actually we have done an experiment for finding deflections in a frame in which the connections of beam and column are not rigid. and also the results of deflections are lies in between of rigid connection and hinge connection.an image attached with question which explains the connection.and also please help me to model this connection in abaqus.
You can model it as beams and define the proper sections but in this case you won't be able to correctly model the joints.
So if you have enough ram to model the whole system by modeling each part as a separate solid and to model the contacts, this will be the best choice.
So create all the parts separately, assemble the parts, model the contacts (interactions), apply the loads (also the bolt pre-load) and run the simulation
Franco
An other similar approach, on top of what Franco suggested above, is to use beam elements for most of the frame, but use solid representation of the connection. This means that your beam elements should be shorter that the length of each member and connected to the solid parts using a coupling. This approach will give you the best of both worlds: accuracy, time/solution efficiency.
I would recommend shells instead of solids in the end regions. Much more computationally efficient. I agree with George's suggestion to use beam elements for most of the frame.
This will be true for 99.9% of the FE models you ever build: the smaller your model is, the easier to run, the easier to debug, and the easier to check.
If your problem extends to a larger framework beyond just this portal, try building a shell model with contact of each connection type, loading it to determine translational (very high, could be rigid) and rotational stiffness, and define a single connector element that captures the joint behaviour, then compare it back to your 3D analysis of the connection.
- Badges
- Science topic