At first, i want to make a thermal-expansion simulation based on the mass center of the model, so i need to minus the displacement of the mass center.So I use the history output to give the mass center displacement. But I found that when I use SC8R element to simulate a panel (edge is 100mm, and thickness is 2 mm, CTE is 1e-4, temperature load is 100K,and the boundary condition is on the angle not the center of the panel), the mass center displacement is 0.000250000011874363,0.000250000011874363,and -nan. Sure, the "nan" appeared. I repeat to submit, the result is -nan, 0.000250000011874363,1.02572344132354e-14.They are not the same and obviously it is wrong.However, the displacement field is right. But when I change it into the C3D8R element, all is ok, the mass center displacement is 0.000500000023748726,0.000500000023748726, and9.99999974737875e-06. They are right. So what is the problem of SC8R element mass center output? Thanks for help