I am currently simulating the forces applied to a POM gear with its steel axis. The torque of the axis is transmitted by means of a flat surface.
My study:
The simulation is an assembly of two parts: the gear and the axis
The gear is fixed with a fixed displacement imposed on the outside of the gear.
An axis connector is placed between the two cylindrical surfaces (Axis + Gear) with connection type: circlip and key.
A torque of 0.2 N.m is applied to the axis.
The aim of this study is to strengthen the plastic gear due to a breakage in the flat area.
My problem:
As shown on the simulation, very few stresses are at the level of the flat spot. I think it's because of the connector and the type of connection "key" which doesn't know the location of the latter. In addition, as for this connector you can only select cylindrical surfaces, I could not select the flat area, hence the low constraints on it.
To solve this problem, I disabled the "key" option to replace it with a plane press move but the simulation fails each time.
If you have a solution to correctly simulate this flatness, I'm all for it.
Personally, I would start by slightly modifying the flat. I suggest you create a radius on the two edges located at the junction of the axis and the flat. In your current setup, you must have a concentration of stress on a point (the length of the edge) which will cause a tear of the axis towards the outer diameter. In addition, the simulation is done in static, dynamic it would be worse because depending on the load and the speed of acceleration the data are very different. It's even worse if you have instantaneous changes of direction at high speed.
Can you elaborate: a) how is the axle fitted to the wheel (hot, glued, tightly fitted, with what tolerances)
b) max temperature of the assembly in operation because the expansion coeff is not the same between two heterogeneous materials. The steel expands more than the POM impeller, which can create significant stress. The wheel bore is solid (without mini spline) which means that nothing absorbs the expansion. (c) how many changes of direction per minute and what speed of acceleration and deceleration for stopping or changing direction.
Looking at the previous ruined piece should confirm my point. We see the beginning of stress concentration on your simulation, it would be worse with the parameters I suggest below!
For the configuration part, I suggest you do it differently at first to not take into account the leader and the leader.
Put a fixed connection on the face of the end of the axle (do not put it on the flat as you will distort the result very slightly). Apply the force on one or two teeth of the gear by delimiting the fulcrum with a sketch made with a dividing line (normally it's a contact line on a tooth but we're not going to quibble since it's the axis junction that is the problem).
Thank you for your answer. Indeed the flat breaks well at this point but the purpose of this simulation is to compare the improvement of a possible modification. Even if the results are not perfect, a Before/After change comparison is sufficient.
Your solution of switching to a single part works much better, the simulation is much more consistent with reality.
