Hello Stressers,

Just want to share my thoughts on this topic.

One of our client have added a requirement in PSV systems just recently. The old spec doesn't show this.

"Relief in closed systems shall consider the force due to change of momentum at every elbow up to the main header. Static analysis shall consider the force in just one elbow at a time. In lieu of static analysis, a time history analysis may be considered."

𝐹=𝜌(𝑉^2)(𝐴)

If I'm not mistaken, the formula above is due to drag force created during the high velocity flow (in a turbulent regime).

But I think, this one is still subject for discussion as most of procedure established in the majority of EPC company's is considering that in the steady state condition the forces are balanced in a closed system. In my previous experience, we can consider (pop-up condition and steady state condition).For closed system, we put the force at the elbow nearest to the header (with the assumption that the header is a large volume and can be considered an open boundary). So basically, in this manner the treatment is almost the same for both open and closed system.

Anyway, just sharing what I have experienced so far. This requirement becomes additional man-hour in our part because it was not included in the old spec. Even in my 1st EPC company I have read this in the procedure: "Only the reaction force of a relief system with the tail pipe open to the atmosphere needs
to be considered. The reaction of a closed system can be ignored."

What I can see in this case is during those days the procedure are simplified as long as good design practice is followed. But nowadays with the availability of powerful computer software in fluid simulation, the physics behind all this can now be coded in these software. Integrating the mechanical and fluid simulation software is even possible now. But in a typical EPC set-up it is the process engineer who do the fluid simulation and the stress engineer who do the mechanical simulation. So as a stress engineer it would be best for us also to know some basic physics in fluid mechanics/dynamics related to fluid flows so we can effectively communicate with the process engineer what we need from them.

Any other opinion is greatly appreciated.

Cheers!!!