Please see the snapshot below, could you please assist to clarify why there are discrepancies in deflection between the element beam and the shell beam.
Did you ensure mesh convergence? Those quad4 shells can be quite poorly behaved sometimes and may require a lot of refinement. It's usually better to use quad8.
The load is only 4.0 kN on the shell because the value you specify in a force item is its total force, not force per node. If it's applied to multiple nodes, it gets distributed equally between them.
You can also save some effort manually specifying the special end node forces if you use the edge "face" instead of nodes, as shown in this attachment.
Comments
Did you ensure mesh convergence? Those quad4 shells can be quite poorly behaved sometimes and may require a lot of refinement. It's usually better to use quad8.
Otherwise, could you upload the .liml file?
Kindly note that I have changed the elements to quad8 and I am still getting the same deflection on the shell beam.
Refer to the attached file.
Best Regards
You can also save some effort manually specifying the special end node forces if you use the edge "face" instead of nodes, as shown in this attachment.