Fan curve exceeding stall pressure or maximum volume flow rate

[Jack]

Hi all,

I am simulating a compartment fire with a mechanical ventilation system. I set fans in the ventilation ducts and the fan curve is a user defined curve. Due to the high pressure in the compartment during the fire, a reverse flow in the inlet duct and a enhanced flow in the outlet duct were obtained. This means: the pressure exceeded the stall pressure of the fan in the inlet duct and the volume flow rate exceeded the maximum volume flow rate of the fan in the outlet duct. In the FDS user guide, it says if the pressure difference increases beyond MAX_PRESSURE, air will be forced backwards through the fan, and if the downstream pressure becomes negative, then the volume flow through the fan will increase beyond MAX_FLOW. In my case, the downstream pressure doesn't become negative, just the pressure difference (downstream-upstream) became negative. I think the effect on the fan curve is the same, right? More importantly, I want to know how does FDS deal with the fan curve when the pressure exceeds stall pressure or pressure difference becomes negative? I can only get the fan curve between 0 Pa and stall pressure from the manufacturer manual , see figure below. And I want to know what the fan curve is in FDS if the pressure is outside this range.

&HVAC ID='Fan01', TYPE_ID='FAN', RAMP_ID='fan1 curve'/

&RAMP ID='fan1 curve', T=0,F=145/

&RAMP ID='fan1 curve', T=0.00278,F=134/

&RAMP ID='fan1 curve', T=0.00556,F=127/

&RAMP ID='fan1 curve', T=0.00833,F=120/

&RAMP ID='fan1 curve', T=0.01111,F=114/

&RAMP ID='fan1 curve', T=0.01389,F=107/

&RAMP ID='fan1 curve', T=0.01667,F=101/

&RAMP ID='fan1 curve', T=0.01944,F=96/

&RAMP ID='fan1 curve', T=0.02222,F=92/

&RAMP ID='fan1 curve', T=0.02778,F=83/

&RAMP ID='fan1 curve', T=0.03333,F=75/

&RAMP ID='fan1 curve', T=0.04583,F=58/

&RAMP ID='fan1 curve', T=0.05000,F=50/

&RAMP ID='fan1 curve', T=0.05556,F=37/

&RAMP ID='fan1 curve', T=0.06111,F=20/

&RAMP ID='fan1 curve', T=0.06667,F=0/

Many thanks in advance!

Jack.

[dr_jfloyd]

See Eq 10.3 in the Tech Reference.  The fan pressure appears in the delta p_j term in the equation. It's value can range between 0 and the stall pressure depending on the current volume flow rate in the duct.  If, for example, you reach a point where the flow in the duct is at the maximum flow, then the fan will not add any additional pressure and the flow rate will solely depending upon the node pressures.  If on the other hand then flow rate reaches zero, then the fan will add the maximum pressure to the node pressure difference. If the pressure difference winds up negative there will be reverse flow. 

[Jack]

Hi Jason,

Thanks for your reply. See figure 9.5 in the user guide, when the volume flow rate becomes negative or exceeds the maximum flow rate, there is still a fan curve. Does this mean the fan still having impact on the duct flow? And how do these extended fan curves determined?

Besides, there is a item called LOSS in HVAC fan parameters. What does this parameter refer to? Does this mean that the fan just act as a resistance in the duct when the flow rate becomes negative or exceeds maximum rate? And how to calculate the loss? I didn't find this parameter in the manufacturer manual. 

Thanks!

在 2018年9月3日星期一 UTC+2下午10:39:45，dr_jfloyd写道：

[Jack]

Is it possible to export the output of the fan curve during the simulation, which could be compared with the input fan curve?

在 2018年9月3日星期一 UTC+2下午10:39:45，dr_jfloyd写道：

See Eq 10.3 in the Tech Reference.  The fan pressure appears in the delta p_j term in the equation. It's value can range between 0 and the stall pressure depending on the current volume flow rate in the duct.  If, for example, you reach a point where the flow in the duct is at the maximum flow, then the fan will not add any additional pressure and the flow rate will solely depending upon the node pressures.  If on the other hand then flow rate reaches zero, then the fan will add the maximum pressure to the node pressure difference. If the pressure difference winds up negative there will be reverse flow. 

[dr_jfloyd]

Figure 12.5 plots the quadratic curve,  the custom curve as defined in the example, and a constant flow curve. These are all made up inputs and did not come from any specific manufacturer.  Manufacturers do not typically measure fan performance curves outside the range of 0 flow (stall pressure) to max rated flow.  The fan blades, fan housing, etc. do impose flow losses. These are already accounted for in the fan curve. If you need to develop a fan curve beyond these points extrapolating assuming quadratic behavior at the ends of the curves should give you reasonable behavior. Unless you have a highly sealed room, typical leakage areas should limit the ability to drastically exceed the limits of the fan curve.

There are no outputs for the fan curve. 

[Jack]

Many thanks, Jason!

在 2018年9月4日星期二 UTC+2下午3:21:58，dr_jfloyd写道：

[Jack]

Hi, Jason. I want to double check about the 'LOSS' setting. Does it mean that we don't need to designate a value for loss if the losses are already accounted for in the fan curve? In the user guide(9.2.4 HVAC Fan Parameters), the loss is designated with the value of 2, how does this value come? Besides, when I change the value (for example setting LOSS=15, LOSS=20), there is no influence on the duct volume flow rate. So, I want to know when will this parameter be used? Thanks!

在 2018年9月4日星期二 UTC+2下午3:21:58，dr_jfloyd写道：

Figure 12.5 plots the quadratic curve,  the custom curve as defined in the example, and a constant flow curve. These are all made up inputs and did not come from any specific manufacturer.  Manufacturers do not typically measure fan performance curves outside the range of 0 flow (stall pressure) to max rated flow.  The fan blades, fan housing, etc. do impose flow losses. These are already accounted for in the fan curve. If you need to develop a fan curve beyond these points extrapolating assuming quadratic behavior at the ends of the curves should give you reasonable behavior. Unless you have a highly sealed room, typical leakage areas should limit the ability to drastically exceed the limits of the fan curve.

[dr_jfloyd]

The value of 2 is just a made up value.

As defined in 12.2.4 HVAC Fan Parameters (you appear to be using an old version of FDS, HVAC is now in Chapter 12):

LOSS is the loss coefficient for flow through the fan when it is not operational.

This means that if the fan is running, the value in LOSS as an input for a FAN isn't used. When the fan is running, the loss coefficient is built into the fan curve. The fan curve is established by measuring the pressure drop across  the fan as a function of the flow rate as one throttles the area for the inlet or outlet of the fan. That pressure drop measurement includes the any losses through the fan fittings and the internals of the fan. When the fan is off; however, we don't have data for that in the fan performance curve and if you want to truly model the flow in the system you would need to estimate the losses of the fan when it is off.

[Jack]

Thanks, Jason. In Chapter 20 (20.14 Summary of HVAC Output Quantities), there is an output quantity named 'FAN PRESSURE'. Is this the output of fan curve? When I set &DEVC ID='Fan01Pressure', QUANTITY='FAN PRESSURE', FAN_ID='Fan01'/ in the input file, there is an error occurred: ERROR: Problem with DEVC number 51, line number 74. What is the problem? And how to get this output?

在 2018年9月6日星期四 UTC+2下午6:34:43，dr_jfloyd写道：

[dr_jfloyd]

This is the output of the pressure the fan is adding to the flow. It is not going to generate the entire fan curve.

Try using DUCT_ID.

[Jack]

Thanks Jason. Problem solved!

在 2018年9月7日星期五 UTC+2下午12:06:00，dr_jfloyd写道：