Simulation Analysis

ON THE SIMULATION PROCESS: PART I

              When the CAD model was completed, we had some trouble working on the simulation of the prototype. The most basic kind of simulation available on Fusion 360 is Static Stress Simulation. However, because our team is trying to build an assembly with three parts moving simultaneously with respect to each other at high speed (the central eccentric hub should rotate at the average speed of a cordless drill, which is between 500 and 800 rpm), it is likely that the Static Stress simulation that we set up is inappropriate for the type of stress load that we are dealing with in our project. The team carried out the simulation regardless and found some interesting result.

              The final CAD model that the team have is as follow:

Figure 1: Final CAD model of the compressor assembly.

Because the final model in Figure 1 has all of the auxiliary components that should not affect the result of the simulation in any meaningful ways, the team decided to use a simplified model (with the same dimension and materials) instead. This will help reduce the part count to the bare essential and speed up the simulation significantly.

Figure 2: Simplified CAD model containing only the four main components: main housing, rotor, eccentric hub, and crankshaft. Because the team is currently only interested in simulating the stress on the components and not the airflow around them, the inlet and outlet holes on the main assembly are also removed to simplify the problem.

              We decided to define the constraint of each of the parts as follow:

Figure 3: Pin constraint on the outside face of the eccentric hub to prevent it from moving in the y direction.

Figure 4: Pin constraint applied to the inside face of the rotor.

Figure 5: Pin and radial constraint on the circular face of the crankshaft. The only motion allowed in the crankshaft is rotational motion along its axis.

Figure 6: The main housing (especially the base) is constrained in all 3 dimensions.

We decided to simulate 2 sources of load in this simulation. The first is the torque load put on the crankshaft. In this simulation, this load is defined in terms of the crankshaft’s rotational velocity of 4800 degree per second (800 rpm).

Figure 7: Angular torque load applied to the circular face of the crankshaft.

              The second load simulated is the increased air pressure in the compression chamber. Since there are too many faces in the geometry of our assembly, we decided to not simulate atmospheric pressure (which would cause the simulation to be too complicated). Instead, we will just simulate the differential pressure inside of the compression chamber. We expect the total pressure inside the chamber to be about five times the atmospheric pressure at most, so the differential pressure simulated is 0.4 MPa, or about four times the atmospheric pressure.

Figure 8: Differential air pressure of 0.4 MPa applied to the portion of the rotor face and the inner housing face in contact with the air in the compressed chamber.

              The result of our simulation shows that the gear teeth of the gear on the main housing that is engaged with the rotor suffer a lot of stress and strain, which resulted in an abnormally low safety factor around the region.

Figure 9:  Low safety factor around the region of the gear teeth in contact with the rotor.

              Figure 10: Stress map of the contact gear teeth. The stress in some regions exceeds or comes very close to the tensile strength of PLA plastic of about 37 MPa.

Figure 11: Displacement map of the gear on the main housing. The region with the most displacement corresponds to the region that suffer from the highest stress, as expected.

Another portion of the assembly that suffer from high mechanical stress is the contact faces between the crankshaft and the eccentric rotor.

Figure 12: Low safety factor in the contact region between the crankshaft and the eccentric hub.

Figure 13: Stress map of the eccentric hub. The stress is particularly high in the corners of the square hole, as expected.

The team concludes that while most of the simulation result does not seem out of place, we should not make design changes based on them. We agreed that the simulation method used to generate these results is too crude and needs to be revised. To this end, we intend to consult the TA in our lab section and our instructor, who are more experienced with other, more advanced finite element analysis simulation approaches that may be more appropriate for our assembly.