After seeing the error, I tried to solve this problem by changing the type of mesh and type of contact algorithm as well. However, the error still remains.
Actually, my impact model consists of two plate (5 mm in thickness) modeled in shell elements, I put them together at their mid-surfaces of each other, this looks like the nodes of two plates are coincident. So, I tried to separate them with a distance of 0.01 mm, then the error disappeared. However, the warnings occur:
***WARNING: In step 1, portions of master surfaces in the general contact
domain have been tied together. Joining disconnected surfaces with *TIE does not alter the surface connectivity and results in a seam in the contact surface. The nodes along the tied surface perimeters have been added to the node set named
"WarnNodePerimTieSeam" in Abaqus/Viewer.
***WARNING: In step 1, some facet thicknesses for general contact were reduced
from the parent element or specified values due to a large thickness compared to facet dimensions. The most significant thickness scaling factor was 5.39498E-02 for a facet on parent element 3746 of instance T-BEAM-1. An element set named
"WarnElemGContThickReduce" has been created for use in ABAQUS/Viewer to locate the regions of reduced thickness.
Lastly, I run the analysis and then I have the results, but the internal energy and kinetic energy are not balance. In particular, the internal energy was increased until end of step, while kinetic energy shown a good result.
Could you explain the problem and give me suggestions?
Since the tie constraints were used to connect two plates instead of bolt connection, so I hope to see the deformed of each plate at other positions, where is out of tie location. Therefore, merging the geometry may be not considered. So could you give me more details of other ways to solve it?