Doors/Exits in FDS6 (+EVAC)
1,387 views
Skip to first unread message
Anass
Jul 7, 2015, 8:29:45 AM7/7/15
to fds...@googlegroups.com
Hi everyone,
I'm trying to simulate the vertical evacuation of a tall building with FDS6 (see the attached file). However, when I launched the simulation, I got the list of errors attached. It seems that is the same error for all the exits/doors used in the simulation.
Could you please tell me how can I correct this ?
Thank you in advance.
Anass
TimoK
Jul 8, 2015, 4:32:36 AM7/8/15
to fds...@googlegroups.com
Check your evacuation geometry.
The temporary FDS+Evac homepage (in a text format):
http://virtual.vtt.fi/virtual/proj6/fdsevac/documents/FDS+Evac_textbased_homepage.txt
I have to made a html-page "by hand". The "old project pages" were removed, when
we had here a web page reform...
Your error is probably due to your XB definitions in your evacuation meshes. Your z-coordinates
in XB are touching the ceiling and/or the floor => your entire evacuation meshes are filled with
obstacles => no place to put anything into these meshes.
So add:
MISC-namelist: EVACUATION_DRILL=.TRUE.
You see just the evacuation meshes in Smokeview. In Smokeview menu show/hide =>
geometry => surfaces => exterior should be taken away, so that the -z and +z boundaries
of the evacuation meshes are not shown (shown as yellow "solid walls" by default).
See the http://virtual.vtt.fi/virtual/proj6/fdsevac/documents/FDS+EVAC_5_Guide.pdf
And there: Figure 33. A 2D evacuation mesh on page 69.
Take all EVSS out (remove the "&" characters, for example). This helps to debug.
Remove ("&" away...) as many exits and doors that you have no errrors. Well, you
could remove them all, but then you should also remove your evacuation meshes.
Then activate one evacuation mesh (say, the bottom ones having exits) and
one exit there (you need at least one exit or door in an evacuation mesh to avoid
errors). See if you manage to see the evacuation mesh. If you get error for the
exit, move the exit XB so that it is middle of an empty space. If this does not
work => probably your evacuation mesh is completely solid obstacle => change
the XB of the mesh.
And do the evacuation mesh activation one by one (from bottom up).
TimoK
The temporary FDS+Evac homepage (in a text format):
http://virtual.vtt.fi/virtual/proj6/fdsevac/documents/FDS+Evac_textbased_homepage.txt
Direct link to the FDS+Evac Readme_InputErrors-file: http://virtual.vtt.fi/virtual/proj6/fdsevac/documents/Readme_InputErrors.txt
I have to made a html-page "by hand". The "old project pages" were removed, when
we had here a web page reform...
Your error is probably due to your XB definitions in your evacuation meshes. Your z-coordinates
in XB are touching the ceiling and/or the floor => your entire evacuation meshes are filled with
obstacles => no place to put anything into these meshes.
So add:
MISC-namelist: EVACUATION_DRILL=.TRUE.
You see just the evacuation meshes in Smokeview. In Smokeview menu show/hide =>
geometry => surfaces => exterior should be taken away, so that the -z and +z boundaries
of the evacuation meshes are not shown (shown as yellow "solid walls" by default).
See the http://virtual.vtt.fi/virtual/proj6/fdsevac/documents/FDS+EVAC_5_Guide.pdf
And there: Figure 33. A 2D evacuation mesh on page 69.
Take all EVSS out (remove the "&" characters, for example). This helps to debug.
Remove ("&" away...) as many exits and doors that you have no errrors. Well, you
could remove them all, but then you should also remove your evacuation meshes.
Then activate one evacuation mesh (say, the bottom ones having exits) and
one exit there (you need at least one exit or door in an evacuation mesh to avoid
errors). See if you manage to see the evacuation mesh. If you get error for the
exit, move the exit XB so that it is middle of an empty space. If this does not
work => probably your evacuation mesh is completely solid obstacle => change
the XB of the mesh.
And do the evacuation mesh activation one by one (from bottom up).
TimoK
Anass
Jul 14, 2015, 5:49:22 AM7/14/15
to fds...@googlegroups.com
Hi Timo,
Thank you for your response. I have fixed the majority of errors but I still have some inconsistencies of results. Indeed, when I check the results of my simulation, I realise that when the agents reach the first floor, they aren't borrowing the stairs to reach the ground floor. I didn't understand the source of this inconsistency.
In the other hand, when I defined doors for the evacuation of the landings stairways (P_A_1_IN ... P_A_11_OUT and P_B_1_IN ... P_B_11_OUT), I realised that when I use the parameter "To_Node" agents do not go down the stairs and otherwise if I don't use it. I don't understund why. In addition, when I don't use it for the tower A, I have this error message :
"ERROR: EVAC: Emesh 25 MESH_EVAC_PALIER_R1 needs at least one DOOR/EXIT. (CHID: cite)"
But when I check my doors, I realise that I have other doors (like PORTE_ESCALIER_A_1_IN) that they have this Mesh like a Mesh_ID. So, I also don't understund why FDS gives me this error.
I know I have a lot to learn about FDS but I hope you can help me to fix this errors.
Thank you in advance
Anass
TimoK
Jul 15, 2015, 4:00:09 AM7/15/15
to fds...@googlegroups.com
Well:
You need TO_NODE in a DOOR line, if this door is a "target door". So, some agents
are going to use it to "go out of the room" type of a door. A door is a transporter (see, e.g.,
https://en.wikipedia.org/wiki/Transporter_%28Star_Trek%29). It moves agents
from a one place (x1,y1,z1) to some other place (x2,y2,z2). The x1,y1,z1 are
the coordinates of the agent at the door1, where they are to be transported to
an another door, named door 2 and having coordinates x2,y2,z2.
Do not use EVSS. See your geometry in Smokeview. See, how the "stairs" are shown
in the two-dimensional plane of the evacuation meshes. These are more or less like
floor plans. The stairs should be "projected" to some evacuation mesh.
A good strategy for a stairs (with intermediate landings between the floors).
Floor numbering: F1 (ground floor), F2 (first floor above the ground), F3 and
so on (here in Finland, "first floor" is usually at the ground level, we count
the floors like that...).
Floor 3 evacuation mesh should contain: The landing at this level. The stair fligth
going down to the intermediate landing 2/3. The stair flight going up to
the intermediate landing 3/4 (if you have 4th floor...).
Floor 2: The landing at this level. The stair fligth going down to the intermediate
landing 1/2. The stair flight going up to the intermediate landing 2/3
Intermediate landing 2/3: Well, you can choose to put this to 2nd or 3rd floor mesh.
This usually does not matter too much. So, here you put the DOOR1 => DOOR2
transporter, where you have a solid wall at the start/end of the stair flight, i.e, where
the one mesh ends and the agents should go to the other mesh.
Floor 1: The landing at this level. The stair flight going up to the intermediate landing 1/2.
Intermediate landing 1/2: Well, you should use the same choice as for the intemediate
landking 2/3. So, the DOOR1 => DOOR2 transporter should be at the same place.
For the Intermediate landings, where you have the door=>door transporter: The TO_NODE
door will place automatically a one grid cell thick OBST behind it. But to be sure that
you have a wall here, your could input the OBST by yourself also. There should not
be any "holes" in the floor plan to "outside". The floor plans should be "sealed".
If you are using DOOR=>ENTR (of DOOR=> DOOR_without_TO_NODE, then
you need to input the OBST behind the ENTR or DOOR_without_TO_NODE to
be sure that this evacuation mesh is "sealed".
And at floor 1 you put your EXIT door where you want in the stairs. Or if the stairs do not
go outside, but lead to first floor, you should have "open hole", where the stair door to
the first floor is. And some EXIT somewhere else in the first floor mesh.
Sometimes your 1st floor stair geometry is such that the EXIT is under the intermediate
landing (high enough floor heights needed, of course). This case you shoud define
the intermediate landing to be in the 2nd floor mesh. And similarly for the other
intermediate landings also. So, it is nice to start doing the stairs "bottom up".
TimoK
You need TO_NODE in a DOOR line, if this door is a "target door". So, some agents
are going to use it to "go out of the room" type of a door. A door is a transporter (see, e.g.,
https://en.wikipedia.org/wiki/Transporter_%28Star_Trek%29). It moves agents
from a one place (x1,y1,z1) to some other place (x2,y2,z2). The x1,y1,z1 are
the coordinates of the agent at the door1, where they are to be transported to
an another door, named door 2 and having coordinates x2,y2,z2.
Do not use EVSS. See your geometry in Smokeview. See, how the "stairs" are shown
in the two-dimensional plane of the evacuation meshes. These are more or less like
floor plans. The stairs should be "projected" to some evacuation mesh.
A good strategy for a stairs (with intermediate landings between the floors).
Floor numbering: F1 (ground floor), F2 (first floor above the ground), F3 and
so on (here in Finland, "first floor" is usually at the ground level, we count
the floors like that...).
Floor 3 evacuation mesh should contain: The landing at this level. The stair fligth
going down to the intermediate landing 2/3. The stair flight going up to
the intermediate landing 3/4 (if you have 4th floor...).
Floor 2: The landing at this level. The stair fligth going down to the intermediate
landing 1/2. The stair flight going up to the intermediate landing 2/3
Intermediate landing 2/3: Well, you can choose to put this to 2nd or 3rd floor mesh.
This usually does not matter too much. So, here you put the DOOR1 => DOOR2
transporter, where you have a solid wall at the start/end of the stair flight, i.e, where
the one mesh ends and the agents should go to the other mesh.
Floor 1: The landing at this level. The stair flight going up to the intermediate landing 1/2.
Intermediate landing 1/2: Well, you should use the same choice as for the intemediate
landking 2/3. So, the DOOR1 => DOOR2 transporter should be at the same place.
For the Intermediate landings, where you have the door=>door transporter: The TO_NODE
door will place automatically a one grid cell thick OBST behind it. But to be sure that
you have a wall here, your could input the OBST by yourself also. There should not
be any "holes" in the floor plan to "outside". The floor plans should be "sealed".
If you are using DOOR=>ENTR (of DOOR=> DOOR_without_TO_NODE, then
you need to input the OBST behind the ENTR or DOOR_without_TO_NODE to
be sure that this evacuation mesh is "sealed".
And at floor 1 you put your EXIT door where you want in the stairs. Or if the stairs do not
go outside, but lead to first floor, you should have "open hole", where the stair door to
the first floor is. And some EXIT somewhere else in the first floor mesh.
Sometimes your 1st floor stair geometry is such that the EXIT is under the intermediate
landing (high enough floor heights needed, of course). This case you shoud define
the intermediate landing to be in the 2nd floor mesh. And similarly for the other
intermediate landings also. So, it is nice to start doing the stairs "bottom up".
TimoK
Anass
Jul 16, 2015, 9:43:59 AM7/16/15
to fds...@googlegroups.com
Hi Timo,
Thank you for your reply.
As you have said, I have removed all EVSSes and Stairs-Holes (see the file aatched) and now I have no idea of how Evacuees will evacuate. You said also that I have to project my stairs to evacuation meshes but I dont know how can I do this. Could you please explain to me (with an example if you can)?
Regards,
Anass
TimoK
Jul 17, 2015, 4:12:12 AM7/17/15
to fds...@googlegroups.com
> As you have said, I have removed all EVSSes and Stairs-Holes (see the file aatched)
> and now I have no idea of how Evacuees will evacuate. You said also that I
> and now I have no idea of how Evacuees will evacuate. You said also that I
I did not say anything for the stair HOLEs. These you should not remove. Just the EVSS
lines so that you see everything at the correct z levels of the evacuation meshes. Then
you can see, that you have defined your doors, entrs, exits, etc on the correct evacuation
mesh. You should have the stair geometry included in the evacuation meshes. So, the
walls should be there. And there should be empty space for the agents to fill the stair
space. The stair space is belonging to some evacuation mesh. And the stair space shoud
be at this evacuation mesh z-level empty (no stair flights, landings, or any other OBST
should be shown in this evacuation mesh).
Well, you could think the stairs as: Have a long red rope. Go to the top of the stairs.
Start going down and start uncoiling the rop. Go to the bottom end of the stairs.
Now you have a continuous rope going from top to down.
What to do next? You should find scissors and cut that continuous rope at
some points in the stairs (between the floors). The pieces of the ropes should
be such, that you can say, that one piece belongs to a floor 1, one belongs
to floor 2, and so on. No piece should have no evacuation mesh, where it
belongs.
How to check that your piece of rope => evacuation mesh is done correctly?
The piece of rope can not be more than a "circle". I mean, than it should not
turn more than 360 degress. This way, if you see the piece of rope directly
above (and no perspective, i.e., you look very far away), you will see
that the piece of rope is not having the ends to be on top of each other
(or any other piece of this rope).
Who to place door=>entr connections? Well, you place these at the points,
where you cut your rope. Now the rope is cutted => it is not continuous
anymore => the agents need something "to jump form one piece of the
rope to the next piece of the rope".
It is up to you to decide, where you want to cut the rope. But at the intermediate
landing/stair flight points are quite nice places to cut the rope.
How to decide to which main evacuation mesh different EVSS describing
stair flights and intermediate landings belong? Well, that is easy. Just
check that which piece of rope you have there.
Check the page:
http://virtual.vtt.fi/virtual/proj6/fdsevac/examples_fds6.html
And see the example cases: Staircase and Stair Example. See their fds files.
And run those cases also and see them in Smokeview.
Check that they have:
And remove the "&" characters from the EVSS lines.
So, you see just the evacuation meshes and every evacuation object (including the agents)
at the correponding evacuation mesh z-level. This should enlighten you how the stairs should
be modelled.
TimoK
lines so that you see everything at the correct z levels of the evacuation meshes. Then
you can see, that you have defined your doors, entrs, exits, etc on the correct evacuation
mesh. You should have the stair geometry included in the evacuation meshes. So, the
walls should be there. And there should be empty space for the agents to fill the stair
space. The stair space is belonging to some evacuation mesh. And the stair space shoud
be at this evacuation mesh z-level empty (no stair flights, landings, or any other OBST
should be shown in this evacuation mesh).
Well, you could think the stairs as: Have a long red rope. Go to the top of the stairs.
Start going down and start uncoiling the rop. Go to the bottom end of the stairs.
Now you have a continuous rope going from top to down.
What to do next? You should find scissors and cut that continuous rope at
some points in the stairs (between the floors). The pieces of the ropes should
be such, that you can say, that one piece belongs to a floor 1, one belongs
to floor 2, and so on. No piece should have no evacuation mesh, where it
belongs.
How to check that your piece of rope => evacuation mesh is done correctly?
The piece of rope can not be more than a "circle". I mean, than it should not
turn more than 360 degress. This way, if you see the piece of rope directly
above (and no perspective, i.e., you look very far away), you will see
that the piece of rope is not having the ends to be on top of each other
(or any other piece of this rope).
Who to place door=>entr connections? Well, you place these at the points,
where you cut your rope. Now the rope is cutted => it is not continuous
anymore => the agents need something "to jump form one piece of the
rope to the next piece of the rope".
It is up to you to decide, where you want to cut the rope. But at the intermediate
landing/stair flight points are quite nice places to cut the rope.
How to decide to which main evacuation mesh different EVSS describing
stair flights and intermediate landings belong? Well, that is easy. Just
check that which piece of rope you have there.
Check the page:
http://virtual.vtt.fi/virtual/proj6/fdsevac/examples_fds6.html
And see the example cases: Staircase and Stair Example. See their fds files.
And run those cases also and see them in Smokeview.
Check that they have:
&MISC EVACUATION_DRILL=.TRUE., /
And remove the "&" characters from the EVSS lines.
So, you see just the evacuation meshes and every evacuation object (including the agents)
at the correponding evacuation mesh z-level. This should enlighten you how the stairs should
be modelled.
TimoK
TimoK
Jul 17, 2015, 10:40:07 AM7/17/15
to fds...@googlegroups.com
Something is not nice in your meshes and how you define door => door and your
staircases.
See the attached images. They are clipped (in Smokeview using Clip dialog)
at z_min=10.75 and z_max=11.25 for the movement case. The XB of the
evacuation mesh defines the movement mesh z. The visibility mesh is
a little bit above (1.6m above "the floor", z_floor= 0.5*(xb_z1 + xb_z2) - evac_z_offset).
(The visibility mesh has z-range, plus/minus 0.29m from the visibility mesh z.)
&MESH ID='MESH_EVACUATION_A_R3', IJK=15,80,1, XB=85.0,92.5,82.5,122.5,10.75,11.25, EVACUATION=.TRUE., EVAC_HUMANS=.TRUE., EVAC_Z_OFFSET=0.75/
So, there seems to be some obstacles in the mid-section of your movement mesh. And your
visibility meshes are solid OBSTs. The agents can not see anything.
You do not need to define separate meshes for the exit staircases. You can include these in
the other evacuation meshes. Use HOLEs and OBSTs with EVACUATION=.TRUE. to make
the things that you need. E.g., the floor => staircase door should now be removed. This
level of the staircase is belonging to the corresponding floor evacuation mesh. So, you just
use a HOLE with EVACUATION=.TRUE. to make a hole to the wall, so that the agents
can flow to the stairs. (You can also model the stairs using the separate evacuation meshes,
but this is a lot of work.)
TimoK
PS. I'm heading now for summer holidays.
staircases.
See the attached images. They are clipped (in Smokeview using Clip dialog)
at z_min=10.75 and z_max=11.25 for the movement case. The XB of the
evacuation mesh defines the movement mesh z. The visibility mesh is
a little bit above (1.6m above "the floor", z_floor= 0.5*(xb_z1 + xb_z2) - evac_z_offset).
(The visibility mesh has z-range, plus/minus 0.29m from the visibility mesh z.)
&MESH ID='MESH_EVACUATION_A_R3', IJK=15,80,1, XB=85.0,92.5,82.5,122.5,10.75,11.25, EVACUATION=.TRUE., EVAC_HUMANS=.TRUE., EVAC_Z_OFFSET=0.75/
So, there seems to be some obstacles in the mid-section of your movement mesh. And your
visibility meshes are solid OBSTs. The agents can not see anything.
You do not need to define separate meshes for the exit staircases. You can include these in
the other evacuation meshes. Use HOLEs and OBSTs with EVACUATION=.TRUE. to make
the things that you need. E.g., the floor => staircase door should now be removed. This
level of the staircase is belonging to the corresponding floor evacuation mesh. So, you just
use a HOLE with EVACUATION=.TRUE. to make a hole to the wall, so that the agents
can flow to the stairs. (You can also model the stairs using the separate evacuation meshes,
but this is a lot of work.)
TimoK
PS. I'm heading now for summer holidays.
Anass
Jul 22, 2015, 8:22:40 AM7/22/15
to FDS and Smokeview Discussions
Hi Timo,
I'm glad to hear from you. Thank you very much for your answers. I will consider your suggestions. I will first reintroduce all HOLEs removed.
I hope you spend a nice summer holidays.
Anass
Anass
Jul 29, 2015, 9:54:52 AM7/29/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
I think you're still on holliday but I wanted to keep you informed about the progress of my work. I have reintroduced all the HOLEs. Then I have changed the z-position of my evacuation Meshes. I have also limited the evacuation to one tower and I would do the same thing for the other tower when the first tower runs as it should.
I have attached some files to better understanding what I'm doing. The first one "cite.fds" represent the evacuation of all floors (1st to 12th) but I have defined the exit at the first floor (see figure 1). The second one, I have added the evacuation of the first floor using a straight flight of stairs (see figure 2) but when I run the simulation I have this message :
forrtl: severe (157): Program Exception - access violation
Image PC Routine Line Source
fds.exe 000000013FFEF180 Unknown Unknown Unknown
fds.exe 000000013FFC01E1 Unknown Unknown Unknown
fds.exe 000000013FE2C66F Unknown Unknown Unknown
fds.exe 000000013F88E249 Unknown Unknown Unknown
fds.exe 000000013F87C056 Unknown Unknown Unknown
fds.exe 000000013F8571B8 Unknown Unknown Unknown
kernel32.dll 0000000076D859CD Unknown Unknown Unknown
ntdll.dll 0000000076FBB981 Unknown Unknown Unknown
and the simulation don't start.
I have also do a simplified example (see the attached file Stairs_4.fds) and it works like it should. I do the same thing with the stairwells of the tall building and it doesn't work. I don't know what I'm doing wrong because it is the same thing as in the simplified example.
TimoK
Aug 4, 2015, 5:18:50 AM8/4/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi!
Check the attached file. The top floors should almost be working. You just need to
adjust the "parlier" mesh geometries. You have "holes" on both sides of the doors.
Well, the other side is not problem, because you should put some kind of
a railing in the "parlier" meshes. But on the outer wall side you should not
have a hole.
I changed your meshes so, that they have dx=0.5,dy=0.5 for the floors
and dx=0.25 for the "parliers". You have overlapping meshes. The parliers
overlap with the floors. Now the meshes match better. Might be that this
was not your problem, but this is nice anyway. You have the OBST at the
same place in the overlapping meshes now. It might have been that you had
empty space in one mesh, and solid obst in the other mesh at the door => door
connections at the sides of the doors. The mid point of XB was nice, but the
door => door algorithm put an agent from an empty space inside some OBST.
This makes "access violation" errors easily.
See the attached file, for example:
Timo: This is now working. PORTE_ESCALIER_A_10_IN', XB=85.25,85.25 ( agent goes in a door at x=85.0, and comes
out of a door one floor down at x=85.25). The x=85.0-85.25 is solid OBST due to the DOOR with TO_NODE set.
These doors generate one grid cell thick OBSTs automatically behind them.
Timo: Well, you need to change your MESH_EVAC_PALIER_R9/10 (and others) geometry. It seems that you are missing
something in the geometry. Some railings between stair flights or something like that? And see that you
do not have holes on either side of your doors. Now you have overlapping evaucation meshes. It might be that
you should add an OBST at the side of the PORTE_ESCALIER_A_10_IN/PORTE_ESCALIER_A_11_OUT doors. One sees there
a green OBST, but this is in the mesh MESH_EVACUATION_A_RXXX (the automatic OBST behind a DOOR).
&DOOR ID='PORTE_ESCALIER_A_10_IN', XB=84.75,84.75,103.5,104.5,28.0,28.5, XYZ=85.0,104.0,28.25, MESH_ID='MESH_EVAC_PALIER_R9/10', IOR=1, EXIT_SIGN=.FALSE., COLOR='BLUE'/
&DOOR ID='PORTE_ESCALIER_A_11_OUT', XB=85.0,85.0,103.5,104.5,30.5,31.0, XYZ=85.2,104.0,30.75, MESH_ID='MESH_EVAC_PALIER_R10/11', IOR=-1, TO_NODE='PORTE_ESCALIER_A_10_IN', EXIT_SIGN=.TRUE., COLOR='BLUE'/
Wbr,
Timo
Check the attached file. The top floors should almost be working. You just need to
adjust the "parlier" mesh geometries. You have "holes" on both sides of the doors.
Well, the other side is not problem, because you should put some kind of
a railing in the "parlier" meshes. But on the outer wall side you should not
have a hole.
I changed your meshes so, that they have dx=0.5,dy=0.5 for the floors
and dx=0.25 for the "parliers". You have overlapping meshes. The parliers
overlap with the floors. Now the meshes match better. Might be that this
was not your problem, but this is nice anyway. You have the OBST at the
same place in the overlapping meshes now. It might have been that you had
empty space in one mesh, and solid obst in the other mesh at the door => door
connections at the sides of the doors. The mid point of XB was nice, but the
door => door algorithm put an agent from an empty space inside some OBST.
This makes "access violation" errors easily.
See the attached file, for example:
Timo: This is now working. PORTE_ESCALIER_A_10_IN', XB=85.25,85.25 ( agent goes in a door at x=85.0, and comes
out of a door one floor down at x=85.25). The x=85.0-85.25 is solid OBST due to the DOOR with TO_NODE set.
These doors generate one grid cell thick OBSTs automatically behind them.
Timo: Well, you need to change your MESH_EVAC_PALIER_R9/10 (and others) geometry. It seems that you are missing
something in the geometry. Some railings between stair flights or something like that? And see that you
do not have holes on either side of your doors. Now you have overlapping evaucation meshes. It might be that
you should add an OBST at the side of the PORTE_ESCALIER_A_10_IN/PORTE_ESCALIER_A_11_OUT doors. One sees there
a green OBST, but this is in the mesh MESH_EVACUATION_A_RXXX (the automatic OBST behind a DOOR).
&DOOR ID='PORTE_ESCALIER_A_10_IN', XB=84.75,84.75,103.5,104.5,28.0,28.5, XYZ=85.0,104.0,28.25, MESH_ID='MESH_EVAC_PALIER_R9/10', IOR=1, EXIT_SIGN=.FALSE., COLOR='BLUE'/
&DOOR ID='PORTE_ESCALIER_A_11_OUT', XB=85.0,85.0,103.5,104.5,30.5,31.0, XYZ=85.2,104.0,30.75, MESH_ID='MESH_EVAC_PALIER_R10/11', IOR=-1, TO_NODE='PORTE_ESCALIER_A_10_IN', EXIT_SIGN=.TRUE., COLOR='BLUE'/
Wbr,
Timo
Anass
Aug 24, 2015, 2:13:18 AM8/24/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
Thank you for your reply. I was on summer holiday. I will check the file attached and tell you if it is ok.
Sincerly yours,
Anass
Anass
Aug 25, 2015, 3:02:44 AM8/25/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
I have checked the file you have attached and I modified some things (see the attached file) : Well, when I introduce the evacuation of the floor R-1, I have this error :
" ERROR: DOOR PORTE_ESCALIER_A_RC_IN problem with VENT_FFIELD (CHID: cite_modAnass) "
and I don't know why ? Could you please enlight me ?
Lots of thanks,
Anass
TimoK
Aug 26, 2015, 6:27:18 AM8/26/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Well, something is not going as it should. Or I did not find the thing easily.
But use ENTR, not DOOR for this "dummy door" that is more or less like
an ENTR. Note, the the IOR for ENTR is the direction away from the
entry (the agents are moving away from an entry). The IOR is
towards a door for DOOR lines.
&ENTR ID='PORTE_ESCALIER_A_-1_IN', XB=76.25,78.0, 106.5,106.5, 0.5,1.0,
IOR=+2, /
DOOR ID='PORTE_ESCALIER_A_-1_IN', XB=76.25,78.0, 106.5,106.5, 0.5,1.0, XYZ=77.6,106.5,0.75,
MESH_ID='MESH_EVACUATION_A_R0', IOR=-2/
Might be that there is something wrong in the source code. But this should not be
the case, because I have used "dummy doors" before also. But I check this using
a simple test case.
Wbr,
TimoK
But use ENTR, not DOOR for this "dummy door" that is more or less like
an ENTR. Note, the the IOR for ENTR is the direction away from the
entry (the agents are moving away from an entry). The IOR is
towards a door for DOOR lines.
&ENTR ID='PORTE_ESCALIER_A_-1_IN', XB=76.25,78.0, 106.5,106.5, 0.5,1.0,
IOR=+2, /
DOOR ID='PORTE_ESCALIER_A_-1_IN', XB=76.25,78.0, 106.5,106.5, 0.5,1.0, XYZ=77.6,106.5,0.75,
MESH_ID='MESH_EVACUATION_A_R0', IOR=-2/
Might be that there is something wrong in the source code. But this should not be
the case, because I have used "dummy doors" before also. But I check this using
a simple test case.
Wbr,
TimoK
Anass
Aug 26, 2015, 8:27:04 AM8/26/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
Thank you for your suggestion. I try it right now.
Before posting my message I have already checked this using a simple test case (see the attached file) and it works. So, I don't know why in my case (complicated geormetry) it doesn't work.
Sincerely yours,
Anass
TimoK
Aug 27, 2015, 9:04:42 AM8/27/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Well, I found a minor bug in the source code. This was in a place, where I check that the
input is correct. So, I was checking for the DOORs without TO_NODE same things than
for a "real agent transfering" DOORs which have TO_NODE set. Now I added an IF statement,
that the source code does not perform all the input checks for the "dummy doors".
This minor bug did not change the calculated results. It was just a bug in the checking
of the input file part. I do not exactly know, how some "dummy door" input was able
to go through the check. It might have to do with the order of the meshes in the input
and if the fire meshes have ID given or not. Mesh name = 'null' was the default in earlier
FDS version, it might be that in FDS 6.2.0 the fire meshes have, by default, some other
name than 'null'.
But, for now, the thing is to change the dummy-DOORs to ENTRs. They perform
similarly, just remember the IOR thing.
TimoK
input is correct. So, I was checking for the DOORs without TO_NODE same things than
for a "real agent transfering" DOORs which have TO_NODE set. Now I added an IF statement,
that the source code does not perform all the input checks for the "dummy doors".
This minor bug did not change the calculated results. It was just a bug in the checking
of the input file part. I do not exactly know, how some "dummy door" input was able
to go through the check. It might have to do with the order of the meshes in the input
and if the fire meshes have ID given or not. Mesh name = 'null' was the default in earlier
FDS version, it might be that in FDS 6.2.0 the fire meshes have, by default, some other
name than 'null'.
But, for now, the thing is to change the dummy-DOORs to ENTRs. They perform
similarly, just remember the IOR thing.
TimoK
Anass
Aug 28, 2015, 3:05:12 AM8/28/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
I have used your suggestion (ENTR instead of DOOR) and it works but when I checked the results, I found that there are several traffic jams in the stairwells (see the pictures attached). I don't know if this is due to the size of stairs or not ? can you please enlight me about this ? Could you please explain to me why this traffic jams are formed ? and how can I avoid this phenomenon without changing the geometry of the building (I don't know if it's possible)?
Thanks a lot,
Anass
TimoK
Aug 28, 2015, 3:29:48 AM8/28/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Well, hard to tell. If you have plenty of agents, you will end up with queueing
at the stairs. But something might also be wrong in your input file. This you
should chekc seeing those places, where you have traffic jam:
1) If the traffic jam is in front of a DOOR => ENTR point in the stairs (or at
the door => entr that is from the floor to the stair mesh), then see, if there
is empty space in front of the ENTR. If there is empty space (no agents),
and the DOOR is not placing agents to this empty space, then something
is wrong in the input.
2) Same as 1), but now you have agents in front of ENTR. Are the agents trying
to move to correct direction? See their faces (you could use avatar "ellipsoid"
in Smokeview to see really clearly where the heads are pointing). If they are
pointing away from the ENTR (i.e., down to the stair, to the downward direction
of the stairs, to the direction where they should go), then the thing is just a
traffic jam in the stairs. The stairs are full of people and the persons at the
floor levels should wait at the stair door so that they can merge the flow in
stairs.
If the agents in the stairs after the ENTR are having "wrong directions" then
something is wrong in your input. "wrong directions" meahs that their heads
are pointing towards walls or towards your ENTR or something like that. This
is due to some bugs in your input file then. Check the "guiding flow fields"
towards the doors in the stairs (stair meshese). The velocity vectors should
go downwards in the stairs. Check the next files for instructions:
at the stairs. But something might also be wrong in your input file. This you
should chekc seeing those places, where you have traffic jam:
1) If the traffic jam is in front of a DOOR => ENTR point in the stairs (or at
the door => entr that is from the floor to the stair mesh), then see, if there
is empty space in front of the ENTR. If there is empty space (no agents),
and the DOOR is not placing agents to this empty space, then something
is wrong in the input.
2) Same as 1), but now you have agents in front of ENTR. Are the agents trying
to move to correct direction? See their faces (you could use avatar "ellipsoid"
in Smokeview to see really clearly where the heads are pointing). If they are
pointing away from the ENTR (i.e., down to the stair, to the downward direction
of the stairs, to the direction where they should go), then the thing is just a
traffic jam in the stairs. The stairs are full of people and the persons at the
floor levels should wait at the stair door so that they can merge the flow in
stairs.
If the agents in the stairs after the ENTR are having "wrong directions" then
something is wrong in your input. "wrong directions" meahs that their heads
are pointing towards walls or towards your ENTR or something like that. This
is due to some bugs in your input file then. Check the "guiding flow fields"
towards the doors in the stairs (stair meshese). The velocity vectors should
go downwards in the stairs. Check the next files for instructions:
Direct link to the FDS+Evac Readme_InputErrors-file: http://virtual.vtt.fi/virtual/proj6/fdsevac/documents/Readme_InputErrors.txt
June 2015: FDS+Evac Homepage (temprorary) http://virtual.vtt.fi/virtual/proj6/fdsevac/documents/FDS+Evac_textbased_homepage.txt
TimoK
Anass
Sep 3, 2015, 5:52:06 AM9/3/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
Well, I have checked my fds file and I can't find where I'm making mistakes. I have also observed my results and I have noted some inconsistencies of the results.
First, I have performed the evacuation of the building when all the rooms are occupied (288 residents) and I have noted that the evacuation of one of the two towers goes as I want but in the second tower I have several traffic jams in the stairwells (It's not easy to explain. So, please check the pictures attached [traffic jams.PNG to see where the agents are blocked, evacuation after traffic jams.PNG this picture shows the moment when the agent decide to continue the evacuation and Quantification of traffic jams.PNG shows the curve of evacuation. There, we can see exactly the moment when some agents are blocked and the moment when they decide to continue their way to the exits] to understand better what I want to say). What intrigues me is that some agents (exactly 17) are able to evacuate from the second tower and I can't understund why the others are blocked. It's realy curious!!!!!! If you have any explanation for this, I would be happy to know it.
After that, I have done the same simulation but with a limited number of residents (2 per floor --> a total of 24 residents). the file attached "limited evacuation.pdf" shows the curve of evacuation for this simulation. As we can see, all the agents succeed to evacuate.
So, I don't know if it isn't the large number of residents that leads to traffic jams. Could you confirm this?
Thanks in advance Timo,
Anass
TimoK
Sep 4, 2015, 8:52:04 AM9/4/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Well, hard to tell without the input file, what is not going nicely.
Check, that the TO_NODEs are correct. And use COLOR_METHOD=5 and see
the avatar colors in Smokeview. Use, e.g., ellipsoid (or some other avatar that
supports coloring). See, if the agents have been able to figure out some
target door or not. And you can use COLOR on the DOOR lines do define
different color for different doors so you could us COLOR_METHOD=4 to
see, where the agents are heading.
Add KNOWN_DOOR=.TRUE. for all doors going downwards in the stairs.
This way the agents should have always a target door in the stairs.
And add the &SLCF PBZ=zz, EVACUATION=.TRUE., VECTOR=.TRUE., QUANTITY='VELOCITY' /
output at the z-levels of your stair meshes.
COLOR_METHOD:
-1) Default color for all agents
0) Evac-line color
3) Pers-line color
4) Target Door/Exit color
5) Door algorithm (7 colors)
Vki Fki Ski (visible, known, no smoke)
1 1 1 black
0 1 1 yellow
1 0 1 blue
1 1 0 red
0 1 0 green
1 0 0 magenta
dead and/or no door cyan
TimoK
Check, that the TO_NODEs are correct. And use COLOR_METHOD=5 and see
the avatar colors in Smokeview. Use, e.g., ellipsoid (or some other avatar that
supports coloring). See, if the agents have been able to figure out some
target door or not. And you can use COLOR on the DOOR lines do define
different color for different doors so you could us COLOR_METHOD=4 to
see, where the agents are heading.
Add KNOWN_DOOR=.TRUE. for all doors going downwards in the stairs.
This way the agents should have always a target door in the stairs.
And add the &SLCF PBZ=zz, EVACUATION=.TRUE., VECTOR=.TRUE., QUANTITY='VELOCITY' /
output at the z-levels of your stair meshes.
COLOR_METHOD:
-1) Default color for all agents
0) Evac-line color
3) Pers-line color
4) Target Door/Exit color
5) Door algorithm (7 colors)
Vki Fki Ski (visible, known, no smoke)
1 1 1 black
0 1 1 yellow
1 0 1 blue
1 1 0 red
0 1 0 green
1 0 0 magenta
dead and/or no door cyan
TimoK
Anass
Sep 4, 2015, 9:22:06 AM9/4/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
I'm glad to hear from you. The input file for the full evacuation is attached. For the limited evacuation, I have used the same file with limited number of residents. I have changed the number initial of persons from 24 per floor to 2. Could you please enlight me as quick as possible. Thanks a lot.
Anass
TimoK
Sep 7, 2015, 8:51:11 AM9/7/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Your geometries for the "left" and "right" stairwells are not the same. See the attached
figure. The agents jam at the narrow space between the OBSTs (red circle).
But your other stair geometry is not good. See the schemantic drawing in the same
figure, which shows how a staircase should look in an evacuation mesh level.
How to correct (easily) your stairs:
Make &HOLE XB=inside of the outermost red rectangle in the schemantic drawing
and set EVACUATION=.TRUE. and have the XB z so that you get this hole in every floor.
The make the "railing" using &OBST EVACUATION=.TRUE., PERMIT_HOLE=.FALSE.
so that the above HOLE does not take the railings away. And set the z-range so,
that one OBST line will make all railings (i.e., make a really tall obst).
Now you probably have a nice geometry for the stairs.
TimoK
figure. The agents jam at the narrow space between the OBSTs (red circle).
But your other stair geometry is not good. See the schemantic drawing in the same
figure, which shows how a staircase should look in an evacuation mesh level.
How to correct (easily) your stairs:
Make &HOLE XB=inside of the outermost red rectangle in the schemantic drawing
and set EVACUATION=.TRUE. and have the XB z so that you get this hole in every floor.
The make the "railing" using &OBST EVACUATION=.TRUE., PERMIT_HOLE=.FALSE.
so that the above HOLE does not take the railings away. And set the z-range so,
that one OBST line will make all railings (i.e., make a really tall obst).
Now you probably have a nice geometry for the stairs.
TimoK
Anass
Sep 7, 2015, 10:09:49 AM9/7/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
Thank you for your reply. Yes, you are right, the geometry of the two stairwells wasn't the same. I have corrected this right now.
I don't understand how to correct my stair geometry. As you suggested to me, I have added two &OBST in the stairs (see the file attached) but I don't understand where I have to add some Holes (if I have correctly understood your suggestion). Could you please explain to me (with an example if you can). I pray you to excuse my ignorance.
Anass
TimoK
Sep 8, 2015, 4:37:02 AM9/8/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Well, I try to have time today to find the XBs for the HOLE that clears the
stair space for the agents and XB for the OBST for the railings.
The problem that you have: The stairs are not well suited to the two
dimensional horizontal "floor geometry" that is "inside" the FDS+Evac.
Everything should be described using two-dimensional horizontal
evacuation meshes, these are like floor plans of your building.
In the floor plans, you have some part of the stairs in some plan
and other parts in other plans.
If you have modelled the stair flights in the fire geometry (like you
probably have, you use Pyrosim to get the fire geometry), then
some part of a stair flight will fit the z-range of your evacuation
mesh XB => this small part of the stair flight is solid OBST in
the evacuation mesh. This you do not want => use HOLE with
EVACUATION=.TRUE. to get rid of these.
TimoK
stair space for the agents and XB for the OBST for the railings.
The problem that you have: The stairs are not well suited to the two
dimensional horizontal "floor geometry" that is "inside" the FDS+Evac.
Everything should be described using two-dimensional horizontal
evacuation meshes, these are like floor plans of your building.
In the floor plans, you have some part of the stairs in some plan
and other parts in other plans.
If you have modelled the stair flights in the fire geometry (like you
probably have, you use Pyrosim to get the fire geometry), then
some part of a stair flight will fit the z-range of your evacuation
mesh XB => this small part of the stair flight is solid OBST in
the evacuation mesh. This you do not want => use HOLE with
EVACUATION=.TRUE. to get rid of these.
TimoK
TimoK
Sep 8, 2015, 5:27:04 AM9/8/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Check the "TimoK:" tags in the attached input file. It has the right staircase operating like it should.
You should be able to do the same things for the left staircase.
TimoK:
Stairwell:
&HOLE XB= 81,86.6, 103.5,106.0, 5.4,40.0, EVACUATION=.TRUE. /
TimoK:
Railing:
&OBST XB= 82.5,85.0, 104.7,104.8, 3.9,40.0, COLOR='RED', EVACUATION=.TRUE. /
TimoK:
MESH_EVAC_PALIER_R-1/0 (railing is also here, see above the z range)
&HOLE XB= 81,85.0, 103.5,106.0, 3.9,4.6, EVACUATION=.TRUE. /
TimoK: Now you could add the EVSS definitions for your
stair flights and intermediate landings.
TimoK:
x=82.0m PORTE_ESCALIER_A_0_IN door moved to x=85.0, where the doors in the upper floor are
(this door is the TO_NODE from the upper door, so it should be about at the same (x,y)
position, only the "one grid cell thick evacuation OBST" behind a door might change
the location little bit, i.e., at most one grid cell).
TimoK
You should be able to do the same things for the left staircase.
TimoK:
Stairwell:
&HOLE XB= 81,86.6, 103.5,106.0, 5.4,40.0, EVACUATION=.TRUE. /
TimoK:
Railing:
&OBST XB= 82.5,85.0, 104.7,104.8, 3.9,40.0, COLOR='RED', EVACUATION=.TRUE. /
TimoK:
MESH_EVAC_PALIER_R-1/0 (railing is also here, see above the z range)
&HOLE XB= 81,85.0, 103.5,106.0, 3.9,4.6, EVACUATION=.TRUE. /
TimoK: Now you could add the EVSS definitions for your
stair flights and intermediate landings.
TimoK:
x=82.0m PORTE_ESCALIER_A_0_IN door moved to x=85.0, where the doors in the upper floor are
(this door is the TO_NODE from the upper door, so it should be about at the same (x,y)
position, only the "one grid cell thick evacuation OBST" behind a door might change
the location little bit, i.e., at most one grid cell).
TimoK
Anass
Sep 8, 2015, 8:31:24 AM9/8/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
Thank you a lot for your help. Now, it's clear for me how to evacuate a stair with FDS+EVAC. I have done the same thing for the left stairwells and it seems that the things are going as it should.
Now, I will integrate the use of the outdoor ladders to evacuate the building ( see the picture attached to situate the position of the outdoor ladders and the file scenarioTK_2.fds ). I have only integrated the outdoor ladders (the stairwells are deleted . I will reintegrate them when the evacuation using the outdoor ladders works). I use "CORR" for modeling the ladders. But it seems that I haven't do this like it should. May be, I haven't understood how to use "corridors". Could you please enlight me?
Anass
TimoK
Sep 9, 2015, 4:53:42 AM9/9/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Well, see the attached PDF figure. It is the top floor of your geometry.
I did not check the direction of your ladders. I assume in the figure that
there is a DOOR with TO_NODE => CORR at the right and at the
left there is the COOR => ENTR (or DOOR, but ENTR is better).
Note also, that I have moved the XYZ of your "Door => Corr" door. Now the
"exit sign" can be seen better in the floor. So, the door is considered to be
visible for a larger area inside the floor. The default location just a little bit
in front of the floor is a little bit "behind a corner" for some agents in the
floor that should see the "exit sign".
No new meshes needed for ladders. The ladders can be part of the floor
meshes. You just extend the floor meshes a little bit and check that
the outer walls of your floors are there, i.e., OBSTs might be needed, if
you original model did not have the outer walls. The mesh boundaries
are by default solid walls.
Now the agents are freely "flowing" from the floor to the landings of the ladders.
So, the agents from the floors and from the above floor ladders (Corr => Entr)
are mixing at the landings.
NOTE: Now you will prefer the agents flowing from the floor. The Corr=> Entr agents
are put to the landing only if there is enough empty space in front of the Entr.
How to correct this? See the other figure. Now the Corr => Entr is more
efficient to put the agents at the landing. How much space you need?
Hard to tell. There is some experimental information on the merging flows
around. But for a fire safety reasons, you should not just try to model
one merging ratio, but use some different ones to see the effect of
the merging behaviour in stairs. A good rule of thumb would be to
assume 50 % to 50 % merging. This is a good first guess. But you should
try a couple of different ones also. Say, about 30% to 70% and the other way
around, 70% to 30%. The %-numbers are not exact, these are not input parameters
in FDS+Evac. But you can "play around" with them a little bit using "tricks" like in
the second figure, where the merging from Corr=>Entr was made easier.
How to set the flow from the floors smaller, i.e., the ladders are prefered?
With the geometry in the attached figures, the only way is to make the
hole to the landing smaller (but it should be not smaller than 70cm due to
the size of the agents). If this is not reducing the flow from the floor enough,
then you could put an EVSS (horizontal one) at the hole and a little bit
inside the floor, where you set the FAC_V0_HORI to a small value. This
should slow down the agents coming to the ladders from a floor.
NOTE: DOOR => CORR => DOOR or ENTR?
An ENTR is better, because you can have just one direction of
travel in the CORR. And in your case all agents are going downwards,
so no need for "upward going doors". You could use DOOR without
TO_NODE, but it might be, that it is not working due a small bug
in the program: The user input checking part had some problems,
that are now already corrected, but not yet in the release version
of fds. In the next release version a DOOR without TO_NODE should
work as well as an ENTR. The problem was: A DOOR without TO_NODE
was read in and then the program was in some cases stopped due to
an input check that should have just be used for doors with TO_NODE set.
TimoK
I did not check the direction of your ladders. I assume in the figure that
there is a DOOR with TO_NODE => CORR at the right and at the
left there is the COOR => ENTR (or DOOR, but ENTR is better).
Note also, that I have moved the XYZ of your "Door => Corr" door. Now the
"exit sign" can be seen better in the floor. So, the door is considered to be
visible for a larger area inside the floor. The default location just a little bit
in front of the floor is a little bit "behind a corner" for some agents in the
floor that should see the "exit sign".
No new meshes needed for ladders. The ladders can be part of the floor
meshes. You just extend the floor meshes a little bit and check that
the outer walls of your floors are there, i.e., OBSTs might be needed, if
you original model did not have the outer walls. The mesh boundaries
are by default solid walls.
Now the agents are freely "flowing" from the floor to the landings of the ladders.
So, the agents from the floors and from the above floor ladders (Corr => Entr)
are mixing at the landings.
NOTE: Now you will prefer the agents flowing from the floor. The Corr=> Entr agents
are put to the landing only if there is enough empty space in front of the Entr.
How to correct this? See the other figure. Now the Corr => Entr is more
efficient to put the agents at the landing. How much space you need?
Hard to tell. There is some experimental information on the merging flows
around. But for a fire safety reasons, you should not just try to model
one merging ratio, but use some different ones to see the effect of
the merging behaviour in stairs. A good rule of thumb would be to
assume 50 % to 50 % merging. This is a good first guess. But you should
try a couple of different ones also. Say, about 30% to 70% and the other way
around, 70% to 30%. The %-numbers are not exact, these are not input parameters
in FDS+Evac. But you can "play around" with them a little bit using "tricks" like in
the second figure, where the merging from Corr=>Entr was made easier.
How to set the flow from the floors smaller, i.e., the ladders are prefered?
With the geometry in the attached figures, the only way is to make the
hole to the landing smaller (but it should be not smaller than 70cm due to
the size of the agents). If this is not reducing the flow from the floor enough,
then you could put an EVSS (horizontal one) at the hole and a little bit
inside the floor, where you set the FAC_V0_HORI to a small value. This
should slow down the agents coming to the ladders from a floor.
NOTE: DOOR => CORR => DOOR or ENTR?
An ENTR is better, because you can have just one direction of
travel in the CORR. And in your case all agents are going downwards,
so no need for "upward going doors". You could use DOOR without
TO_NODE, but it might be, that it is not working due a small bug
in the program: The user input checking part had some problems,
that are now already corrected, but not yet in the release version
of fds. In the next release version a DOOR without TO_NODE should
work as well as an ENTR. The problem was: A DOOR without TO_NODE
was read in and then the program was in some cases stopped due to
an input check that should have just be used for doors with TO_NODE set.
TimoK
Anass
Sep 9, 2015, 9:33:58 AM9/9/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
I hope that I have understood your suggestion. So, I have removed all the new Meshes and I have extended the floor Meshes.I have added some OBSTRUCTIONS to model the walls indicated in the pictures you have attached in your message. I have also replaced the doors between the floors and ladders by HOLES. And I expanded the size of the ladders and I have modified the position of the corr=>Entr as you said. You can check this on the file attached. But when I run my script I have this error : " STOP: FDS was improperly set-up (CHID: outdoor_ladders)ERROR: Specify orientation of VENT 1, MESH NUMBER 2 (CHID: outdoor_ladders) "
I don't understand this error because I haven't any Vent in my file. Could you please enlighten me ?
Your faithfully,
Anass
TimoK
Sep 10, 2015, 7:32:34 AM9/10/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi!
A practical hint: Start to build the model bottom up. Just the first two lowest floors first. And for
the DOOR => CORR =>ENTR, you do like:
1st step: Put EXIts at the DOOR positions (no need to define CORR at all). See, if things work
like they should. (Well, you could do this for all floors, not just the bottom ones)
2nd step: Add ENTR in the input. See in smokeview, that these are at the correct position. You do
not see their IOR (no "small green cone"), but using COLOR='some valid color, not cyan'
you should see the IOR side (the face facing the empty space) as the given color. The face
facing the solid OBST (or mesh boundary) is cyan. (Well, if I remember correctly, might be that
the colors are not like that in Smokeview.)
3rd step: Change the EXITs to DOOR with TO_NODE =>
The error message means usually, that your DOORs or EXITs are not at a air/solid boundary. They are totally
inside a solid OBST or if they are at an air/solid boundary, then the IOR is not correct. The attached file should
run:
STOP: Set-up only (CHID: outdoor_ladders)
I have the T_END=0.0, so just the geometry is read in and processed. Nothing is calculated.
I added holes for your ladders:
TimoK:
&HOLE EVACUATION=.TRUE. XB=70.0,71.5,85.25,86.0,5.0,50.0, /
&DOOR ID='VERS_ECHB_A12', XB=71.5,71.5,85.25,86.0,33.0,33.5, XYZ=71.3,85.6,33.25, MESH_ID='MESH_EVACUATION_B_R12', IOR=1, TO_NODE='ECHB_A12', EXIT_SIGN=.TRUE./
And similar holes for other ladders also.
And your ENTRies were one metre too high. They were not in the given meshes, actually, they were not
in any evacuation mesh, so their XBs were wrong. I changed them all like below:
&ENTR ID='SORTIE_ECHB_A12', XB=71.5,71.5,85.25,86.0,30.5,31.0, MESH_ID='MESH_EVACUATION_B_R11', IOR=-1, MAX_FLOW=0.0/
ENTR ID='SORTIE_ECHB_A12', XB=71.5,71.5,85.25,86.0,31.5,32.0, MESH_ID='MESH_EVACUATION_B_R11', IOR=-1, MAX_FLOW=0.0/
TimoK
A practical hint: Start to build the model bottom up. Just the first two lowest floors first. And for
the DOOR => CORR =>ENTR, you do like:
1st step: Put EXIts at the DOOR positions (no need to define CORR at all). See, if things work
like they should. (Well, you could do this for all floors, not just the bottom ones)
2nd step: Add ENTR in the input. See in smokeview, that these are at the correct position. You do
not see their IOR (no "small green cone"), but using COLOR='some valid color, not cyan'
you should see the IOR side (the face facing the empty space) as the given color. The face
facing the solid OBST (or mesh boundary) is cyan. (Well, if I remember correctly, might be that
the colors are not like that in Smokeview.)
3rd step: Change the EXITs to DOOR with TO_NODE =>
The error message means usually, that your DOORs or EXITs are not at a air/solid boundary. They are totally
inside a solid OBST or if they are at an air/solid boundary, then the IOR is not correct. The attached file should
run:
STOP: Set-up only (CHID: outdoor_ladders)
I have the T_END=0.0, so just the geometry is read in and processed. Nothing is calculated.
I added holes for your ladders:
TimoK:
&HOLE EVACUATION=.TRUE. XB=70.0,71.5,85.25,86.0,5.0,50.0, /
&DOOR ID='VERS_ECHB_A12', XB=71.5,71.5,85.25,86.0,33.0,33.5, XYZ=71.3,85.6,33.25, MESH_ID='MESH_EVACUATION_B_R12', IOR=1, TO_NODE='ECHB_A12', EXIT_SIGN=.TRUE./
And similar holes for other ladders also.
And your ENTRies were one metre too high. They were not in the given meshes, actually, they were not
in any evacuation mesh, so their XBs were wrong. I changed them all like below:
&ENTR ID='SORTIE_ECHB_A12', XB=71.5,71.5,85.25,86.0,30.5,31.0, MESH_ID='MESH_EVACUATION_B_R11', IOR=-1, MAX_FLOW=0.0/
ENTR ID='SORTIE_ECHB_A12', XB=71.5,71.5,85.25,86.0,31.5,32.0, MESH_ID='MESH_EVACUATION_B_R11', IOR=-1, MAX_FLOW=0.0/
TimoK
Anass
Sep 15, 2015, 8:28:23 AM9/15/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
I followed your advices and now I have one outdoor ladder (see the picture attached : the outdoor ladder B_A is working) which work as it should. I have transposed the same thing for the other outdoor ladders (A_A, A_B and B_B) but it's not working. I don't know why ? could you please check the file attached ? Thank you.
Anass
Anass
Sep 15, 2015, 10:08:36 AM9/15/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Hi Timo,
I have resolved the problems. It was a problem of mesh size. Have a nice afternoon.
Anass
Anass
Sep 15, 2015, 10:14:40 AM9/15/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
I have one more question. What value can I take for the speed factor and the effective length for my corridors ? I suppose the effective length is the length of the ladder, is that true ? how can I calculate the value of the speed factor? Thanks a lot.
Anass
TimoK
Sep 16, 2015, 7:53:26 AM9/16/15
to FDS and Smokeview Discussions, rahout...@hotmail.fr
Yes, the effective length is the length of the ladders. Or you could use it.
The time that an agent spends inside the CORR is calculated as:
time_inside = EFF_LENGTH / (FAC_SPEED * v^0_i)
where v^0_i is the free (unimposed) walking speed of the agent on a horizontal
surface. Time_inside is the minimun time. The actual time could be longer,
if the TO_NODE of the CORR is "busy". This means that in front of the
ENTR (the TO_NODE in your case) there is no empty space to put the
agent. So, the agent stays inside the CORR and in the next time step
it is once again tried to put throught the ENTR.
Set also MAX_HUMANS_INSIDE to "correct" value. It should be about
the maximum number of humans that can be on the ladders. Depends
on the width and length of the ladders and the geometry of the stairsteps.
FAC_SPEED?
See, e.g., the SFPE Handbook, 3rd Ed, Table 3-14.4 on page 3-370
Maximum (Unimpeded) Exit Flow Speeds.
If your "ladders" are more like ladders than like stair steps, then I do
not know what the speed should be. Search the literature.
TimoK
The time that an agent spends inside the CORR is calculated as:
time_inside = EFF_LENGTH / (FAC_SPEED * v^0_i)
where v^0_i is the free (unimposed) walking speed of the agent on a horizontal
surface. Time_inside is the minimun time. The actual time could be longer,
if the TO_NODE of the CORR is "busy". This means that in front of the
ENTR (the TO_NODE in your case) there is no empty space to put the
agent. So, the agent stays inside the CORR and in the next time step
it is once again tried to put throught the ENTR.
Set also MAX_HUMANS_INSIDE to "correct" value. It should be about
the maximum number of humans that can be on the ladders. Depends
on the width and length of the ladders and the geometry of the stairsteps.
FAC_SPEED?
See, e.g., the SFPE Handbook, 3rd Ed, Table 3-14.4 on page 3-370
Maximum (Unimpeded) Exit Flow Speeds.
If your "ladders" are more like ladders than like stair steps, then I do
not know what the speed should be. Search the literature.
TimoK
Reply all
Reply to author
Forward
0 new messages