Hello Guys,
When I dealing with safety relief valve load case, I put concentrated force F1 at first elbow after safety valve. AM I should put an equal force F1 on the back of relief valve(at relief valve discharge flange) but in opposite direction ?

Generally, for static simulation of such dynamic event, the reaction force stated in PSV datasheet is multiplied by DLF of 2 and applied in opposite direction of the PSV outlet. Some companies also recommend to apply the force in upward direction (without any multiplication).

Thank you but can you more clarification, do you mean that I put the reaction force from data sheet in first elbow in only downward direction unless the company recommend to apply force in upward direction only in psv discharge flange?

For open relief systems, forces are applied at the PSV outlet and exit point to atmosphere (you may refer API RP 520 or ASME B31.1 ). Refer datasheet or PSV drawing. If not, then you can calculate based on fluid properties using API RP 520.

For close relief systems, generally these reaction forces are ignored for equilibrium condition (though this is a vast topic and can't be generalised).

so do you mean in open discharge system I should Put F1 in two nodes one at free end point with -y and another one at PSV discharge flange with +x if its direction in +x is that right?

In local, open discharge systems, the force is applied at the discharge of the piping. See API 520 diagrams. This simplified method represents the steady state condition, some time after the valve has opened, and assumes that the piping lengths are sufficiently short enough that the pressure wave cancels itself early enough to be insignificant by the time the final discharge occurs. ASME B31.1 has guidance regarding DLF here.

In closed discharge systems, a force will occur at every bend, but the magnitude of that force will be a fraction of the previously calculated force, which considers opening time for the valve, the speed of sound of the fluid, and the lengths between bends upstream of each bend. All forces are applied in the opposite direction of flow out of each element. (If the elbow is traveling +Y and turns +X, force is applied in the -X direction). Whether you apply the force to the "upstream" bend or the "downstream" bend is immaterial, as forces applied axially to a pipe in CAESAR will be transmitted perfectly through the "thin stick." This methodology is spelled out in L.C. Peng's "Pipe Stress Analysis." Within, he specifies considering only the elbows within total length of the PSV equal to speed of sound times valve opening time.

In remote, open discharge systems (quasi-open/closed), apply both logic: small forces on the bends, large force on the discharge. You should calculate pressure losses and velocity increases in the line for the discharge.

In all cases, the magnitude of the force is (more or less) mass flowrate times velocity plus pressure times area. Typically, this is taken at the valve as a matter of conservatism and simplicity.