I'm doing some packing simulations on the Aspherix Solver/GUI and am running in to a puzzling problem. I have normally been modelling complex particles as Concave shapes, but have recently moved to modelling 'simpler' cylindrical, Convex, pellets.
The particles I am using are essentially the same - effectively comparing a cylinder (Convex) against the same cylinder with various sections cut out (Concave). The Concave shape is decomposed into 34 separate elements.
My understanding was that Convex shapes were computationally simpler than Concave shapes, and thus I'd expect my simulation to run quicker, but I'm seeing the exact opposite. The whole, Convex cylinder is running much, much slower than the Concave counterpart. Does anyone have any thoughts on why this might be the case?