Spring idealisation and contact interface

A contact interface must be between surfaces.  I suggest a washer or disk part rigidly connected to the end of the idealized spring as representing the spring contact interface.  On the other hand since you say there are a number of these connections, maybe you can use non-linear spring definition instead because contacts will be much more difficult to run successfully.  I don't have time at the moment to mock up this type of model and prove what works but these are my initial suggestions.  I have not done any non-linear springs that I recall so I am not sure how to help that part be successful.  Use a low stiffness for extension lengths and normal spring stiffness for compression lengths.  You may need to keep the spring curve a smooth function near the break between T/C.  Hopefully someone in the community can enlighten on the possibility of using non-linear springs to model contact when you do not care about local contact areas/pressures.  Also keep in mind that point connections such as to a spring will generate very large stress so you may need to rigidize a small area right where the spring attaches or use weighted links to spread the load over a real area (the reason simply is a divide by zero problem.  Force/Area if A is zero = undefined/infinite stress)

good luck and welcome to the community!