E-mail address:olivier.vauquelin@univ-valenciennes.fr(O.M!e gret).
0379-7112/02/$-e front matter r2002Elvier Science Ltd.All rights rerved.
PII:S0379-7112(02)00014-0
with jet fans(Fig.1a),the problem is rather simple:the standard procedure consists in blowing the whole smoke towards a tunnel exit,while avoiding the occurrence of backlayering[1],i.e.of a stratified
layer developing clo to the ceiling in the opposite direction as against the one impod by ventilation.In the ca of tunnels featuring transver ventilation systems(Fig.1b),the mechanical exhaust ducts located along the tunnel should be ud in order to confine theflowing smoke and then extract the smoke over as short a longitudinal distance as possible.
In this paper,we focus on the issue of smoke extraction for a fairly simple configuration:there is no natural longitudinal ventilation within the tunnel and only two mechanical exhaust ducts are activated;one on each side of thefire. Experimental simulations are carried out on a reduced scale model allowing a correct reprentation of the duality between natural and forced convection effects. For different simulatedfire heat relea rates(HRR)and for different locations and shapes of the ducts,the quantity of smoke is extracted via the exhaust ducts enabling the asssment of the system efficiency.This study compares the efficiency of veral simple configurations with respect to the extracts volumicflow rate(VFR).
2.Experimental t-up
Fig.2shows a schematic view of the experimental t-up.Thefire smokes are simulated using a low-density gas(a mixture of air and helium)relead into a channel of rectangular cross-ction.The cha
nnel is10m long,250mm high and 500mm wide.The dimensions correspond approximately to a1:20th scale reduction for a standard two-way road tunnel.The air–helium mix is injected through a circular opening(diameter D)located in the channel,on the axis and at ground level.Q s will be identified as the air–helium mix VFR.Finally,two exhaust ducts are t up and activated at3m on both sides of the atÀ60and
+60m on full scale).The exhaust duct VFR will be denoted as Q t:
The air and helium flow rates for the buoyant source are controlled by two independent flowmeters.Velocities into the channel and at the exit of the ducts are measured using a hot wire probe.Finally,the two fans ud for the smoke extraction are controlled by an electronic variator and their volumetric flow rate is associated to a voltage value.For a given value of the voltage,fans are always working with a constant volumetric flow rate.Conquently,the buoyant mixing is extracted only by a mechanical method and the buoyant effects due to the chimney effect can be neglected.
We have to note that this experimental simulation ud a densimetric plume to reprent a thermal plume (thermal transfers by conduction and radiation are ignored).Limitations of such an approach are widely prented and discusd in [2].The simulation of the full-scale phenomena is focud her
e on the strict reprentation of the buoyancy effects as compared to the inertia effects.Froude and Richardson relationships are thus modelled.Therefore,for a scale reduction ratio a ;both the VFR and HRR are affected by a 5=2relationship,whereas the density differences will be maintained at their full-scale value,by imposing the suitable helium proportion within the injected mix.The main difficulty in such a simulation consists in a proper asssment of the smoke VFR and
temperature童谣手抄报
Table 1
Full-scale values and reduced scale model adjustments as a function of fire HRR Fire HRR Full-scale values Reduced scale model values
(MW)using mi-empirical model [3]
稚子弄冰诗意(a ¼1=20)
Source Smoke T of Source Mixing Air ratio Helium
diameter (m)flow rate (m 3
/s)smoke (1C)diameter (cm)flow rate (l/mn)(%)ratio (%)
1
0.98 5.8267 4.919354.945.14 1.6519.43698.365137.562.510 2.4645.444612.3152431.868.220
3.33
84.3
515
16.6
2827
27.6
72.4
O.Vauquelin,O.M !e gret /Fire Safety Journal 37(2002)525–533527
relevant to a fire of a given HRR.With that aim,we have ud a mi-empirical
model [3],which propos adjustments in good agreement with the International Recommendations [4].Starting from the adjustments and taking into account a 1:20th scale reduction ratio,both the full-scale values and the model values have been drafted in Table 1,for the four simulated HRR in the experiments (1,4,10and 20MW).A 4MW HRR is taken to correspond to a large car on fire and a 20MW HRR is generally associated with a bus or a lorry on fire [4].The HRR values have to be considered as order of magnitude only and they still remain debatable.As an example,during the HG
工程合作协议书
Vfire test performed during the EUREKA project [5],the evaluated HRR reached an instantaneous value >120MW.
3.Efficiency of the extraction system
By definition,the efficiency of an exhaust duct (or alternatively of a t of ducts)is calculated by the ratio of the extracted smoke VFR to the produced smoke VFR.It can be expresd as follows:
团校培训心得体会e ¼
Q es
Q s
:ð1Þ
As a conquence,the efficiency may theoretically reach 100%whenever the exhaust VFR equals the smoke VFR generated by the fire.In a fire situation,a mixture of smoke and gas combined with fresh air would be extracted.Therefore,100%efficiency will be achieved only by increasing the VFR of the smoke extract system.亘古绵长
美白吃什么
Efficiency asssment requires the measurement of mixture concentrations and mass flow rates,both at the source of the fire and at the outlet of the ducts.r s will be identified as the air–helium mix density associated to the VFR Q s ;r t will be identified as the density associated to the total exhaust VFR Q t :
In operational mode during a fire,Q t corresponds,in fact,to the sum of both VFR:on the one hand,a pure smoke VFR Q es (density of pure smoke is r s )and on the other,a fresh air VFR Q air (density of fresh air is r air ).As a result we obtain
Q t ¼Q es þQ air :
ð2Þ
The mass-flow rate conrvation at the duct level can be written as follows:
r t Q t ¼r s Q es þr air Q air :
ð3ÞBy combining (2)and (3),efficiency given by relation (1)can be also written as
e ¼
r air Àr t r air Àr s Q t
Q s
:
ð4ÞIf ðw air Þs and ðw air Þt reprent the air volume fractions,respectively,at the source and at the duct exit,r t and r s can also be expresd as
r t ¼½1Àðw air Þt r helium þðw air Þt r air ;
ð5a Þ
O.Vauquelin,O.M !e
gret /Fire Safety Journal 37(2002)525–533528
r s¼½1Àðw airÞs r heliumþðw airÞs r air:ð5bÞFinally,relations(4),(5a)and(5b)yield
e¼1Àðw airÞt
1Àðw airÞs
Q t
Q s
:ð6Þ
In practice,during the experiments,air volume fractions are measured via the oxygen percentages.The apparatus ud for the measurements works by sampling and analysing.Knowing that the proportion of oxygen in air is21.0%,relation(6) thus becomes
e¼21Àð%O2Þt
21Àð%O2Þs
Q t
Q s
:ð7Þ
延河
4.Results
4.1.Influence of the duct location
In thisfirst t of experiments,the simulated HRR is4MW(full scale).The system efficiency is determined depending on the extract VFR,for three different locations of ducts:
*Location L1:the ducts have a central location at the ceiling,
*Location L2:the ducts are located at the ceiling,but at one-third of the total width,
*Location L3:the ducts are located in the upper part of one of the walls.
In all locations,the extract ducts are equivalent to a1mÂ1m=1m2opening.In configuration L3,the top edge of the extract opening is located300mm from the ceiling.
The direct measurements give the oxygen percentage of the extracted gas mixture. The converted results(in terms of efficiency,using relation(7))are given in Table2 and plotted on the graph in Fig.3.
Table2
Values of the two-ducts system efficiency as a function of duct location
Q t2Q t System efficiency有关的英文
Duct VFR System VFR Location L1Location L2Location L3
Centered(%)1/3moved(%)top wall(%) 0.25Q s0.50Q s191914
0.50Q s 1.00Q s333424
1.00Q s
2.00Q s525239
2.00Q s 4.00Q s777861
3.00Q s 6.00Q s908977
O.Vauquelin,O.M!e gret/Fire Safety Journal37(2002)525–533529
