What is the benefits and limitations of using node to node contact (CONTA178)?
Sometimes, people get different results when using CONTA178 vs. CONTA174. Here I explained the cause of the difference. Feel free to comment if you have any questions.
Node-to-node contact (CONTA178) has a big limitation in the fact that contact is only detected between two nodes I and J of CONTA178. Thus, one CONTA178 can’t interact with another CONTA178 – only the two nodes of a single CONTA178 can interact with each other. Thus, for any situation where you expect sliding, you shouldn’t use CONTA178. (For CONTA174, one CONTA174 can interact with any TARGE170 in the same contact pair, so that is why it is more flexible. One way to think of it is that the “node I” of CONTA178 is contact, “node J” of CONTA178 is target, so it’s like we separate CONTA178 into two groups of elements CONTA174 and TARGE170 if we were to compare the two approaches.)
This begs the question, if finite sliding is not supported, why do we have CONTA178? There are many situations we can imagine where two parts are in contact (not bonded – one surface can lift off from the other surface) without finite sliding. CONTA178 requires a similar mesh pattern between the two surfaces, but it is extremely computationally efficient. There is no algorithm requiring searching for which contact detection points may come into contact with which target surfaces, computation of contact ‘normal’ and ‘tangential’ directions, smoothing computations, etc. Thus, if you don’t have large sliding and have the same mesh pattern between two surfaces, then CONTA178 is very attractive due to its computational efficiency. However, because we do need the same mesh pattern, that is why you don’t see CONTA178 much in practical usage (and explains why it’s not supported in Mechanical directly).