You are on page 1of 30

~"~~

l; ..

,01:= U

~---

THE LIBRARY

'I

FIRE Ri'5EAR,_H STATiON


nORt}-lf\i~1

WOOD

I~~::

419~960

tloF. R.
No.
---Research Programme Objective E1/9

- DEPARTMENT OF SCIENTIFIC AND INDUSTRIAL RESEARCH AND FIRE OFFICES' COIdMITTE&


JOINT FIRE RESEARCH ORGANIZATION
.-

ROOF VENTING OF BURNING ENCLOSURES


PART III.

P. H. Thomas,

VENTING FIRES OF CONSTANT HEAT OUTPUT

-,
D. L.

by
S~',
.-:.-

.-

P. L. Hinkley and C. R. Theobald

This work has not been published and


should be considered as confidential advance
information.
No reference should be made to
it in any publication without the written
consent of the Director, Fire Research
Station, Boreham Wood, Herts.
erelephone: Elstree 1341 and 1797).

"

This work has been done in conjunction


with Messrs. COLT VENTILATION LTD.

Summary

Investigations using a 1/12 scale model representing one bay of a


factory have shown that, where the area or the rire is small and its heat
output is known, use can be made of Bernoulli's theorem and the theory of
turbulent Plumes to calculate the rates of discharge of heat-and air through
a vent and beneath a curtain extending from the ceiling, the temperatures
near the ceiling and the depth of the layer of hot air.
Formulae are givenfor the effect on these quantities of varying the size of the vent and _the
depth of the curtain.
Some of the effects of the shape and position of
the vents are also discussed.

February 1960 F.1 000/1 2/346

Fire Research Station,


Boreham Wood,
HERTS.

BRE Trust (UK) Permission is granted for personal noncommercial research use. Citation of the work is allowed and encouraged.

Notation and units used in this aerie.

Symbol

Dimensions
Conversion
I--------......,.--------t
faotor used
P.P.S.

Meaning

unit.

F.P.S. unitl

He'

Effective area of fire'


Area of inlet
Area of vent
Coeffioient of
disoharge through the
vent
Coefficient of discharge beneath the
curtain
Specific heat at
constant pressure
Depth of layer of
hot air
Depth of layer of
hot air beneath
curtain
Acceleration due to
gravity
Height of opening
Height of model
Depth of curtain

H~

H + 1.5

'i

C d
C
p

d
d

h
H

r-r:
At,

(effect~ve

1
At

To
t
V

929
929
929

B.T.U Ib -1

F-1

it

cm

30.5

it

cm

30.5

ft/s

30.5

it
ft
ft
, ft

30.5

cm
cm
cm
cm

B.T.U.s-1 ft- 1 F-1 Hal


~t

30.5
30.5

30.5

'~:\iDi:10c:1
cm

lb/s

'gm/s.
C8.1/s
cal/s

B.T.u/s
B.T.n/s

Oa

OX

OR

OK

s
ft 3/s
ft/s

cm3/ s
cm/s

--ft
it

15
30.5

om

it

Subsoripts
H = Near the ceiling
c
Level with the bottom of the curtain
o = Ambient condition
s = Discharge under the edge of the curtain
v
Discharge throUghtbe vent
y = At a distance, y, above the floor
b = At the base of the hot layer
i = Inlet

to
C.G.. S. unit.

height

of model)
Thermal conductivity
Characte ri s ti c
height dimension
Path length of air
in model
Mass flow rate
Rate of flow of heat
Heat output of fire
Absolute temperature
" ambient
"
Time
Volume flow rate
Velocity of gas at
any given point in
'prototype and model
Width of inlet
"
Distance measured
vertically from
floor
Temperature above
ambient at any given
point in prototype
and model
Density

C.G.B. units

"

30.5
454252
~52

0.555
0.555
2.84 ; 104

30.5

em
em

0.555

0.0160

ROOF VENTING OF BURNING ENCLOSURES


PART III.

VENTING FIllES OF CONSTANT HEAT OUTPUT

by,
P. H. Thomas, D. L. Simms, P. L. Hinkley and C; R. Theobald
"

1.

;"

.,.

:.r

......

,';

':~'.

-,

Introduotion '

1/12
A previous note(1) desoribed the construction of'a
scale model of
a bay of a factory for use in studying the effect of venting burning
enclosures.
The present paper desoribes the experiments that have been
made with this model to examine the infiuence of the" size, 'shape and position of the vent; the depth of' any auxiliary curtains and the'shape and
area of the heat source on the ,temperature ,distribution in :~the 1I10deland
the heat flow through it.: The results are shown to agree :wi,th c::alculations based on a suplified theoretical approach to the ven,H~g of fif8S"
derived from Bernoulli's theorem and the theory of turbulent plumes 2).
The application of the theory,tolarger and differently shaped enolosures, larger fires and the design of venting systems ..Will. be discussed, in
later papers.
,..
,';
;

.:

-:",

..:
."

Experimental method and results

"

2.1. Apparatus and procedure


Tl).e model is shown in fig~ 1.
Essentially it is an asbestos
box insulated on the outside with exfoliated vermiculite and lined
on the inside 'with aluminium foil.
Heat was supplied by various
types of' gas burner (rig. 2) the gas input being regulated to Jive
a gross heat output of 2.37 B.T.U/s.
The earlier experiments~3),
show,ed that there were two time constants governing the equili~rium
temperature rise; a short one associated with the air flow and a '
long one of less importance associated with the walls.
In eacjh '
experiment the walls were allowed to reach equilibrium.
lIeasurements
were then made of the air temperatures at various points inside the
model and of the distribution of air velocity and temperature along
the vertical centre line of the inlet.
The temperatures were
measured using ,44 S.W.G. chromel-alumel thermocouples distributed as
shown in fig. 1A except for a ,few of the early 'experiments where they
were shown in fig. 1B.
Air velocit:j.e(l were measured with a radiation compensated Simmonds anemometer\4).
Since the horizontal dis.,.
tributions of air velocity and temperature across ,the inlet were uniform e~cept for some reduction in velocity within 2 in. of either
sideO), the air input to the model and the output of air and heat
under the curtain oould be obtained from the velocity and temperature
distribution, along, the vertical c~ntre of the inlet.
2.2. Experiments ,and. results
Two series of experiments were carried out; inth,e first the ,
shapes of burner and vent were varied; in the seoond the size and ,/
position of, the curtain and the size of the vent were varied.

...

2.2.1.
"

';'"

Shape of burner and vent

Three types of burner and three' shapes of ,vent wore used in


, these experiments.
The burners (linear, circular and double
line) are shown in figs. 2A, 2B and 2C respectively. , The vents
were rectangular haVing different proportions but the same area
of 24 in 2
Details of these experiments are given in Table 1,
whioh also gives the air temperature near the ceiling and the
rates of air input and disoharge under the edge of the curtain,
which were obtained from velocity measurements, as explained'
above.
,

,.

, " TABLE 1
:)~x~rl..ments'.' on eft'eot Or' shape or vent 8114 burper
,~ -. . ,
,(Firat series)
,

"r, ,

'.

, . ".'
,

,
Vent
position
H: over
he,"ter
"I,: ' Dear
inlet

Vent
dimensions
in. x in.

4 x

Heater.,

..

22~571.': '
1;1

.... -

'Ra:tl' or
diBo~ee
or ,air
thrOugh
,vent, (by:'

..

0.05
0.10

,Jfeasurecl ,
_'temperature '
,- near oeUbg
OF

diff'erence)
"ft3/s
'
,
.

'

288
270

0.37' "
,0.31

' Line
Circular

'

0.40
'0.42

0.10
0~15

0.30
0.27

,8
, 8

0.39
0.43

"

'"

Double liDe

t -.

iI'-',',

"

~.

0.42
0.41

8'
8,

"

H-

"

........

'"

ORate or
disoharge
or air
under e~
of ourtain
ft3/s

0.13
0.16

0.26
,0.27

0.05
0.05
0.32
0.35

0.33

243
243
288
275

1:

8
.;8

-,
..,

,-

,
.

"

'Circular

"

,4 x

'

..
Line
,.. Circlilar

,"

Line
,

H,
.

Bate of
air input
ft 3/ .

depth
in.'

"

. ,. :

'

-.

"

..
"

"Curtain

12 x ~

....

"

ff

"

",

'

."

c'

I,:'
r"

..

..

12
,12
.."8
-.. 8,

"

',ft
.-

..

'.ft

,.. ,

",

'

..~:

0.38
0.38
0.59
0.64-

!'

0~33

221
220
,,64162

. 'O~27

:0,,29
..

..

"

.'

".

','

Volume meaaured at ambient temperature.


'-

fThermooouples arrlUl8ed as in fig. 1B.

~ermooouples arranged as in fig. 1A.


t.

"

"

Obtained from 'meas~ments of'air velooities.';'''


'.'

'

..

.\

of S3iGI\1,n

2.2.2. Sise and psa;!.tioA o( vEl!!.t epA !l9RIib

The f'actors ~ild in thisseoom arias of' uparimeQta were'


the poaition of the vent, the vent area and the depth of' the ~
Throughout, the ventwaa .approx1lllateJy square -and the oiroular
heater was used.
A sUlJllllBl"Y of results is given in fable 2. (pap 4).

,.1. fest et

-,.'

,
t.

peW.:!;;!.o,n o( unt,

't . ~

The rate of disaliarg~ of' hOt air throush the vent (Table: 2), was,' ;', ';,:
gena'fl.lly gt'ElBter when the ventoWas .over the heater thanwhen' it-was near':
the inlet, althou,gh the differenoe ,was only important with the 1 in. Oul-',:
tain.
In that osse it Dan be explained by the hot gas having a veloaiV'" ;
suffioie1'1t to oause a tendency to overshoot the vent when this ,W8,S. near
;
the inlet (fig. ,) and to pass under the curlain.
This resulted in ii -"" ~::
oorrespondillly high rate, of discharge under the ,ou.rteJ.xt..
~:. .
, .: "_";
.

.:

The position of the v~t


near the oeiliD.

had

..1..

"'.

.. '

:-:

-.

"

no Irl,gniticant ef':feot on' the, t~~~"~ :


'
"'
.. :: ": ~, ,
~.

'.2. J?!'!p::1buU0A of ye1o1 t.y beata tRa

i\tl1NJ!.

.,t
"

In all the above experiments the rates of' air:flow -were

t .

~ouiB.tea.

:from veloeity measurements; a typical distribution of velooii;y 'beneath


"the curtain is shown in fig. 40,
The velooity near the f'loor of' the, ,", ',:

mpdel was 0.5 ft/s with the linear andoiroular burners' ana. 0.65, fila With
the double line burner.

El:am1JIation of the air f'lew pattern in the inlet uaiJlg a BIIIOk:e trac"
showed that in those experiments where there was a high rate, of discharge
of hot air beneath the aurtain a vortex existed between the inf'lowiJlg and
outflowing air streams (fig. 5).
This sometimes led to an overestimate
of the rate of flow of. air into the model as the non-direational anemameter
measured the speed of the air in the vortex instead of, the mean hor:l.llontal
velooity.
'

The aoeural:\Y of the experimental measurements of air :flow and tlllllperature may be estimated _by oompar:l.ng values of' the heat passing tbroush the
vent oalculated in two different way.,.

(a)

(Qt) lass the heat discharged under the edge of the curtain ~Q); Q is obtained,
:f'rom measurements of air temperature andSVelooi'-r beneath the

From the net heat input to the'model

ourtaino

(b)

.9.v: =

~ - Qa

~__

( 1 )

r-

Prom the temperature and the mass of gas vented, obtained by


subtraoting the rate of discharge of air under the curtain f'rcmI
the rate of input of air (Tables, 1 and. 2).
."
"

Q' v=
'l_ C (M - M)
'-lip
is

(2)

In fig. 6, Q' is compared with~.


The sOBtter is r",ther large,
extending from aboi{t -25 per oent to + 50 per oento
,
"

The rate of flow of heat through the v-ent could. not be' obtained'from
the temperature and velooity of the air flow through the vent because the
anEllllometer calibration was uncertain at' the high air temperature existing
there.
However, the above comp~n shows that the vent heat :flow Dan
be obtained :f'rom measurEllllents of' a:!,ll! velooity and temperature beneath the
curtain without systematio error exoept that aome very high values of Q'
were obtained.
These are attributed to the difficulties in measuring tKe
air input due to the vortex desaribed in the 'previous section.
Beeauae
the first method ilWolves fewer measured quantities in the air streams, it
has been assumed that it is more reliable than the seoond method.
:

,.'

..

1'ABLB 2

Be.u1ta of experi!!enta on aue and position of vent


(Sec0n4 aerieal

.:.
'-

,
Curtain

depth

Vent area'

abe -of curtain


-

-,

"

ta2

in.

tD4

*Rate' of
air~t

41scltarp,
of air -.

n/ts,

16
12

24

..

,~

"

..

H
I

16

24

,I

40

H
I

"
"

:.1

:I

16

H
I
H

24-

40

.,
,',

.:- ;.

..
,,'

..

~~~~'_f~ 1IIll~UI:ements

H
I

of-

0.26
0.36
0.29
0.40
0.35

0.01
0.03
0
0.01
0.03
0

0.4-1
0.4-2
0.4-9
0.4-2
0.4-6
0.4-8

0.15
0.22
0.13
0.05
0.03

0.85
0.67
0.75
0.93
0.75
'0.80

0.51
0.53'
0.4-6

0~29

"

';

;40

.. 'I

..

0~07

0~55

.:j

wi.

"

0.28
0.23 fj
0.36 '

....
,

-,

"

-,

,~

"

340

354:
286
283
'243

_
'

0~28

-~';;_::,"'":.' s :".:,-.'

,: ._.,~:p:.::'l, .

0.37
0.35

0~74-

0.34-

Dear

ceili.DB

thrOu&h

"

.-

vent (by
din'erence) ,

uD4ercurtain
"
tt3/s:

Te.p,rature

.. - 41s~~'
of air \..'"

"'teof,
- Veftt poeitioft '
H: ever boater
I: near in1et

hate or

"

24.8

0.26 ~
0.20, '
0.36
0.37
0.4-3
0.4-1

299
308
279
262
232
241

0.340.11.
0.29
0.19
Q.4-1
0.25

183
183
191
180
165
176

-,

,.:

',: !,j,,,;
"
',I: .' .1

: '.'

"

" ;.

"

ho1~e ..,uured at8lllbie~t temperatUre.

~.:

..'

';;.

,,
"
"

-... roo:
\' ,

,;:

"

,I

'

"
r

~':'

.\

..

:~

c,

,..

" ,

'-,"

z,
.,'

ve1ocit,-.
'"

'

~f;

'.: .

',

3-:40 .Distribut;i.on ot, ~e.mppWe ;la Sie fII238l,


,

' .

..

. ...

'

o.

~ '~~

A.-:,typioal vertical temperature distribution 8S reco~' b;y-~.tliermooouple8


1'-12 (figo rs). is shown in fig. 7.
There iSQ layer ofhot'atr"beneath the
oeiling, a layer of cool air near the floor -ends .central,. layer'where the hot
and cold layers are JIli#ng.
The variation in tempera.ture! ~v~ any horizontal
plane',ll the mooel was found 'to be small compared W;~ .the, vert:L~ variation it
" temp.era.'ttJ.r e &
-.
-, ;
' -. '
. ".
."
'. ..
3.. ,5. Ef't"ect .otabape~-am.j?08it.iQA, of vgnt
v

'

Both ~ -shap&.&Jd poaiti~n .-Of the vent had _op:).:y ~..aliiht effect on the
temperature -within the enolosure'"-ana. the heat,vented from it; 'there tended to
be sli-gh~mor-e discaar-ge beneath the curtain-when' the ,v~nt 'was near(itO'
These results are very-similar to those obtained in other experiments 1 -' 5).

4.

tsnm~

Calcu:I.a;t!0BS of ,air flow-a,n,d

The foregoing experiments have. determined the heat-and air flow through the
vent and under the curtain.and the temperatures near ,the roo~ for ~ous oonfigurations of vent and curtain in a small scale -modelo
In this section a simple tbeor.Y'
is developed which enables these quantities to be predioted, i.l?- 'certain cases, from
the dimensions and geometry of the model and the size of the fire, and it is shown
that the experimental results are oonsistent with this theoryc> .
.

disclJ,a.rfl~

Co'efficients of

4.10

;'

"

..

In the following argument the pressure drop due to the loss of velocity
head of the inlet flow beneath the ourtain is negleoted..
This is juatif'ied
sinoe the layer of hot air beneath the gailing was at least 4 in. thick and
its temperature was of the' order of 180 F while the pressure drop across the
inlet was equivalent to a head of only about Oa2 in. of air at this temperature.
The rates of discharge of hot gas through the vent and under the curtain
may be calculated from the vertical distribution of temperature in the model.
by using Bernoulli's Theorem (Appendix I)..
It is, however, found that the
a.ctualdiacharge' measur-ed experim~tally is less than the theoretical by a
factor whj.g{l is known as the coefficient of discharge. . This is usually
0116 - 0 0 7l 6 ) but may vary with the size and shape of the orifice and the
velocity and orientation of the tlowo

4.2. Determipation of coefficient of discharge through the vent


It is shown. in the appendix that

. '.
"

. '.

Vv

..

(:.

Cd

= A 't/

. The integral (H 9/T


I

.)0

00 0

To

may be evaluated gra~hically by plotting 'IT

against y.
This was done for several different conf'iBurat1ollBof vent and
ourtain (fig. 8), and by inspection of the graphs, it was found that to a first
approx:ima.tion

0 0 0 0 0,

"

(4)

where d was measured to the Pj)int of in.f'lea:tion of the temperature distribution


curve, ioe. the shaded area 'in fig. 8 is approximatel.y equal in area to the
rectangleo
Hence .~guatio~ (3) simplifies to
,

V"
C=
d

0&0000000.

(5)

VJ(' as oaloulated from equliUon. (5) is plotted against the measured


rlteJ.of c11scharge of gas through the vent (fig. 9) (obtained fran the'
subtraotion of the meas~d mass spilled under the ourtain 'frOm the .
measured mass input (Tabl~ 2.' The results for those l;lXp8riments'
.
where the disorepancy between Q.. 'and Q' 'exceeded 25. per oent are"
omitted.
Lines oorresponc11ng ~o ooeft'IoientB of disoharge of 0.6 and
0.7 have been drawn on this graph. . AJ.though there' is a large soatter,
the experimental results are consistent with the assumption that the.'
ooefficient of c11soharge is about 0.6 and this value .is assumed in all
subsequent caloulations.
In applying the theory to full scale
'
builc11ngs and to vents c11ffering great],y in shape from any discussed .
in this note, other values. for the discharge coefficient may be ~quired.
4~3.

Determination of coefficient of c11scharse under. the curtain

It' the velocity head of the gas in the upper part of the enolosure
may be neglected the ourtain behaves as an inverted weir. (Appendix)

;..-10 ~tl, ~) di

....,..... (6)

This equation was evaluated graphioally for a few. oases from' the
experimental temperature c11stributions and it was found that a s1a11ar
approximation to that given in the previous seotion,oould be uaed,if .
the bottom of the hot layer extended at least about 2 in~' beloW .the
. bottom of the ourtain.
Then as a first approximation .equation (6)
reduoes to
.
.

v
_

~.

C d'"

,.' 3 . ..

=
2
'd' -~2 /'
/2g "H'
.. -w
3.
o,.~.

<v

S/C' a

.,

"

.......

. : ,:. ~ .:,

f.

'

, .

".
~

(7)

was plotted against the measured volume disoharged over the edge
of the ourtain (fig. 10) . Again there is a large soatter, but there
,
is no systematio variation about a line correspcinc11ng to a value of 0~6"
for the' discharge coefficient, though two reSUlts in 'particular are of
doubtful value.
4.4. Rate of entrainment of air

The air flow through the vent and under t;he curtain may be oal-.
culated from equations (5) and (7) respectively provided that Qg and d
are known.
These latter quantities may be'calculated by oonsidering
the rate at whioh the hot air layer is replenished by the rising plume
of hot gas and entrained air from the burners (fig. 1.1)"
The effect
of the rate of input of coal gas in these experiments (0.0175 ft3/s) may
be negleoted, in oomparison with the muoh' greater rate of fJ,ow of air
into the hot layer.
The mass of air (14) entrained in a distanoe (H' - d) by hot gases
.riSing(ffOm a point source of heat may be obtained from a f01'l!liUa given
by Yih 2) which in our notation may be written
. ".

Il = 0.153 fH
.

Q g
C T

{fH

1i (H(-d/h

(8)

Sinoe the fire is not a point source a correotion may be introduoed '(')
based on the angle. of spread of the plume' as' suggested by .G,' I. Taylor 7

H~ = H +1.5 .;-;:;

. (9)

The evaluation of Af, the effective area of the fil-e, 'ilic11soussed in


the following seotibn.
. .

1+.5. Caloulation of temperature beJieatb, the oeji .."


The temperature beneath the oeiling can be oaloulated by 8U~Sti tutin8

in equation (8)

'\I

(9u + T

. !.

6 . ,'i

..l =

,~?

c.2.

)..

. i'

?t

(11) .

0.000

1-

.,2

o
(~o p)"3 (gl' 0)"3 (H .a:)~
.
Since the effective area of the fire in these experiments was nOt oertain,
values of 9u calculated from eqy.ation (11) were plotted against a for a
number ofailbitrary values of H' and compared with the'experimental points
. plotted in the same graph (fig. 12). .:The most satisfactory agreeaent
. between theory and experiment was found i f HI,. was. taken to be 25 in~, which
gives At ,;, 21 'in2 from equation (9)."

ev

Aa the vaiues of both


and 9H obtained using the ciroular burner did
not differ significantly from. those obtained using the linearburne~ with
.the same Vllnt and curtain, it fol,lows.,that the ~ear burner had approximately
the same effective area as the cirCular one.
.

The values of d and 9H obtained experimentally using the double line


burner were put into equation (11) and it was found that this burner had an
effective area of 100 iJi2 0
-:',
.
J+~6. ,.Calculation of depth 'of la.yer of hot 8as

The depth' of the la,yer cit 'hotgas be!1eath the, ceiling is obtained by ,
equating, the mass of hot il.irentering the layer. to that disoharged thrOugh
the vl)nt and under the curtain (fig. 11). "
"
.When there is nodisc~rge under; the

ourt81n

,.

(12)

.11=11

hom equation (5)


. 11,/

and from

=c

equatio~ (8)~

fRy?g d~,i
"

' (13)

. To,'.,

(10), (12)'imd (13)

CdAv d =.0.043 (H-d)2


.

" .

. '

Th~ relevant experiaientiu.result~ are oOllpared with equation (1J+) iD fig. 13."
Equation (14) will not apply if d}:1Io sinoe therato .of discharge of hot " .
il.ir beneath the ourt&1nJlust then be t8kenintoii.ocoUnt~ ,
, "
.,

:'

,,\....:..

.11 = ~ +

".

.'

"

............ (!,s)

lis
'

..

.. .
. .....
~

..

"

, ,7'

"

and from equations (8), (10),(13), (15)8.nd(16)


,

Cd A d
V

32 Cd~'
w (d - HO)
,

, """,
~

'= 0.043 (H,-d)

The relevant experimental results are oompared with equation (17) when
d).Ho and with equation (14-) when d(H in t'ig~ 14-.
The agreement
1:ie_n th~ory and praotice is very sa~isfaotory in view of the
approximatiom
made.
'
- ,
I:t there is no ve'nt open and all the heat is disoharged beneath
,the ourtain equation (17) reduoesto
'
':
'
"

"

2-

C~'W (d:- Ho)2 = 0.065 (H~od) 2

r.
"(;,,.

4-,7. ,Rates of .disoharm of air


4-.7.1,'

Throufsh vent - no spill under 'ourtain

'The theoretical rate of disoharge th1-oughthe 'Vent when


-;;'" ,:;there is .no spill under the ourtain is obtained by putting the
',values of d and Qj{ obtained from aqUationa (W) and (11) in aqua';';,<tion (5). ' The variation in rate of'disoharge with, ventaroa '
'oaloulated in this w8J and, oorreoted for disobarge ooeffioient,
is oompared with the experimental points in fig. 15~,' ' , ,
4.7.2.

upder,ourtain- no vent

,The theoretioal rate of disoharge una.er the curtain when


,there is no vent is obtained by putting the values of -d and 9H
obtained from equations (18) and (11) iD. equation (7)~ The
variation in rate of discharge nth ourt&1ndepth oaloulated in
this W8J and oorreoted for disohargeooet:rioient is oompared with
the experimental points ,in fig.',16.
In'both figs 15' 8.nd16, the
agreement is, satisfaotory.
4.8. Rate of disoharm

of

heat through the

vent

The theoretioal rate of disoharge of heat throil.gh the vent is


, obtained, by putting the values of d and Qj{ obtained from equations (17)
, and (11) in the equation (19) below whioh is derived from equation (5),

Qv = Cd ~, eH fH C/'26 d
"
P.
T
.

'

9H

The variation' in the theoretioal value ~ heat disoh~rged with vent


area is oompared with the experimental points ,for two ourtBJ.n depths
in tig. 17.' Again, the agreement between theory and practioe is
satisfaotory.
"
'
4.9. Summsl'V
of Seotion 4'
,

It has been shown in seotion 4 that it is 'possible to oaloulate a'


, number of quantitiell allsooiated with the heat and ,mass flow in the,
Ilodel and, although there ill in some oases a wide sOatter in the"data,
the faot that no signifioant systematio differenoabetween,observed and
measured quantities has been found shows, that the methods ~ oaloulating thelle quantities, based only on the heat and mass, bal&noe using
elementary buoyancy oonllideratioas and plume theory, are satiBfaotory.

\,

. 5.

Applioations of theoq
..

A th~ory was developed. in the previous section .and it was ":fQ~nd, by experiment to apply to a particular model"
This section shows tha~ the theory may be
expected to apply to larger scales of the 'same mode1 and, in some. cases, to other
.shapes of.model~
.

The~. 18 nothing in the foregoing theory that restricts it to any particular


scale, provicied .that the scale 'is large eno.ughtorthe flow to be tur~ulent and
that external wind conditions do not influence the flow.
There is a restriotion
on the shape of the enolosure beoause theexperiments. have been oarried out with
a flow which is essentially two-dimensional, 1 0 e . the- gases flow along the model
ox:'. in" a vertioal direotion, but there .is.no radial tlow.,;
.This is because the assUmption was made.in the .theory that the temperature
is a funotion ,of the height above the floor but is '~therwise, independent of .
position; i.e. no mixin~ ~ccurB between' the hot and cold laYers. . Work on two-'
~mensional flow systems~8} has' shown that .little:niixing "would be expected to
.
ooour even when .eaoh of the layers is t.urbulent unless the relative velocity is
quite high.
Consideration ~llbe givento the.problem of radial flow in future
work.
.
-5.1. 'Soaling
,For geometrically similar enclosures of "linear dimension L, the velooity
of the ex! t hot gases due to th~ buoyancy
of. the, hOt layer is of the form
,

:;.-v

,and' the

/TtL. " ".. ..; ';.

:X

..

..

',.'.:.'
,

..~

".:.

1~.

.,.' ',
. .' . :

. ~.

:~.'

.. .i,e ' ,*

(20) .

-, ',"-,

'

heat' balanoe is :ofthe form.-

......

0'

Examples of equations similar to equatd.on (20r are (1) - (6) and similar t'o
(21) is (19).
Other equations can readi~'be shown to .be combinations.of
these two types.
From equations (20) and (21),.

\.'

< . '

.. ,.

.~'

(J"

.:J -x

L3

. 0 0 . 0 . ,

(22)

0 0

(23)

Equation (11) is of this type

x,.oc. i!
TI

Li

Equation (8) is of this type .(strioe M(~:v L2 ) 0 .


These., ~J.ations have been confirmed for. the' present modei'-u~ing different heat
input8~3).; they can be used to extend the .range of equations, ,(8) - (18) when
the linear scales are varied.
These. soaiing laws are derived for' an enc.Iosure: in:'wMch the exit .is the major
resistance to flow.
For more general systems the .relations between L, ,~, g and
Qarethe same but their dependanoe on T will be different and usually less.
~o avoid this complication it is desirable to keep the .. temperatures ina model the
same 808 in the prototype which is b~ing modelled.
.
. ~,:.-:. ,~.,
l'

~,

~. ~.r ~~: :~

r.

9
. '.:.::. I:

'

....

.....

\r,.-' -
. . . . .~~ .

'5.2. Fioor

...

are!
~

"

..

'Bquationa (1-13) do not oontain the term ,"floor area", so that noll1-

nally they m~, be applied to buildings of an:y si.e.. However, as t11e


,floor area 1.nonIases, mixing of' the hot and, oold laYers inofOases ~d

u.v be ilJ.ightly less effeotive as their distanoe from the ftie


inoreallell(1.51.
A similar etfectmust be expectedU heat losses by
conduction and ,radiation through the walls and oeilings are' ot, ~;Lativoly
greater iIIIportance than in this, model.
In general, on lal"ge bUildings.,
one needonJy oonsider the oeiling.
This problem has yet to be
'
investigated.
"

ve~ts

5.3. mect of area of fire


, The depth of the hot layer hlis been caJ,culatedtrom equation C14-) as
a f'unC1;ion of the ettecUve area of' the fire, assWlling no discharge"
bei1eaththe ,curtain and is shown iil fig. 18, for. various vent areas.
, Asth~ ettectiTe area ot the tire increases' the depthof' the layerof'
hot air :1l1creases" and i f discharge beneath' the curtain is to be pre'vented, larger vents are required.
With fires of' large area the vent
area required m~ become impractically large; and the discharge of' hot
gas can onJybe prevented by restricting the air inlets.
~",
5.4-. Bttecit of inlet openiIls of' limited sizeoD temperature and air flow
,

'

Under the oonditionsOf' the experiments desoribed in this paper, it


is possible to,negleot the ettect of the pressure drop across the inlet,
and therefore to regard the neutral pressure plane as coinoidiilg with
the bcittom ot the"hot gas layer.
In generaJ.., ,hoWever, the, neutral, '
planel!ill be higher than the bottom of the hot' gaB l8yer in order to
.'. l'rorlde &buoyanoy head to balance the velooity headaorossthe iDlet .'
,{fig.19). ' ,Equations to allOw for this cianbe easily derived, but are '
" generally dit't'ioult to solve.
There is a simple solution when the
bottom .of'the hot air layer is above the bottOm: of, the oUrtain and there '
is no spill under the ourtaiD; ,the mass rate of' flow of gas through the
model is thenl (See A~enciix 2h_ _' _

2", cll~

II=CdA;.d
, '

o;~ T (1 + ~
H

Equati:on (14-) then becomes


.
.,'' . 1~..

Cd d
,

" f

,
A.2~H

Av

=0.04-31c--2
""
, ....1

"

, .

T)

(24)

.,(25)
"

,To

,,'

Thus, it the size of the ,inlet is reduced so that TH' ~ cannot be neg, lected inoomparison with
the l~er of hot
,
'air will be
,
inoreaaod in thickness and hence from equation (11) its temPerature
will be higher.
A simple solution is also possible ,when it can ,be
assumed that,theCmQciol is fullCOf)hotgas at~'i. Thip Q~se has,bee~'
treated by Yokei 9) and Dyakov 10
There is ..vidence'11} that this "
approxi.mation can be used when a fire covers most of tho floor area
of',the model although it did not appear to hold for' the lf~rtments
reportedolsewhere with a large coal gas fire ,in a model\5)., These
experiments, therefore require some future 'oonsideration and this will
be:reported later.

6.

Conolusions

, The eiperiJllents .have shown that, the ehape of theverit is not important
and that it~poSiticn within a small curtained area is' oJu,y of aiilor iltPor,o,
tance 'provided that the hot lrlr does not have a suf'ficiently:highhon-'
Bonta! velooi1;y to cause e. tendoncy to overshoot the 'vent.
These quanti~ative resultsoolltirm. the qualitative results obtained by earlier experi_
ments using a',smoke tracer, and it can therefore be inferred that a Uletul
guide m~"be obtained from simple experiments with a smoke tracer.,
10

The layer of hot gas beneath'the oeiling 'can be, regarded as uniform in
and Bernoulli's ,theorem, together with' the theory of turbulent
plumes}oan be applied tO,estimate its depth and temperature.
This leads to
the following formulae for predicting the efficiency of a vent when the pressure drop across the inlet is negligible and the air flow 'is linear in a
horizontal plane!
temperatu~

Depth of layer ,of hot air (d)

(a)

0.6 A d 2",0.043

(H!'';'d)~

(fig. 13)

when d (H

O.~Ac
dt + O.4'd(d:'H )~ '" 0.043(Hic.)~(fig. 14)
, v
c'
-',,'

(b) 'Temperature of hot air layer~IlH)


2 '

.:

6.5 Q,~3
.

(foCr)(gTsrt
,(c)

(HI-d)~

(fig. 12)

Rate of discharge of heat throughthe vent,


'-'-~'''''

,"

'Q" '06'" 6 .. / " To \~ ',!2g ,d~Il,H


ov'" , ' fo "r \T +9 I "H
.'

When the
":" used,

O.

HI.

:)
...

. To .

,"

}l,

(fig. ,17)

!lV

..

mOdelis:fu~10~hoig;;:~tli.e ari~lySiS~f'YO~Oi(9) may be


.....-, ,,;,-, .

~.

,"

"

"'.'

, There is ,satisfactorY agreement', betweerime~ureci: and calculated data


and it is therefore possible to design' a,ventiD:g system provided the heat
output of the, fire and the, ,various dimensions ,of the building are known.
Conside'rationwill be given to s.ome problems..of de'sign' in a later paper
.uafng .the' methods given here. '
'
AoknOwledgment
.. ..
'

The authors would liketo'thanJiMr;', R. ,W. Olaridgefpr his v8.luable


assiStanoe with the experiments' and 'Miss V. 'MoNaughton for undertaking the
. oalcuia~':i.ons~,.,- .. '
. ..... '
. <, .
"

,"

'

"

:\

1. 'SIMMS; D. L., HINKLEY, P.L., EindBisSETT, A;' ,Roof venting of burning


, Elnolosures,'Part 1.
Joint
Research Ore;lmization., F. R. 'Note No.
',"
'390/1959.
'
.
'
."

Fire

-.,'

.~

2.' YIH, C; S; , 'Free conveotion due, to a point sourceof"heat.


.. ,:,
, U.S. Nat. Congr.Appl. Meeh, 1952, 941\-7;

3.

Froc. 1st

SIMMS, D. L., HINKLEY; P.L; and BISBFrT, A.


Roof venting of burning
enclosures Part II. '. Joint Fire Research Organization. F.R. Note No.
391/1959;
,
R. H.
A radiation' compensated anemometer.
Ore;anizatiori. F. R. Note No; 297/1957. ,

4~

~IGMORE,

5.

THOMAS, P. H. and SIMMS, D. L.


on the prevention of flame spread beneath
ceilings in fUlly developed fires' by roof vents.
Joint Fire Research
Ore;anization F. R., Note No. 376/1959.
,

6.

Joint Fire Research

LEWITT,~. H.

Hydraulics and the mechanios,of riuid8; London, 1958,


( 10th Edition), Sir Isaao Pitman and Sons 'Ltd.
" ,
11

TAYLOR, G. I.
The dispersal of fog from airfield runwil,ys. ed.
'WALKER, E. G. and FOX, D. A. Ministry of SupplY. London, 1946, p.230
et seq.
8.

Some Interim Notes on Methane Roof ~rs.


Safety in Mines
Research Establishment Research Report No. 164 Feb. 1959.

BAKKE, P.

YOKOI, S.
A
construction.
No. 29, March
10.

DYAKOV, N. M.
Influence of area and space layout of building" ,
construction in relation to direction of fire spread.
Collected,
information on fire resistant'building construction. "'Central Research
Institute for Fire Protection, Moscow, 1958.

11.

SIMMS, D. L., HINKLEY, P. L. and THEOBALD, C. R.


Roof venting of
burning enclosures, Part IV.
Joint Fire Research Organization; F. R.'
Note (in preparation).

.:

",'

_:'.3 ._

12

_'

Appendix I .

')

-, v. .....

i .

Derivation of formulae for rates of flow of air 'under the curtain


,',

1.

,I

'

..

and through the vent' , ..,

"

_'0'. " .... "

...~ . '.

:~

'Disoharne i ofhot'air through the vent (equation 3)

The' 'air .'is' regarded 'as being stationw.7 inside the 'model. :The ';p~iB'~'re' .
beneath the ceiling (PH) differs only slightly fro,m that. outs.ide ,the JI104el,.:{po}
so that it is possible to negleot the change in density of the air with 'pressure
on passing through the vent.
Bernoulli's ~quation (whioh.is given together with
examplfi)8' of. its . use in m8J'137 text books{ 6)} may then be ~itten in a simplified
form.

,,.:,
(26)
.
..
.'.
.

. . : .r'.

-~,

. ,'';'~.'~

~.'

-:

:.

To obtain (% - p ) it is assumed that the temperat~, and hence the density,


of the air in themodel is a funotion of the height above the floor (y) but is
othetrwiae indepel1dent ,of position.
Then (referring to fig. 20)
.

"~

~~

f;:

't~

rl

e
e:i

.)

Ceiling

.PH
-:"1---

iI
l~j

.. '

"

doc-

Column of
air of ~t
oross section

,
]

:. <,

,.;.

",

'

..

Floor

Fig. 20
But

pr

r:

'oTo

==

P.a

a'

~.~' ~ ,. ~ .' . (28)

and T - T

=,G

._.

:.::. ~ "\

.':10:

(26, 27, 28, 29)

Then from equations

,~

. .~

::

.\:~J

,"

. -

,'..

"'~

,.

.:~

.. ,r:,

"~"'''~r

........

', (30)
"".~

..

"""'

'-_-~I'

Now

c..

&llowiD6. tor the coeffioient of discharge as

. . . . ".

(~1>,

explain~d

.~_

ip. sectioJ1 4.1 .

Bquation (3) can be derived from (30) and (31)

v .- Cd

~.);;: f9/T<\Y

.. ~! " .

',t

To

Disoharge
of hot air beneath. the ourtain (equation 6)
.
i
.
i
i
..

.....

Consider a horillont8.l strip of the opening beneath the curtain at


distanoe, r, above the floor and having an area W dy (fig. 21)...

Ceilins
Curtain
'(width w)

' . ' "vr. - .,...+...--

-. ".
- - - - I - -. - - - - -

I
I

Outside air
p

(,.)

:')

Floor
l'i6. 21

'nUs may be regarded as a vent at a height, y, above the floor and


t"rom equation (3) the disoharge (dV) thrOUgh the strip will be
.
W

(1

dy I-g i'i ""


I

. dV = Cd

\I

WT

dy

........... (32)

To

and the disoharge ~hrou~~ the ~ntire opening (Vs ) will be


H
HI'
~
r-

'i."

Vi'

r.

= CdJodV ..

. This 1s equation

i (J

C~ w'

ai /2 g ~t '&~ itT & '

. To

..

)dy

6~

,"

"

air

" p;privation of formula for mass flow 'of

. .. ,

...

through the model '

, (e9uati~n 24-) ,
-

"

A ~Pr.1I1~~~~ .. ~~er~ ot)~ot air of. tempera:ture Ta, densi ~~Pu; ~ depth ,'"i:
a <.(H - .ao > i8 assumed to exist be~eath 'the ceiling., The restUof. the .odel '
'90ntains UJ"
at T0 and D~o (Fig. 22).'
"
.
'
, '.

......;~'; ~:~,~.;~.

~..:,:i:-

,-

Ceiling

\,.;,

'

",' -

.J . ,
c;.

Outside
.~)

....

"

'.
,"

'air

"

~oin'l~

. . '.' ....

:- '. ". '.

.-r....
~

..,
,--'1,.-

. . ~.~.

-- -

,U

-.' '~.

r""~_ ....,
-'''ri
'

H'
.

~-!--

'

-_:.-.;..--"
,

<.;

.:
,,'

'.:

..

---.0-1...........;._'--

~.-

'f1

t1

Pressure

distribution
~sid8

model'

I,

..

Appendf!II

1'.

Fis.'22

Par the disoharge through the vent, Bernoulli I 8 equation m~ be written

, l' +'H'2g
1V

.,

<,>.

0,

....

t,

,_

=ta

and tor air entering through the inlet beneath the ourtain

.......

,
,

t) ~d)
: 0
I h

+ (J

.3!L
2g

'(3~)

ITo. (33) and (34)

,ftJ - fl, =
But

PH

.2

~H.lL +
2g

eo 2g'*'~

,?.

(35)

~Ft ~ (. eo - eH)d
eo,To
'9,

...........

= ,~ TH'

= TH -

(36)

(37)

........ , (8)

To

~n trom equations (35, 36, 37, 38)


.~.~

Vi 8D~ Vv OaD Qe obtained in terms .of:~~he,~&88 flow (M) and the area, of ~
vent and i~et (A", and Ai)' s~poe._, ."..
I

~ - ,~.::.

.:

_,

"

_.~.

"

,.'

:f

II = Ca" Ai fa "i. .:.'_...._.

_<' _ :

,":

. '

.~'. :.:~ ~'.

.;,. -a_;;

..

~. (4-9)

. ,. . ..,~. 'c'dJ, A,; fa:~', =C


'Av" Po0 '!2'l'v
(fram 37)""" ~'~.
'. ~ .:.;.'-' (41:)
d .-,
TH. .. . ' : '
::,', ,..' )J

!L

-t

r";

",

Provided tha~ ~he shapes aJl.d areas of the vent,


greatly trolD e~oh other,

,.:

D9t cUtter

>

C ~d.
and

~d ,"nle~ .40

.A.,

~-

. e-.-.

tro. t39, 4-0, 4-1, '4-2) equati~n .(24) oan

(2;)

.2
'\
{1 +!L TH ~
"-... . AV2To.'"
-

;L'. "

,I

.v

(~)

be derived

28 :49ft

+,."

....

-,

"

":

. .,
..

,1.

."
......

'.~:"

"

"

.;

'
I

.~. -! . :

..

.;

.......

. '.:
"

.''''

-1/3578

F.R.419

Roof tray
Depth
of curtain

Roof tray

1,90,100,110
I

2 3

x90

of curtain

-5

I
I

6 7

x eo

I
.1

)(6

Depth

x1

10

I(

8,11

xl2

1
.l

t -

)(5

10 II
r1-----1...~----------J
Probe
ELEVATION

xl4

ELEVATlON
I-thick glass fibre

I
I

x "90

I
I
'I

1
I

x 100
x r -'11
x 110

-15

I
1

.~ J 90-12
,

-x2
"x

-x

)(14

)(10
-Vent

)(6

-j

-I
PLAN

FIG.IA. DIAGRAM OF MODEL SHOWING


POSITION OF THERMOCOUPLES
FOR MOST EXPERIMENTS

PLAN

FIG.IB. DIAGRAM OF MODEL SHOWING


POSITION OF THERMOCOUPLES
IN SOME EARLY EXPERIMENTS

II
1

FIG.2A. LINEAR

BURNER

~G"

0'

FIG. 2B. CIRCULAR

CI5

Jets

BURNER

- - - - - - . 2' - - - - - -~l
,!

. +t

; _ \L

-+

4-

...

.... +

+! \

rantail burners

.+

1\

,.....

...

FIG.2C. DOUBLE LINE BURNER

Shallow curtain -vent over hecter (No


tendency to "ov~rshoot" vent)

'" "

----.,..-.-.-,.-~~

----..-...

~C"~
--.R.~Pidiy movin~
~

"

'

Olr

"

'

Shallow curtain - vent near inlet


.""...
(Considerable tendency to 'oversnoot vent)

Slowly moving
air

" I

Deep curtain ~vent' near inlet' (No"


tendency
,
. ' to ov~r5hoot vent)
.
-

F'IG.3. TO

ILLUSTRATE TENDENCY OF AI"R STREAM


TO OVERSHOOT THE VENT

11 3 58 1

F.R 419

----:l----:::::::::::===-EE:ddiOcgeZ"Oiof curtain

o
l!J

UJe

> .-

~I

<{~

llJ

~...J

IZ
,,UJU-

IO

-ln -Out

05

05

15

VELOCITY OF AIR -It/s

FIG.4. TYPICAL VELOCITY DISTRIBUTION


IN INLET (CIRCULAR BURNER)

:e - -.....
&

Hot air

- - " .Cold .oir

FIG.S FLOW INSIDE MODEL WITH


GAS HEAT1NG

'W

'I-

---

~>
""0

2-0

-:
-- l/
.V

........
O~
0: 0

lJ..~

'5

~a:

...

["T"

+
+

LLl <t:'

>lL

IO

l?

I' 0

::>LLl

" 0 a::
c!=>

\Dill

J:~
.... 0::

.~ ~g~

l11'

Ill!

.5 /

'LLuJ"r

~=
<to =!

I-

WZ~

IdJ

()

r-o

05

2 5

20

1-5

-2

HEAT FLOW THROUGH VENT MEASURED BY DIFFERENCE 0v-B.tu. ft

VENT

CURTAIN DEPTH -in

AREA - sq in.

Circular

Distributed

burners

burners

24

16

40

16

40

24

8
12

- -+
+

Ii

++

OF Ov WITH, O~

FIG.6, COMPARisON

18

J.

12

0::

lL.

IJ.J

()
co

lI

cczillng
.

/'

16

/'

----~

.:

-I

I/

- -

--_._-Hot layer

EdglZ of
curtain

- -

---.-

Mixing layer
-I-

-- -

-- -

Cold lover

I
I

100

200

300

AIR TEMPERATURE ABOVE AMBIENT(9) -

FIG.7 TYPICAL VERTICAL TEMPERATURE

400
o

DISTRIBUTION

Hot
layer

c
~

a:
......
'-'

q~

CO<;~~~*~~~~~~~~~~4L~----JLEdge
of
cu rtoin

l.1J

beo

Mixing
loyer

I-

J:

llJ

::r

Cold
layer

05

0'1

FIG.8. TYPICAL VERTICAL DISTRIBUTION

OF

/'

-+-

-V
I-/
/. A+
*
.>
-/'

.~

,.~

/'

51+

3r:/'

2
/'

1.- /

i/

0'/

.r>

-it-

02
03 0'4 0'5 06 '07 08 09 .',0
C;:ALCULATED RATE OF DISCHARGE THROUGH VENT- ft 3/5
(UNCORRECTED FOR DISCHARGE COEFFICIENT)

Discharge coeticient of 0'6


CUR.TAIN DEPTH -in

<,

Discharge coeficient of 07

VENT AREA - sq in.


16
24
40

12

C""J

--+

-+

+ '+

FIG. 9 c;oE;:FFIC1~NT OF DISCHARGE THROUGH THE VENT

a:
~

(J')
-

...
III

O~

I.

'-'-1'

ll--

1
Wz
~-

~~
tr:

O~

0 5.----.,.--+--

wU

0::

:Jet

(J')LLJ

iJ50

~5

CALCULATED

RATE OF DISCHAR<;;E

(UNCORRECTED
BURN~P.

DOUBLE LINE
,

UNDE~

J
CUP.TAIN _ft / s

FOR' DISCHARGE COEFFICIENT)

CURTAIN

CIRCULAR

1-0 ---05

VENT AREA ~

DEPTH ""':'in
I

.
.

16

24

-4;-

sq in.

40

---

._- 1 - - - - -

FIG.IO COEFFICIENT OF PISCHAR~E UNDER


EDGE OF CURTAIN

Hot air discharged


thr:qugh vent (Mv )

---

~~ entrolned by
rising column of
hot gas (M)

Burner.

\ I

Effective
/'W'
point source

RIG.

II.

---1

TO ILLUSTRATE PROCESSES MAINTAINING


CONS~ANT DEPTH, OF HOT AIR UNDER
CEILING

Lines are of equation

r%

<D
fZ

"-"

(8

8b
2/. =
375
3
(HI_d)~3
h+500)

where HJ-H+!-S

dJ
~

24

fAf

.~

300l-------t------t---~~

I
26 "':r

>

odJ
<{

19
Z

:J

2001t---

--:---

- t ---'7" --7""0-"----:*":.......-7""'--

...,....-l

w
u
W

I
f-

a::

IOOI~-----+-------t----____j

a:

c!

W
Q.

O''------;;-----~:-----~

f-

et

CURTAIN DEPTH-in

16

24

--+

--

12

FIG,12, TEI\~PERATURES

-4-

~~

a:

40

NEAR THE CE!LING

-.r-,

c
0"

VENT AREA -sq in.

~.

<,

a:

oJ:

~O'6

Av -

O' O~% (H -d)

~
2

I'--

where H'-25 in,

I..L

o
I

x Experi mental points


2

b:w

-0

10

VENT AREA
ttl

a>

.t.n
('t')

40

30

20

(Av) -

in 2

FIG.13. DEPTH OF HOT AIR LAYER


(1\.1("\

r't1c.rI-JAor::c-

111\.1r'tl:'D

rIIDT!l.II\.I\

so

c r<.,
,

--a:
'0

---~
0

~
", ......
9

o
I

8 in. c urto: n

I'- -,

<,

........

......

........ ~

\J..

---

'0:

w
~

~dge of

x.-

.=
...... ......

'U
:I:
J:

-- ti:
8

...J

Iz

U.

~
c::

o
I

0-

20

10

30

40'

VENT AREA (A..,)- in

50

-...,...-- Bottom of hot Jover below edge of curtain.

06 Ayd} +9(d-Hc)~ =0'043(25-d)~"'


-:' - - -

Bottom of hot layer above edge of curtain

06

Av d~

=0'043(2S-d)51
0 Curtain d~pth 8 in.

x Curtain depth Iin.

FIG,14. DEPTH

OF LAYER OF HOT AIR (GENERAL CASE)

075

,......

-:t

9. 5

>

.>:

~
B

0
,

::J

025

w
o
a:
.~

'U
!J1

10

20

30

VENT AREA" (Av)-in

40

50

o Experimental points

- - Theoretical line (see section 4:S)

FIG .15. QISCHARGE' OF AIR THROUGH. VENTS


(NO SPILL UNDER CUR.TAIN ')

~
.~.

075

::>

c!
<{

' O'~-----+-----'''''''-:---+---e-

u,

I,

' . 10

15

CURTAIN DEPTH ( He) -, in


o Experimental point,
- - Theoretical line (se~ section 4- 8)

FIG.16. DISCHARGE OF AIR UNDER EDGE


OF
CURTAIN (NO VENT)
.,
.

,..,.., ...,.

/~

'

-;- 1"'"

-:-

"-

.d
/

- -

1/

I
~

-5

/
/

/
0

"

/
/
/.

'5

0,

->

'

V,x

)(
'

10

20

30
,

VENT AREA
CURTAI~

DEPTH ~in

40

'

i,,~

(Av ) -

EXPEP..IMENTAL
POINTS

THEORETicAL
LINES
(see section 4-9)
,

FIG.

17

HE:AT

----

VENTED

50

Vtr:t area

". ". ". .. l


~

.. _- ..,.

"~'--'--

..

--,..-~._---.

.1

-I

,
- t - - - - - + - - - - --I

t-----t-------

?CO
EFFECTIVE' AREA

9F

FIRE (Afr..- inl

Lines ore obtained from equation (14)

FIG.IB, E:FFECT OF FIRE ARq:A ON DePTH


',OF LAYER OF' HQT AIR. (NO SPILL
UNDER CURTAIN)

'250

..'

-1/3589

F.R 4f9

Vent

in

.-

-~

out

VELOCITY

PR.ESSURE (qbove otmospher!c)

Theoretical velocity
distribution in inlet

Theoretical pressure distri but ion


away from inlet and vent

F1G.I9. VELOCITY AND

PRESSURE DISTRIBUTION

IN MODEL

You might also like