Manvin,
Drej is 100% right. You can rely on linearity of your problem (if you are in linear case, as it seems) in order to compare the "trial" stress solution you get with your "trial" displacement/force/pressure with the "target" stress solution you want to compare with: simply, if trial stress is half the target, the load pattern that will give the target stress is twice the trial one. Of course due to calculation deviations, it's unlikely that you will get a stress pattern 100% compatible with your target one in every point: so, interpolate btw points to compare the stress patterns. Creating the "trial" load pattern requires knowing "how" the beam was loaded to obtain the stress pattern you want to re-create.
Claudio