Due to their unique
structural and mechanical properties, randomly
cross-linked polymer networks play an important role in many different
fields, ranging from cellular biology to industrial processes. In
order to elucidate how these properties are controlled by the physical
details of the network (e.g., chain-length and end-to-end distributions),
we generate disordered phantom networks with different cross-linker
concentrations
C
and initial densities ρ
init
and evaluate their elastic properties. We find that the
shear modulus computed at the same strand concentration for networks
with the same
C
, which determines the number of chains
and the chain-length distribution, depends strongly on the preparation
protocol of the network, here controlled by ρ
init
. We rationalize this dependence by employing a generic stress–strain
relation for polymer networks that does not rely on the specific form
of the polymer end-to-end distance distribution. We find that the
shear modulus of the networks is a nonmonotonic function of the density
of elastically active strands, and that this behavior has a purely
entropic origin. Our results show that if short chains are abundant,
as it is always the case for randomly cross-linked polymer networks,
the knowledge of the exact chain conformation distribution is essential
for correctly predicting the elastic properties. Finally, we apply
our theoretical approach to literature experimental data, qualitatively
confirming our interpretations.