Generally every established MD engine is sufficiant for simulating polymers, especially when realizing that proteins or DNA are polymers, too.
Away from that, what is more important in case of polymers without extisting forcefields or parameters is the respective toolchain to parameterize new molecules. Here, e.g. CgenFF + VMD + NAMD or Gaussian + AmberTools(GAFF2) + NAMD/SANDER migth be reasonable choices.
As mentioned before, many MD programs can be useful in such calculations (it depends on how accurate you're searching to be).
However, for a relatively friendly program, i recommand material studio. It contains a wide range of fields and options and it's easy to use. I already used it for long chain polymers (check my poster DOI: 10.13140/RG.2.2.27410.45760 ) and for starch (data not available for the moment). The results were very close to experimental data.