Unfortunately I have 2019 R2, so I can't open that archive. However, I notice on the graph you posted that the displacement of P appears to be settling somewhere around 2e-2m. Since the joint displacement is 5e-2, point P should settle to 5e-2. I think this is due to the joint not being able to displace freely for some reason. Instead, the blocks are pushed apart by ~0.5x the displacement.nHowever, I found a solution that worked quite well for me. I added a bar between blocks 1 and 2 with an orthotropic secant coeficient of thermal expansion in the X direction equal to 1/L where L is the original length of the beam. I also gave this part a very high stiffness relative to the other parts. I then set the reference temperature by body for this part to 0 C. This means that each degree of temperature change equals 1m of expansion (delL/L = alpha*delT --> delL = alpha*L*delT, alpha = 1/L --> delL = delT).nWith this method, I can assign a tabular temperature change to this part. Here is a gif of the response and below is a plot of the response. As you can see, point P (TCP) settles to the correct commanded value.nn