HVAC Calculation

[ShehabGamrah]

With regards to HVAC calculation I have tested this setup of duct with exhaust surface in room at a set flow rate and got a pressure drop close to the hand calc based on the fds duct loss equation in user guide. I have rotated the duct from horizontal to vertical while keeping all parameters constant and found the pressure in the room to become much different than before. Is there a reason as the pressure losses should be the same regardless of duct orientation.

[mcgratta]

Off the top of my head, are you considering the effect of gravity? There is an additional term in the equation to account for a change in the elevation of the entry and exhaust vents.

[drjfloyd]

Looking at the logic for the various combinations of inputs (ROUND, SQUARE, AREA, PERIMETER, and DIAMETER), I think I over complicated it. Let me simplify it and see if that fixes this issue.

[ShehabGamrah]

Thanks. Please let me know the best way to input this to obtain an accurate hydraulic diameter. Also, if I specify a node as an ambient endpoint and apply a flow loss, what area would be used for the K-loss calculation? And, Can I specify a area for this node loss calculation?

[drjfloyd]

You can try the latest FDS test bundle (https://github.com/firemodels/test_bundles). Read the User's Guide section on HVAC Ducts in the bundle for details on the updates for ducts.

Node losses are based on the velocity in the attatched duct. The User's Guide has details on how to scale losses if you need to account for a different real world area vs. modeled area.

[ShehabGamrah]

Thanks. Just to clarify, does this mean that in the previous releases, the rectangular or circular diameter was not represented accurately, whereas this approach treats a rectangle using its hydraulic diameter and a circular duct using its actual diameter?

Secondly, in the context of HVAC, I would like to replicate a shaft with an opening at the head and an opening into the corridor through the HVAC system, so that I can control the friction loss. What would be the most accurate representation?

I know that for the damper/opening at the head, the inlet loss is K=2.37, and for the exit into the corridor, the loss is also K=2.37. The shaft itself has frictional losses only, with no other losses.

I attempted to represent the head as an ambient endpoint, and also as a node connected to a vent with the vent connected to the external atmosphere, but I obtained different results. Could you please advise on the most suitable way to represent this in FDS?

[drjfloyd]

Possibly not correct before. You had reported that you felt it was not using the correct diameter. The code for that part had become convoluted over time. I simplified it down, and it should now be correct in using the hydraulic diameter for a rectangular duct for friction losses if it wasn't before.

An AMBIENT node only every sees the hydrostatic pressure at the node height. Since there is no CFD calculation around an AMBIENT node, there is no velocity field to perturb the hydrostatic pressure. A node connected to FDS domain sees the hydrostatic pressure plus pressure perturbartion due to the velocity field. This will result in some differences which could be signficant depending on the local flow field that develops (you also should have enough domain around the exit that you are not overly perturbing the pressure field by the choice of domain).

The most accurate representation would be to model the shaft with sufficient resolution and domain exent at the outlet to model the pressure losses directly. This of course may not be an affordable calclulation. The HVAC model has many simplifying assumptions in it and requires you to specify losses. Your knowledge of those losses is likely incomplete. You say that you know the losses are 2.37, but that is based on some specific test condition(s) and test setup and the real loss for your specific case could vary somewhat.