Flame Height and Plume Flame Height
690 views
Skip to first unread message
Sven Kolbe
Oct 4, 2011, 5:30:16 PM10/4/11
to CFAST
Hi,
sorry for asking again, but im a new User of CFast and some of the
calculation are not so clear for my mind.
Sorry for the bad question about the Lower Oxygen Limit, now I
unterstand how it works. I think in a complete wrong way and don't
unterstand that CFast only cut off the Fire at the given Point of
Oxygen.
But, now i there are two question which a canot solve by reading the
manuel. I try to solve it by reading other puplication, but theres no
sulution in it.
When i use the Correlation of McCaffey, there are three Zones in the
Plume (Flamming, Intermitied,and the Plume itselfs). I think Flamming
zone is the reaction zone. But why its diffrent to the flame height
Correclation which included in the modle. Or there are no connection
between.
An what happend when the smoke Layer descents unter the flame height.
Because, for my oppinion, there is no plume. Is the two Zone modle
still valid in this case.
And what happend when the flame is heigher than the room itselfs. When
i wachting it in smokeview, the flame is trough the ceiling. Is the
zone model still valid in this case (special when the flame is heigher
than the ceiling at the beginning of the simulation, or there is no
way to use the zone model in this case).
The last thing is, if the smoke layer has a temperature above 600
degree, there will be a flash over. For my opinion, the model change
in a complete room fire, and at this point CFast is not more valid.
Thanks for your help.
sorry for asking again, but im a new User of CFast and some of the
calculation are not so clear for my mind.
Sorry for the bad question about the Lower Oxygen Limit, now I
unterstand how it works. I think in a complete wrong way and don't
unterstand that CFast only cut off the Fire at the given Point of
Oxygen.
But, now i there are two question which a canot solve by reading the
manuel. I try to solve it by reading other puplication, but theres no
sulution in it.
When i use the Correlation of McCaffey, there are three Zones in the
Plume (Flamming, Intermitied,and the Plume itselfs). I think Flamming
zone is the reaction zone. But why its diffrent to the flame height
Correclation which included in the modle. Or there are no connection
between.
An what happend when the smoke Layer descents unter the flame height.
Because, for my oppinion, there is no plume. Is the two Zone modle
still valid in this case.
And what happend when the flame is heigher than the room itselfs. When
i wachting it in smokeview, the flame is trough the ceiling. Is the
zone model still valid in this case (special when the flame is heigher
than the ceiling at the beginning of the simulation, or there is no
way to use the zone model in this case).
The last thing is, if the smoke layer has a temperature above 600
degree, there will be a flash over. For my opinion, the model change
in a complete room fire, and at this point CFast is not more valid.
Thanks for your help.
CFAST Development
Oct 5, 2011, 9:21:44 AM10/5/11
to CFAST
In CFAST, the flame height calculation and the plume calculation are
totally independent of on another and simply based on different
empirical correlations that are documented in the Technical Reference
Guide for the model. The flame height calculation is based on the
correlation by Heskestad and is simply calculated at the time of
printout. It has no impact on any other part of the model. A
calculated flame height higher than the ceiling indicates that the
flame would spread out over the ceiling, but the model does not
account for this.
CFAST includes several algorithms to account for different burning
conditions. When the base of the fire is in the lower layer, the
standard plume correlation is used to calculate entrainment of air
into the plume that is then transported to the upper layer. If there
is not enough oxygen entrained to completely burn all the fuel mass,
this unburned fuel is also transported to the upper layer where is may
burn if there is sufficient oxygen in the layer and if the temperature
is above the specified volatilization temperature. When the layer
descends lower than the base of the fire (again not related to the
flame height calculation), then there is no lower layer entrainment
and only the upper layer combustion takes place. An additional source
of combustion takes place at vents with entrainment calculated as if
it were a normal fire (again assuming there is sufficient oxygen and
the temperature of the doorway flow is above the volatilization
temperature.
I hope this helps.
Richard Peacock
totally independent of on another and simply based on different
empirical correlations that are documented in the Technical Reference
Guide for the model. The flame height calculation is based on the
correlation by Heskestad and is simply calculated at the time of
printout. It has no impact on any other part of the model. A
calculated flame height higher than the ceiling indicates that the
flame would spread out over the ceiling, but the model does not
account for this.
CFAST includes several algorithms to account for different burning
conditions. When the base of the fire is in the lower layer, the
standard plume correlation is used to calculate entrainment of air
into the plume that is then transported to the upper layer. If there
is not enough oxygen entrained to completely burn all the fuel mass,
this unburned fuel is also transported to the upper layer where is may
burn if there is sufficient oxygen in the layer and if the temperature
is above the specified volatilization temperature. When the layer
descends lower than the base of the fire (again not related to the
flame height calculation), then there is no lower layer entrainment
and only the upper layer combustion takes place. An additional source
of combustion takes place at vents with entrainment calculated as if
it were a normal fire (again assuming there is sufficient oxygen and
the temperature of the doorway flow is above the volatilization
temperature.
I hope this helps.
Richard Peacock
Sven Kolbe
Oct 11, 2011, 11:49:35 AM10/11/11
to CFAST
Hi,
thanks for your fast answer.
But, there are still some open points. I read the users manuell two
times, same with the technical referecne guide.
But i didnt find a answer, as well as not in different publications on
the internet.
One question is about the plume models used in Cfast. When i use the
Heskestads Plume, there is a diffrent entraiment below and above the
Mean flame hight. In CFast, as you say by yourself, there is no
connection Between the Flame height and the Correclations for the
Plume. But it seems there is one, because in the many puplications
there is a Conection between Mean Flame height und Plumes.
The next thing is that in all puplications i read the MacCaffey plume
was used to determine the Temperature und Volecity, but not the air
entraiment. So form where are the correlations for the air
antraiment.
So a all i have key Questions
1. ) When the Flame is heigher than the room, can i used CFast or is
the modele incorrect for this case
2.) When the Smoke Layer descents under the Flame height, what
happend ? (is there a Plume, because when i read from your statemend
above, there will be only combustion, but no plume). So is the modle
incorect for this case (Because in the literatur all the Plume
corrlections are testet without this case. The tested models are not
included in Cfast)
3.) Is the Plume in CFast a Part of the Upper layer all the Time ? Or
is it a seperate Zone
4.) When i take a look into the .out File, there is a column for Fire
in Upper in Lower Layer. How is it incooperated in the Modle ?
5.)For Calculting the Clol there is a entraiment rate Used. Which
entraiment rate is it ?
Thanks for your effort in advance.
Maybe some one of you worked with CFast an can give me a Answer,
becaue i have to decide can a use CFast for my Problem or not. Because
if not Maybe i use FDS, in this means longer Calculation times at
all.
Thank you
Sven
> > Thanks for your help.- Zitierten Text ausblenden -
>
> - Zitierten Text anzeigen -
thanks for your fast answer.
But, there are still some open points. I read the users manuell two
times, same with the technical referecne guide.
But i didnt find a answer, as well as not in different publications on
the internet.
One question is about the plume models used in Cfast. When i use the
Heskestads Plume, there is a diffrent entraiment below and above the
Mean flame hight. In CFast, as you say by yourself, there is no
connection Between the Flame height and the Correclations for the
Plume. But it seems there is one, because in the many puplications
there is a Conection between Mean Flame height und Plumes.
The next thing is that in all puplications i read the MacCaffey plume
was used to determine the Temperature und Volecity, but not the air
entraiment. So form where are the correlations for the air
antraiment.
So a all i have key Questions
1. ) When the Flame is heigher than the room, can i used CFast or is
the modele incorrect for this case
2.) When the Smoke Layer descents under the Flame height, what
happend ? (is there a Plume, because when i read from your statemend
above, there will be only combustion, but no plume). So is the modle
incorect for this case (Because in the literatur all the Plume
corrlections are testet without this case. The tested models are not
included in Cfast)
3.) Is the Plume in CFast a Part of the Upper layer all the Time ? Or
is it a seperate Zone
4.) When i take a look into the .out File, there is a column for Fire
in Upper in Lower Layer. How is it incooperated in the Modle ?
5.)For Calculting the Clol there is a entraiment rate Used. Which
entraiment rate is it ?
Thanks for your effort in advance.
Maybe some one of you worked with CFast an can give me a Answer,
becaue i have to decide can a use CFast for my Problem or not. Because
if not Maybe i use FDS, in this means longer Calculation times at
all.
Thank you
Sven
>
> - Zitierten Text anzeigen -
CFAST Development
Oct 11, 2011, 3:00:49 PM10/11/11
to CFAST
In a "real" fire, I'm sure there is a connection between the flame
height, entrainment, and other plume properties. In CFAST though,
there is no direct connection. The flame height calculation is just a
side calculation done at the time of the creation of the printout. In
CFAST, the plume calculation determines how much oxygen is available
for combustion. While the correlation has an implied dependence on the
fire size (by virtue of the heat release rate) and flame height (which
has a role in determining the steady burning, intermittent burning, or
plume region of McCaffrey's correlation for example), the flame height
is not used in the actual calculation, just the emperical plume
correlations. In the lower layer, the entrainment is determined by
the net heat release rate constrained by the available entrained
oxygen (an iterative calculation). In the upper layer, the
entrainment is driven by the heat introduced by the lower layer
burning plus any burning which can take place due to air entrained
from the upper layer. Thus, the plume calculation is indeed used in
both layers.
To attempt to answer your specific questions:
1) There is no exact number, per se, where the model cannot be used.
Certainly, a very large fire in a compartment, as would be indicated
by large flame height, is an indication that care should be taken in
using the model. However, there are a number of post-flashover fire
examples in the validation tests for CFAST, quite often with excellent
results.
2) Most typically, the fire will not descend totally below the base
of the fire as specified by the user since some entrainment must take
place to transfer mass and energy into the upper layer in the model.
Thus, for fires high in a compartment, again, care should be taken in
using the model. There is little is any experimental data for
validation of fire high in a compartment.
3) The plume in CFAST is used in both layers. For McCaffrey's plume,
for example, an appendix to the journal article referenced in the
technical reference guide specifically details how this is done.
4) The 'Fire in Lower' and 'Fire in Upper' are simply the amount of
heat released by the fire due to entrainment in the lower layer
(lower) and from burning available oxygen in the upper layer (upper).
One or the both may be zero depending on the available oxygen
entrained from the control volume.
5) The calculation in the technical reference guide for the lower
oxygen limit for burning is for both layers, done separately for each.
height, entrainment, and other plume properties. In CFAST though,
there is no direct connection. The flame height calculation is just a
side calculation done at the time of the creation of the printout. In
CFAST, the plume calculation determines how much oxygen is available
for combustion. While the correlation has an implied dependence on the
fire size (by virtue of the heat release rate) and flame height (which
has a role in determining the steady burning, intermittent burning, or
plume region of McCaffrey's correlation for example), the flame height
is not used in the actual calculation, just the emperical plume
correlations. In the lower layer, the entrainment is determined by
the net heat release rate constrained by the available entrained
oxygen (an iterative calculation). In the upper layer, the
entrainment is driven by the heat introduced by the lower layer
burning plus any burning which can take place due to air entrained
from the upper layer. Thus, the plume calculation is indeed used in
both layers.
To attempt to answer your specific questions:
1) There is no exact number, per se, where the model cannot be used.
Certainly, a very large fire in a compartment, as would be indicated
by large flame height, is an indication that care should be taken in
using the model. However, there are a number of post-flashover fire
examples in the validation tests for CFAST, quite often with excellent
results.
2) Most typically, the fire will not descend totally below the base
of the fire as specified by the user since some entrainment must take
place to transfer mass and energy into the upper layer in the model.
Thus, for fires high in a compartment, again, care should be taken in
using the model. There is little is any experimental data for
validation of fire high in a compartment.
3) The plume in CFAST is used in both layers. For McCaffrey's plume,
for example, an appendix to the journal article referenced in the
technical reference guide specifically details how this is done.
4) The 'Fire in Lower' and 'Fire in Upper' are simply the amount of
heat released by the fire due to entrainment in the lower layer
(lower) and from burning available oxygen in the upper layer (upper).
One or the both may be zero depending on the available oxygen
entrained from the control volume.
5) The calculation in the technical reference guide for the lower
oxygen limit for burning is for both layers, done separately for each.
Sven Kolbe
Oct 13, 2011, 2:20:35 PM10/13/11
to CFAST
Hi,
i am a little bit confused, because in a other post describes a
complete different way of calculating the entrainment. It states that
if the fire is in the lower layer, the entrainment is calculated from
the top of the mean flame height to the interface layer. When the fire
is in the upper layer, is the distance between the base of the fire
and the ceiling. And so, there is a connection between the flame
height und the plume.
So, I tried to understand the process with your very good
documentation, but I don’t get the point.
So I tried to use a little example, maybe, the situation is clear
than.
I hope someone can give me answer.
So I have a Room, 20 meters width, 10 meters deep and 5 meter height.
There is a fire located in the middle of the room with a HRR of 900
kW. The area is 0.2 m².
So I try to make the plume Calculation by Hand. First, I see that
there a two formulas used. In the documentation the correlation from
Hesekstad is the fowling:
Me = 0.071 * Qf,c^1/3*(z-z0)^5/3+(1+0.026*Qfc^2/3*(z-z0)^-5/3)
In all other documentation I found the correlation is state as :
Me = 0.071 * Qf,c^1/3*(z-z0)^5/3 + 1.92*10^-3 * Qf,c
And there is a other correlation if z is below the Mean flame height.
The correlation in CFast is net valid below z=1 because its incorrect
for a negative base with a rational exponent if the exponent is odd.
Maybe I make a mistake in some way, but that is what I spot.
If a use the MacCaffey correlation the flaming region is up to 1.2
meters, the intermitted from 1.2 to 3 meters. So, my mind the reaction
is finished there and the plume begins to clam up. So the Plume begins
a 3 meters.
After 20 seconds the interface layer is at 4.26 meters.
So I try to calculate the air entrainment.
I build up an Integral from the correlation used in CFast with the
boundaries 3.06 and 4.26 because I think that’s the entrainment
region.
CFast states that there is a entrainment of 6.45 kg/s. I don’t
understand this value, because the result from my calculation is much
higher. I am complete wrong with my way ?
So, CFast works very well and so I try to understand where my
failure is
Thankyou
Sven
On 11 Okt., 21:00, CFAST Development <cfast...@gmail.com> wrote:
> In a "real" fire, I'm sure there is a connection between the flame
> height,entrainment, and other plume properties. In CFAST though,
> > > standard plume correlation is used to calculateentrainmentof air
> > > - Zitierten Text anzeigen -- Zitierten Text ausblenden -
i am a little bit confused, because in a other post describes a
complete different way of calculating the entrainment. It states that
if the fire is in the lower layer, the entrainment is calculated from
the top of the mean flame height to the interface layer. When the fire
is in the upper layer, is the distance between the base of the fire
and the ceiling. And so, there is a connection between the flame
height und the plume.
So, I tried to understand the process with your very good
documentation, but I don’t get the point.
So I tried to use a little example, maybe, the situation is clear
than.
I hope someone can give me answer.
So I have a Room, 20 meters width, 10 meters deep and 5 meter height.
There is a fire located in the middle of the room with a HRR of 900
kW. The area is 0.2 m².
So I try to make the plume Calculation by Hand. First, I see that
there a two formulas used. In the documentation the correlation from
Hesekstad is the fowling:
Me = 0.071 * Qf,c^1/3*(z-z0)^5/3+(1+0.026*Qfc^2/3*(z-z0)^-5/3)
In all other documentation I found the correlation is state as :
Me = 0.071 * Qf,c^1/3*(z-z0)^5/3 + 1.92*10^-3 * Qf,c
And there is a other correlation if z is below the Mean flame height.
The correlation in CFast is net valid below z=1 because its incorrect
for a negative base with a rational exponent if the exponent is odd.
Maybe I make a mistake in some way, but that is what I spot.
If a use the MacCaffey correlation the flaming region is up to 1.2
meters, the intermitted from 1.2 to 3 meters. So, my mind the reaction
is finished there and the plume begins to clam up. So the Plume begins
a 3 meters.
After 20 seconds the interface layer is at 4.26 meters.
So I try to calculate the air entrainment.
I build up an Integral from the correlation used in CFast with the
boundaries 3.06 and 4.26 because I think that’s the entrainment
region.
CFast states that there is a entrainment of 6.45 kg/s. I don’t
understand this value, because the result from my calculation is much
higher. I am complete wrong with my way ?
So, CFast works very well and so I try to understand where my
failure is
Thankyou
Sven
On 11 Okt., 21:00, CFAST Development <cfast...@gmail.com> wrote:
> In a "real" fire, I'm sure there is a connection between the flame
> there is no direct connection. The flame height calculation is just a
> side calculation done at the time of the creation of the printout. In
> CFAST, the plume calculation determines how much oxygen is available
> for combustion. While the correlation has an implied dependence on the
> fire size (by virtue of the heat release rate) and flame height (which
> has a role in determining the steady burning, intermittent burning, or
> plume region of McCaffrey's correlation for example), the flame height
> is not used in the actual calculation, just the emperical plume
> correlations. In the lower layer, theentrainmentis determined by
> side calculation done at the time of the creation of the printout. In
> CFAST, the plume calculation determines how much oxygen is available
> for combustion. While the correlation has an implied dependence on the
> fire size (by virtue of the heat release rate) and flame height (which
> has a role in determining the steady burning, intermittent burning, or
> plume region of McCaffrey's correlation for example), the flame height
> is not used in the actual calculation, just the emperical plume
> the net heat release rate constrained by the available entrained
> oxygen (an iterative calculation). In the upper layer, theentrainmentis driven by the heat introduced by the lower layer
> burning plus any burning which can take place due to air entrained
> from the upper layer. Thus, the plume calculation is indeed used in
> both layers.
>
> To attempt to answer your specific questions:
>
> 1) There is no exact number, per se, where the model cannot be used.
> Certainly, a very large fire in a compartment, as would be indicated
> by large flame height, is an indication that care should be taken in
> using the model. However, there are a number of post-flashover fire
> examples in the validation tests for CFAST, quite often with excellent
> results.
>
> 2) Most typically, the fire will not descend totally below the base
> of the fire as specified by the user since someentrainmentmust take
> from the upper layer. Thus, the plume calculation is indeed used in
> both layers.
>
> To attempt to answer your specific questions:
>
> 1) There is no exact number, per se, where the model cannot be used.
> Certainly, a very large fire in a compartment, as would be indicated
> by large flame height, is an indication that care should be taken in
> using the model. However, there are a number of post-flashover fire
> examples in the validation tests for CFAST, quite often with excellent
> results.
>
> 2) Most typically, the fire will not descend totally below the base
> place to transfer mass and energy into the upper layer in the model.
> Thus, for fires high in a compartment, again, care should be taken in
> using the model. There is little is any experimental data for
> validation of fire high in a compartment.
>
> 3) The plume in CFAST is used in both layers. For McCaffrey's plume,
> for example, an appendix to the journal article referenced in the
> technical reference guide specifically details how this is done.
>
> 4) The 'Fire in Lower' and 'Fire in Upper' are simply the amount of
> heat released by the fire due toentrainmentin the lower layer
> Thus, for fires high in a compartment, again, care should be taken in
> using the model. There is little is any experimental data for
> validation of fire high in a compartment.
>
> 3) The plume in CFAST is used in both layers. For McCaffrey's plume,
> for example, an appendix to the journal article referenced in the
> technical reference guide specifically details how this is done.
>
> 4) The 'Fire in Lower' and 'Fire in Upper' are simply the amount of
> > > into the plume that is then transported to the upper layer. If there
> > > is not enough oxygen entrained to completely burn all the fuel mass,
> > > this unburned fuel is also transported to the upper layer where is may
> > > burn if there is sufficient oxygen in the layer and if the temperature
> > > is above the specified volatilization temperature. When the layer
> > > descends lower than the base of the fire (again not related to the
> > > flame height calculation), then there is no lower layerentrainment
> > > and only the upper layer combustion takes place. An additional source
> > > of combustion takes place at vents withentrainmentcalculated as if
> > > is not enough oxygen entrained to completely burn all the fuel mass,
> > > this unburned fuel is also transported to the upper layer where is may
> > > burn if there is sufficient oxygen in the layer and if the temperature
> > > is above the specified volatilization temperature. When the layer
> > > descends lower than the base of the fire (again not related to the
> > > flame height calculation), then there is no lower layerentrainment
> > > and only the upper layer combustion takes place. An additional source
CFAST Development
Oct 13, 2011, 3:18:47 PM10/13/11
to CFAST
Again, in no case is the mean flame height used in the calculation
other than in an indirect way to determine which of McCaffrey's
correlations to use. The height over which entrainment takes places
is either 1) the distance between the base of the fire and the layer
interface (for fires in the lower layer) or 2) the distance from the
layer interface and the ceiling (for burning in the upper layer). the
term z/Q^2/5 is used to decide which of the three McCaffrey
correlations to use (continuous flame, intermittent, or plume). The
only inputs are the height over which entrainment takes place (z) and
the nominal convective heat release rate (Q).
Heskestad's correlation is a bit different, but gives very similar
results over then range of use. You are correct that it is possible
for the Heskestad calculation of virtual source location to be
negative. In CFAST, we use a check to make sure it is not less than a
very small number.
You may want to examine the code for the calculation. It's available
on the CFAST Google Code site at
http://code.google.com/p/cfast/source/browse/trunk/cfast/trunk/CFAST/Source/fire.f
> ...
>
> read more »
other than in an indirect way to determine which of McCaffrey's
correlations to use. The height over which entrainment takes places
is either 1) the distance between the base of the fire and the layer
interface (for fires in the lower layer) or 2) the distance from the
layer interface and the ceiling (for burning in the upper layer). the
term z/Q^2/5 is used to decide which of the three McCaffrey
correlations to use (continuous flame, intermittent, or plume). The
only inputs are the height over which entrainment takes place (z) and
the nominal convective heat release rate (Q).
Heskestad's correlation is a bit different, but gives very similar
results over then range of use. You are correct that it is possible
for the Heskestad calculation of virtual source location to be
negative. In CFAST, we use a check to make sure it is not less than a
very small number.
You may want to examine the code for the calculation. It's available
on the CFAST Google Code site at
http://code.google.com/p/cfast/source/browse/trunk/cfast/trunk/CFAST/Source/fire.f
>
> read more »
Nico
Mar 1, 2013, 2:36:38 AM3/1/13
to cf...@googlegroups.com
Hello,
I have a question concerning the plume entrainment.
I have checked the difference between Heskestad and McCaffrey's correlations and as it has already been emphasized in other messages, there are some differences. I am not going to discuss about that since it is clear that two different empirical correlations might provide different results in some conditions (depending of the range of application).
My problem is that I tried to estimate the rate of air entrained using the equations for Heskestad and McCaffrey's plume that are provided in the Technical Reference Guide and my hand calculations do not match at all with the predictions from CFAST (output "Plume Entrainment Rate" output from CFAS_n.csv). The fact that the output integrate the rate of air entrained and the mass loss rate can not explain such difference.
To performed the hand calculation, I took the layer interface from the CFAST_n.csv file.
I went to your code and as far I can see, the equation are the same so my question is why I cannot found back the rate of air entrained? What is the z that you take to perform the calculation? My question is not related to validation but to verification.
Basically, I want to understand where my understanding of your code is not correct.
Moreover, I realized that the virtual origin is affected by the width and length of the fire object but not by the fire size defined in the table. Unless, I have not seen it reading the user guide, I think this would might be worth to underline it in the user guide.
Thank you in advance
Nicolas
Reply all
Reply to author
Forward
0 new messages