Unrealistic mass flow in an HVAC duct when transporting smoke from one space to another

[dev-ltk]

The model is shown in the image below and contains two spaces, i.e. one 'smoke generation space' (at the bottom of the image) and one 'discharge space' (on the top of the image). The smoke is extracted from the 'smoke generation space' and discharged to the 'discharge space' via an HVAC duct in FDS with a constant flowrate (200m3/s). Two makeup vents (100m3/s each) is provided at two sides of the 'smoke generation space'. The four sides and the top of the 'discharge space' are set to open boundaries. The fire is set to ramp its peak quickly from the start of the simulation so as the HVAC duct volume flowrate.

The HVAC duct quantities are logged. It is observed that the duct volume flow and temperature reach steady state quickly after the simulation starts. However, the mass flow and density inside the duct are monotonically increasing. The density is exceeding 2.5kg/m3 at the end of the simulation. It does not seem realistic.

Is it the intended behaviour of an HVAC duct in FDS? Or does it exceed the capacity of an HVAC duct in FDS?

The FDS file is attached.
hvac_setup_test2.txt

[drjfloyd]

Think carefully about your inputs. You are supplying 200 m3/s of ambient air at ambient temperature and removing 200 m3/s of hot air. Are these mass flows equal?

[dev-ltk]

It is true that the density should increase with this setup. Given that the fire ramp up instantaneously at the start of the simulation, when the temperature in the duct is almost steady at the later phase of the simulation, it is expected that the density and mass flow are steady too. However, the density and mass flow are still increasing.

[drjfloyd]

You are using a SURF for the inlet and not HVAC. The SURF boundary condition uses for its density the density of the gas phase cells adjacent to the VENT, not the ambient density.

[dev-ltk]

Could you elaborate more on "You are using a SURF for the inlet and not HVAC."? Do you mean the "Makeup " SURF in the FDS file?

[drjfloyd]

Yes. Your inlet in the lower MESH is a SURF with a fixed VOLUME_FLOW. The defalt temperature for a SURF is the ambient temperature. The density is taken as the gas density adjacent to each wall cell.

[dev-ltk]

How is this related to the mass flow and density recorded in the duct in the simulation?

Meanwhile, the setup of the lower mesh is not very good. It makes more sense to switch the inlets in the lower mesh to 'OPEN'. However, the initiative of this setup to later apply wind to the upper mesh. If these inlets in the lower mesh are set to 'OPEN', the wind will also disrupt the flow there. Is there any better boundary condition can be considered?

[drjfloyd]

The solution is to not fully specify both supply and exhaust as the same volume flow. This is non-physical in terms of how a real world HVAC system would operate. You also have no leakage in the closed space. Unless this is something like the inside of a reactor containment building, there will be leakage. If you want the supply to be ambient air at ambient density you should use HVAC for the supply. You should add some leakage.