How to Simulate "Discrete Rigid" Support?

Hi all, I am modelling a 4-point bending test for a composite material called Timber-Concrete Composite. I have simulated the test and obtained some results. Since the results have more than 10% error from experimental result, I tried to vary the boundary conditions for the model. I created "discrete rigid" cylindircal parts as support and loading for the bending test. 

However, after I have made these changes, I receiving warnings that said "displacement increment for contact is too big" and convergence issues arose. The model was immediately aborted due to "time increment required is less than the minimum specified" which is of size 1e-8.

Could someone give me some help and help to diagnoise the problem behind. Any kind of help would be greatly appreciated. I will give more detail about the problem I am facing below.

4-point Bending Model:

I was initially applying simple support condition which restricts translastional displacement and vertical displacement loading (-100mm) directly on the beam member. I would partition the member and then apply the B.C. on the partition lines. Since I am simulating half of the beam, plan of symmetry B.C. was applied to the right end of the beam. The interaction for the interface of the composite employed "hard contact" for normal contact and a 0.57 frictional coefficient for tangentail contact.

I later thought applying B.C. directly on the partition lines of the beam might affect results due to stress concentration, so I tried to create "discrete rigid" parts to simulate the support and loading. I followed tutorial I found on YouTube to create the "discrete rigid" parts and then create general contact interaction that excludes the interface of the composite to define contact property between the beam and the supports and the loading tool. "Hard contact" and a frictional coefficient of 0.2 were defined for the general contact. With other B.C. remained the same, I defined "encastre" at the face of the support cylinder to fix it and the same vertical displacement loading while keeping all other DOF fixed for the loading cylinder.

After these modificaiton I then submitted the job and received the aforementioned warnings. The modell was aborted due to "time increment required is less than the minimum specified".