how to model 3 point bending test on Ansys

[aizatul_ain]

hello and good day,
I would like to know how to positioned the constrain correctly on 3-point bending test on Ansys APDL. in order to mimic exactly the 3 point bending the i have to constrain the bottom end of my model. but Ansys keeps selecting automatically the middle line instead.

And what is the exactly constrains that I should apply. I have tried to constrain all DOF for both ends and partially constrain in Ux& Uy at first end and Uy for second end. unfortunately, both gives me absurd result in terms of displacement vector sum.

my model is in 3D.

can someone suggest more correct technique to done it?

Thank you.

 

[rb1957]

by constraining all freedoms (that'd be 12, at two nodes) you are creating a double cantilever.

At a minimum release the bending freedom that its reacting to your load; this'll reduce your constraints to 10 (still overconstrained).

The best solution would have only 6 constraints (to take out rigid body motion). Maybe pin one end, and at the other end 2 translations and axial rotation (torque along the beam).

Then check the reactions to ensure they are what you expect.

another day in paradise, or is paradise one day closer ?

 

[rb1957]

why is your model in 3D (I assume that means 3D elements). this should be something that can be modelled with 1D elements ? (unless it's an RC beam).

But if you have multiple nodes at each end, then either
1) use an RBE "spider" element to join all the nodes to a single node (at the centroid would be best) then follow the previous post, or
2) let each node react the applied shear, and constrain a limited set of nodes in the other freedoms, you need 4 more constraints. If the centroids of the two ends are not aligned then you'll have kick reactions .

another day in paradise, or is paradise one day closer ?

 

[mronlinetutor]

you have both ends of the beam constrained so it cannot move in the y-direction. even though you have release the second end in the x-direction, the beam is still fixed in place. it is a pin pin connection. so it cannot moved at all and the internal stresses has built up to give you the wrong result. rb1957 is right. you should have a roller at the second end instead. you should note that there is a different between beam end release and release of the degree of freedom. beam end release only releases the beam in bending but release of degree of freedom allows you to implement support like rollers. since it's a 3 points bending, that means another roller is needed between the front and end supports. you should model the beam as two beams where the end of the first beam and the front of the second beam are not released but restrained in the y-direction.

disclaimer: all calculations and comments must be checked by senior engineers before they are taken to be acceptable.