Learning Forum

[dawidmezy]

I’m working on a bike frame model for FEM analysis in Ansys and I have a 3D model ready.
How does symmetry work in ANSYS? When I divide an object in half, do I need to apply boundary conditions, such as loads, at half the value? Specifically, does ANSYS consider the applied force as acting on the entire model, even though the mesh only represents half? I'm concerned that I'm getting results that are twice as large as expected. Could I be missing something? Thanks!

 

[peteroznewman]

If you take a 3d surface model of the tubes that make up the frame, and you slice it in the center plane and support the cut edges in that plane with a Symmetry object and region in Mechanical, then you would apply only half the force because you have only half a frame.

Pinned is not the correct boundary condtion for a frame on wheels since the wheel can roll as the frame is loaded at the seat.  The correct BC to apply at the rear axle is simply Y=0 leaving X free assuming the XY plane is the plane of symmetry.

The force going into the seat would more correctly be modeled as a remote force scoped to some elements at the top of the seat tube at the height of the seat above the frame. The force would have the weight of the rider with a negative sign in the Y component and there would be no X component if the rider is just sitting on the seat with no weight on the pedals.

Another load case is with the rider standing on the two pedals equally. That would be a Y component applied at the pedal axle.

A more severe load case is with the rider applying force to just one pedal. This cannot be done on a symmetry model becuase the load is not symmetric. When the rider pushes down on the right side pedal, they push down on the left side handle bar while pulling up on the right side handle bar. You don't need the handlebars to apply this load, you can have the front tube fixed. The load on one pedal has the effect of twisting the 3D frame.

[dawidmezy]

Thank you very much!
I'm working on a surface model of a frame, and everything is going well. However, I'm thinking about the mesh and want to use squares instead of triangles for better results and mesh quality. Could you give me some advice?

[peteroznewman]

The mesher will automatically maximize quad elements and use triangles only where it needs them.

[dawidmezy]

One more question, where I indicated - can it be a fixed support, or is it better to have e.g. cylindrical support?

[peteroznewman]

A better support for the front tube of the frame is a Remote Displacement scoped to the short tube. In the Details window, there are some global coordinates that represent the center of that short tube.  Edit the numbers so that the X and Y values are the global coodinates of the front wheel axle. That is some distance in front of the short tube along the X axis and has the same Y coordinate as the rear axle. For the 6 constraints, if you are working on a symmetry model, 3 Degrees of Freedom (DOF) have been supported by symmetry. The rear axle supports 1 DOF (Y=0) so the front axle Remote Displacement only needs X=0 and Y=0 and leave the others Free.  This will allow the front tube to rotate about the front axle as the load is applied to the seat or the pedal axle.

If you are doing a 3D model without symmetry, then you need that Remote Displacement to also have Z=0, Rx = 0 and Ry = 0 in addition to what I said above.

[dawidmezy]

...

I'm trying to make a carbon fiber composite for a bike frame, but I'm having trouble with the XYZ orientation.? Can you help me?

[dawidmezy]

Thank you for your time!

[peteroznewman]

Is the carbon fiber a fabric or a unidirectional layup?  If it is unidirectional, how many layers are there and what is the angle of each layer relative to a 0 degree reference direction?  There is a software environment called ACP (Pre) in the Components section of the Workbench toolbox where you define the layup and angles.

[dawidmezy]

I know, that ansys have a composite material but for the frame a need use ACP to configure the thickness - layers. I have a problem with defining the coordinate system for each frame element (pipes). Because I can't give a common system for the whole frame?