You would generally model both as pinned conditions; however, just because you model them that way doesn't mean that's the way they react. For either condition, you can create a force couple between the bolts, thus induced moment in the angle.
What is the error for assuming one condition or the other? Run them both way and you'll see the differences. If you are looking at bending, a pinned condition will give higher midspan bending moment than fixed condition. If you're looking at sizing the angle based on bending, then that's more conservative. If you are looking at stress concentrations around the bolt/connection area, then a fixed condition might give you a more conservative result there.
In reality, the actual answer is somewhere between. There will be some moment developed, but probably not enough to reduce the midspan moment appreciably.
Sorry, I forgot to mention that these angles are purely compression members. I would like to know what "K" value is considered the right value among most of you structural eng. guys. K=1.0(pinned)? or K=0.65(fixed) or some value between 0.65 to 1.0?
Ron,
I'd be suspicious of z-axis eccentricity in Figure 2. Unless you had slip-cirtical bolts, there is the potential that some rotation of the angle could be allowed.