Simulate Early, Simulate Often... In Rhino
I´m evaluating SnS so I tested it on a hollow beam of 12 meters long, 0,45 m wide, 0,45 m deep and 10 mm thick. It is under uniform load and must be simply suported at the edges.
I performed analytical hand calculation and the deflection is around 70 mm at the centre.
Running SnS I tested X, Y and Z restrains in different ways at one the edges and just restricted in the load direction (Z), on the other edge. The deflection obtained varied from 30 mm to 15 mm aprox.
The deformed shape of the beam always showed the max deflection point slightly moved from the centre and the deformed beam slope was different from one edge to the other.
It seemed that some kind of clamping was happening on the edge with X,Y and Z restrictions.
Seems that the necessary X restriction applied on one surface at the edge prevents edge section from rotating as it should if simply supported.
Sorry the picture should be upside down
Is there any nice way to simulate a simple support on the edges of the beam?Thanks
It looks like you are very close. You just have to re-arrange your assignment of boundary conditions a little.
At each end of the beam, restrain the end face in Z and X. This will allow small rotations of these faces about the X-axis, simulating a "roller" support at each end.
For numerical stability, you still need to restrain the beam along the Y-axis. You can do this by applying a Y-restraint to a small horizontal face at one end as shown below. This will correspond to a "pinned" restraint at this end of the beam. The purple wedge in the figure represents the pinned support.
Does this help?