FDS+EVAC Delay at door connection compared to an opening with the same size

75 views

Skip to first unread message

Spitzel

unread,

Feb 25, 2020, 2:09:34 PM2/25/20

to FDS and Smokeview Discussions

Servus,

in a couple of FDS+EVAC projects it appeared that the delay in front of a door seems to be larger than in front of a hole or an opening with the same size. This appeared especially when comparing different door widhts. Also in front of stairs (evss) this flow phenomena is quite obvious.

To break this topic down to its core problem, I made some test simulations with the following scenario.

Room 1: A = 10 m x 5 m

Door or opening to connect the rooms: b = 1,20 m

Room 2: A = 5 m x 5 m

EXIT: b = 1,20 m

(see also the attached files)

I simulated two set-ups:

1) Opening (hole) to connect the two rooms.

2) Door-Door connection to connect the two Rooms.

I varieed the number of People in steps of 50 from 50 to 200.

I made 10 monte-carlo-simulations for each variation.

The result is as follows:

The egress time with 50 people is "almost" the same. With an increasing number of people the egress time also rises.

Sometimes problems like barricades or putting people into the other room appear, which leads to rised standard deviation (stabw in the graph).

Question: How is this delay defined? Is it possible to manipulate/set it? Sometimes it could be usefull, sometimes not. For example at a door connection on the same level the delay should (maybe) be zero, but at a door connection in front of stairs/EVSS it could model the real waiting time in front of stairs (somehow). It could also be a factor to simulate the time for someone to open a door etc. In any case it would be useful to set it manually. Is it possible?

I am very happy for every professional answer / hint.

Thank you very much.

Bastian

attached:

1) FDS+EVAC-File (opening):

2) FDS+EVAC-File (door connection):

FDS+EVAC-File (opening):

opening_200P.fds
Generated by PyroSim - Version 2019.3.1204
25.02.2020 14:10:53

&HEAD CHID='opening_200P'/
&TIME T_END=200.0/
&DUMP DT_RESTART=300.0, DT_SL3D=0.25/

&MISC EVACUATION_MC_MODE=.TRUE.
EVACUATION_DRILL=.TRUE./

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
AUSGABE    $$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

&SLCF PBZ=1.0, QUANTITY='VELOCITY', VECTOR=.TRUE., EVACUATION=.TRUE./

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
MESH $$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

&MESH IJK=198,54,1, XB=-10.0,9.8,0.0,5.4,0.9,1.1,
EVACUATION=.TRUE., EVAC_HUMANS=.TRUE., ID='EMesh_EG', EVAC_Z_OFFSET=1.0 /

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
PERSONEN $$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

&PERS ID='Adult',
      FYI='Male+Female diameter and velocity',
      DEFAULT_PROPERTIES='Adult',

      HUMAN_SMOKE_HEIGHT=1.60,
      DET_EVAC_DIST=1,
      DET_MEAN=0.0,DET_LOW=0.00,DET_HIGH=0.0,
      DENS_INIT=4.0,
      COLOR_METHOD=0, I_HERDING_TYPE=0,
   AVATAR_COLOR= 'BLACK',
   OUTPUT_SPEED=.TRUE.   /

$$$$$$$$ Etage OG2 $$$$$$$$

&EVAC ID = 'Personen',
      NUMBER_INITIAL_PERSONS = 200,
      XB=-9.1,4.5,0.7,4.7,0.9,1.1,
   PRE_EVAC_DIST=0,
   PRE_MEAN=0,
      AVATAR_COLOR = 'YELLOW', AGENT_TYPE=1,
   KNOWN_DOOR_NAMES = 'EXIT',
   PERS_ID = 'Adult' /

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
TÜREN UND AUSGÄNGE    $$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

&EXIT ID='EXIT', XB=9.7,9.7,2.1,3.3,0.9,1.1, IOR=+1/

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
GEOMETRIE $$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

&OBST ID='Obstruction', XB=-10.0,9.8,5.3,5.4,0.9,1.1, SURF_ID='INERT'/
&OBST ID='Obstruction', XB=5.0,5.1,0.1,5.3,0.9,1.1, SURF_ID='INERT'/
&OBST ID='Obstruction', XB=-10.0,-9.9,0.1,5.3,0.9,1.1, SURF_ID='INERT'/
&OBST ID='Obstruction', XB=-10.0,9.8,0.0,0.1,0.9,1.1, SURF_ID='INERT'/
&OBST ID='Obstruction', XB=9.7,9.8,0.1,5.3,0.9,1.1, SURF_ID='INERT'/

&HOLE ID='Hole', XB=4.9,5.2,2.1,3.3,0.0,2.0/

&TAIL /

FDS+EVAC-File (door connection):

door_200P.fds
Generated by PyroSim - Version 2019.3.1204
25.02.2020 14:10:53

&HEAD CHID='door_200P'/
&TIME T_END=200.0/
&DUMP DT_RESTART=300.0, DT_SL3D=0.25/

&MISC EVACUATION_MC_MODE=.TRUE.
EVACUATION_DRILL=.TRUE./

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
AUSGABE    $$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

&SLCF PBZ=1.0, QUANTITY='VELOCITY', VECTOR=.TRUE., EVACUATION=.TRUE./

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
MESH $$$$$$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

&MESH IJK=198,54,1, XB=-10.0,9.8,0.0,5.4,0.9,1.1,
EVACUATION=.TRUE., EVAC_HUMANS=.TRUE., ID='EMesh_EG', EVAC_Z_OFFSET=1.0 /

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
PERSONEN $$$$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

$$$$$$$$ Etage OG2 $$$$$$$$

&EVAC ID = 'Personen',
      NUMBER_INITIAL_PERSONS = 200,
      XB=-9.1,4.5,0.7,4.7,0.9,1.1,
   PRE_EVAC_DIST=0,
   PRE_MEAN=0,
      AVATAR_COLOR = 'YELLOW', AGENT_TYPE=1,
   KNOWN_DOOR_NAMES = 'EXIT','Raum1_out',
   PERS_ID = 'Adult' /

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
TÜREN UND AUSGÄNGE    $$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

&EXIT ID='EXIT', XB=9.7,9.7,2.1,3.3,0.9,1.1, IOR=+1/

&DOOR ID='Raum1_out', IOR=+1,
      EXIT_SIGN=.TRUE.,
      TO_NODE= 'Raum2_in'
      XB=5.0,5.0,2.1,3.3,0.9,1.1,/

&DOOR ID='Raum2_in', IOR=-1,
      EXIT_SIGN=.FALSE.,
      TO_NODE= 'Raum1_out'
      XB=5.1,5.1,2.1,3.3,0.9,1.1,/

$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
GEOMETRIE $$$$$$$$$$$$$$$$$$$$$$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$

&TAIL /

TimoK

unread,

Feb 27, 2020, 3:28:18 AM2/27/20

to FDS and Smokeview Discussions

Yes, a hole and a door (or exit) are not same. Well, an EXIT might be quite close to a HOLE. But the door is actually

a connection between two evacuation meshes, typically different floors. SO, a door is a DOOR1 =>DOOR2 connection.

So there the flow is affected whats is in front of the DOOR2. And there are some agents there, because the

door algorithm has already placed some agents there (if there is a queue in front of the DOOR1).

And there should be a queue in front of an EVSS (if the FAC_SPEED_DOWN and/or FAC_SPEED_UP < 1).

The walking velocity is slower on the EVSS => there should be a queue in fornt of it.

An EXIT is just checking, if some agent crosses the line defined by XB and then just removes the agent from the

calculation.

If you want to model an exit door of a building "correctly", then do something like:

1) Make a HOLE to the outer wall and extend your evacuation mesh to include some outside space.

2) To make the guiding flow field calculation more or less good, put some OBST to the outside

on both sides of your door HOLE, say some 0.5m away or so. and then put a wide EXIT to the

mesh outer boundary (hole width + 0.5m + 0.5m, because you have the OBSTs at the outside).

This way the EXIT is going to be wide enough so that it is not slowing down the agent flow rate.

So your limiting factor to the agent flow rate is determined by the HOLE at the building outer wall.

This thing should be done also for some other evacuation simulation prgrammes, like Simulex, for

example. You get different flows if you place the exit directly at the wall or make a hole and place

a wider exit further away.

As said, look the FDS+Evac manual: Figure 13. Test geometries used to calculate the specific flows through doors and corridors

There you see, that the FDS+Evac is tested for a exit door that is done like points 1 and 2 above. And

this kind of an exit door produces flows that are given in the Figure 15. Effects of different model parameters.

And in "Figure 32. The specific flows through doors." the flows trough a "HOLE type door" are shown.

In order to model a door correctly, you should model also the space after the door. In real world, the

humans are not vanishing after the door anywhre. Well, the only case I can imagine is a door hole

in a high rise building at the upper floors. The humans will drop suddenly down.

All this said, you have still a problem with the flows in a DOOR => DOOR connecting two different

floors of the building (or in staircases done with the EVSS, not STRS). There are some things

you could do:

1) make a test calculation (similar to "Figure 32. The specific flows through doors." analysis),

where you see, what kind of flows you get, when you change the width of the DOORs in

your door => door connection. An after this, you choose a door width that is giving your the

flow that your are supposing to have in the case of a HOLE.

2) You put the door=>door connection in the staircase, for example on the intermediate landing.

And make the door=>door to be wide enough so that it is not the bottleneck of your case.

For example, make the intermediate landings a little bit wider than they are in the reality.

You can test this quite simple. Model just the stairs (and put enough agents going to the

stairs) and see, where you get bottlenecks. E.g., see, if the door=>door connection at

the intermediate landing is able to feed enough agents so that there is enough agents

trying to go to the stair fligh going down at this landing. The witdth of the stair flight

is typically the rate limiting factor.

TimoK

Reply all

Reply to author

Forward

0 new messages