The computational models used for compression members in the non-linea
r linear analysis of space trusses are classified by decreasing order
of computational effort into finite, segment, finite-difference, and s
ingle-member solutions. The common assumptions and derivations used wi
th single-member solutions are reviewed. A chronological re view of th
e models is given indicating the contributions each has made. Single-m
ember solutions can often give results similar to those that use more
computationally demanding methods. However, the success of a model at
correlating with experimental behavior depends on the assumptions made
. Single-member solutions that include partial plasticity appear more
accurate than those that assume full plasticity at midlength. It is ob
served that consideration also needs to be given to the relationship b
etween the member and the structure to which the member is attached. T
hus, internal snap-through or dynamic jump can occur in the postbuckli
ng regime of compression members in a relatively compliant adjacent st
ructure.