COMS Antenna Door FEA
We performed Static Structural, Modal, and Random Vibration Finite Element Analysis (FEA) in Ansys Workbench to validate design changes to the COMS Antenna Door (UMS-0358 Rev 1). This analysis consisted of an initial static structural analysis to determine the stresses exerted on the door by the antenna, torsion spring, and burn wire. These results were fed into a modal analysis to determine the door's first six modes, and then finally three random vibration analyses were performed for each principle axis.
A Young's Modulus of 0.5 GPa was used, the lowest strength Delrin formulation at 60 C, as well as a density of 1410 Kg/m^3, and a Poisson's ratio of 0.35.
Based on this analysis, the antenna door has sufficient margin (FOS>20) to survive launch and the changes to the design will not affect its performance.
Boundary Forces
First, the boundary conditions were modelled by:
1) Measuring the force that the antenna exerts when coiled, using a Newton Spring. The total force was approximately 0.4 N. We assumed that this force is being applied equally from the antenna onto the door via the two bolted connections.
2) Calculating the force applied by the UMS-0183 torsion spring applies to the antenna door when the door is rotated 79 degrees, the maximum that the antenna can rotate before coming into contact with the UMS-0361 "bucket."
From the supplier, the torsion spring has a maximum torque of 0.099 in.-lbs when rotated 225 degrees. Assuming a linear torque, rotating the spring 79 degrees should yield 0.0348 in.-lbs of torque. Assuming this torque is applied at the end of the spring's arm, 0.75" from the spring's centre axis, the spring should apply 0.026 pounds of force (0.116 Newtons).
3) We calculated the force applied onto the door by the burn wire chord by assuming that the chord was in sufficient tension to counter the antenna's spring force, and the torsion spring's torque. This force was 0.28 Newtons.
Boundary Conditions
The boundary conditions shown in the figure below consisted of the various forces described above, point masses representing the antenna, and boundary conditions to represent the connection between the door and the rest of the spacecraft. The torsion spring force was applied over an area equal to the spring's wire diameter, the antenna spring force was applied over two areas, equal to the bolt-head diameters which attach the antenna to the door, and the burn wire force was applied over the curved area where the wire will pass through the door.
The antenna's mass was modelled as two 15 g point masses, applied to the two bolted connections.
The hinge joint was modelled with one frictionless support, one cylindrical support, and one fixed support. The frictionless support prevents movement in the z-axis but allows for translation in the x and y, as well as rotation in z. Similarly, the cylindrical support allows rotation and translation in z but restricts translation in x and y. The fixed support was necessary for the random vibration FEA to solve but is not fully representative of the physical system. This approximation causes inaccuracies in the stresses near the hinge but will cause higher stress concentrations than should actually be expected. As such, this model is conservative.
Figure 1: Boundary Conditions
The model was then meshed as shown in Figure 2. All mesh conditions were set to their defaults. The factors of safety for the FEA results were sufficient that a sensitivity analysis was not necessary.
Figure 2: Meshing
Static FEA Results
A static analysis was needed to incorporate the applied forces into the random vibration. The results of this static FEA are shown in Figures 3 and 4 below. The maximum von Mises stress was 576 kPa, well below Delrin's 55 MPa yield stress at 60 C, and the maximum deformation was approximately 0.4 mm at the far end where the burn wire is attached. The resulting factor of safety was approximately 95.
Figure 3: Von Mises Stress for Static Loading
Figure 4: Strain for Static Loading
Modal Analysis
Once the static FEA was completed, the results were fed into an Ansys Workbench Modal analysis using the same boundary conditions. The resulting modes are shown in Figure 5.
Figure 5: First 6 Modes
Random Vibration Analysis
Finally, the above modal study was used to create three random vibration FEA studies. Each study applied the Nanoracks soft-stow vibration profile, shown in Figure 6, to each principle axis. The same boundary conditions were used as the previous studies.
The maximum von Mises stress was 2.46 MPa in the y-axis, as shown in Figure 7. However, this stress is near the fixed boundary condition and likely overstates the actual stress that the system will experience. The local maximum near the antenna storage slot, "Max 4" in Figure 8, was approximately 1.64 MPa which is acceptable.
Despite the conservative boundary conditions, this maximum stress has a conservative factor of safety of approximately 22. These stresses are also below Delrin's fatigue curve at 60 C, shown in Figure 9, and so we should not expect the door to fail under fatigue.
Figure 6 : Nanoracks Soft-Stow Vibration Profile
Figure 7: Antenna Door von Mises Stress During Y-Axis Random Vibration FEA
Figure 8: Local Maximum von Mises Stresses During Y-Axis Random Vibration FEA
Figure 9: Fatigue Curve for Delrin from Dupont