1. The problem statement, all variables and given/known data For a larger project we had to redesign a truss to prevent certain members going beyond yield stress. I've found that due to the nature of the reaction forces and the supports used, the only way to do this is to introduce an X-brace in the bay nearest the supports (as well as some other changes that prevent zero force members from occurring in that bay). It's the only logical thing that will work. Otherwise all of the horizontal reaction force at one of the supports will go into the tension or compression of a single member and that will be beyond the yield stress. However, when I do the hand calculations for the truss now, when I get to the last bay with the X-brace, it becomes indeterminate. The truss simulator doesn't mind, it simulates it just fine, no singular matrix errors. But I can't figure out how to analyse it. I've tried running with simplifying assumptions about making certain member stresses equal to each other, I've tried cutting the bay through the half way mark and applying method of sections, etc. Everything I've done results in singular matrices. Here's the system: imgur link: http://i.imgur.com/5WAHblS.jpg red arrows are known forces, green unknown. Those little red arrows along the members are member self-weights. Fd is not a reaction force at B, it's part of a distributed load along the top of the truss. 2. Relevant equations Sum of moments to zero Sum of forces to zero 3. The attempt at a solution So many. But this is a first year mechanics/statics unit, and we're not supposed to go accounting for displacement of joints or bending moments in our hand calculations. Am I totally stuffed? I just cannot see how else to get the truss into bounds.