again, accelerometers are placed on the Test POD and CubeSat to record all test data.

Development Model Test Campaigns

While most satellite missions utilize simulation software to test their preliminary designs, the ELFIN team did not have that luxury. Without proper guidance and the knowledge of how to properly set up the simulations, the Structures and Mechanisms team instead opted to build relatively cheap Development Models of the satellite and put them through vibration testing in order to analyze the structural integrity of the spacecraft bus.

The first DM build was composed entirely of mass models and had no functional hardware. Due to the magnetic sensitivity of the instruments on board ELFIN, only brass screws were used to fasten all the subassemblies to the chassis. However, all the screws used to mount the Electronic Particle Detector (EPD) to the chassis sheared completely, leaving the EPD mass model suspended in the chassis. The number of fasteners used to constrain the EPD was tripled going into the second DM build.

DM 2 faired much better than the first but was still deemed a failure. One of the fasteners used to mount the EPD sheared, and another backed out. A fastener QA program was implemented to inspect and torque test all screws before installation in an assembly. Additionally, non-magnetic A286 and 316SS steel fasteners were used in place of brass where they were deemed structurally necessary and significantly far from any magnetically-sensitive instruments.

Again, DM 3 vibration testing was not a success. This time, only one EPD screw baked out of its staked position. After inspecting the DM 3 build, the team decided to stiffen the +X “tophat” and add two additional mounting brackets to the EPD. Despite all these efforts, DM 4 still resulted in mechanical failures.

The team could not risk any more failures, so major design changes were implemented after the fourth Development Model was tested. Additional steel fasteners were used in place of brass to mount the EPD and stacer can to the chassis as they are the heaviest subassemblies. Open-cell foam was placed between the circuit boards in the instruments stack and avionics stack to reduce “drum-heading” and keep the boards in their correct positions. The Fluxgate Electronics Board (FGE) was also stiffened with an aluminum H-brace and a PEEK strap across the “gold brick.” Initially the brick was only fastened to the board by the solder joints; the strap takes the strain off the joints and better secures the brick to the board.

Cal Poly SLO’s shock testing system.

Cal Poly SLO’s shock testing system.

EM 3, FM A, and FM B in their Test PODs awaiting shock testing.

EM 3, FM A, and FM B in their Test PODs awaiting shock testing.

Engineering Model Test Campaigns

In the summer of 2017, the first Engineering Model underwent both shock and vibration testing. For the first time in the history of ELFIN, the EPD was assembled with live particle detectors. While mass models were still used for the FGE and the EPD board stack, this was the most complete an ELFIN assembly had ever been.

The 2-56 brass screws that were used to fasten the solar panels to the chassis posed quite an issue during EM 1 assembly. One of the final screws on the last panel sheared during installation. This was the straw that broke the camel’s back and pushed the team to use steel fasteners for all external panels.

Despite the difficulties during assembly, the EM 1 build survived shock and vibration testing without a single mechanical failure. The satellite also passed functionality tests after testing and was successfully able to deploy the “tuna can” antenna.

The second Engineering Model, built to spec with fully functioning hardware, was tested alongside the two flight units; all three spacecraft passed shock and vibration testing. Disassembly of EM 3 showed no damage or wear from testing, giving us confidence that the same is true for the flight units. FMA and FMB went through one more vibration test prior to launch.

It is no secret that ELFIN had many mechanical failures in the early days. However, all these failures led to insight that helped fortify the structural design. Proper simulations would have revealed many of these issues but assembling mass models and putting them through vibration testing gave us the results we needed in the time we had.