Idea Transcript
Aerosol Can Explosion Test (caution: graphic material)
Problem: want to eliminate Halon 1301 from use in aircraft cargo bays 1. Halon 1301 (CF3Br) => high ODP, high GWP.
Compound Halon 1301 (CF3Br)
HFC-125 H
(CF3CF2H)
2-BTP (CH2CBrCF3)
FK-5-1-12 (CF3CF2C(O)CF(CF3)2)
Atmospheric Lifetime (yrs)
ODP
GWP100
65
12
6,900
29
0
3,400
0.008
0
N/A
0.014
0
1
Problem: want to eliminate Halon 1301 from use in aircraft cargo bays
2. But in one FAA-mandated qualification test, the possible replacements make things worse.
Gas Analyzer Probe Pressure Transducer
10 m3 Pressure Vessel
Fan Igniter
Propane Ethanol Water
Agent Port
Nitrogen Port
Thermocouple
Aerosol Can Simulator View) FAA Aerosol(TopCan Simulator
Video Camera 1
Goals Understand the overpressure phenomena in the FAA Aerosol Can Test 1. Why is the overpressure occurring with the added suppressants? 2. What can be done about it?
2-BTP
FK 5-1-12 HFC-125
FAA Aerosol Can Simulator No agent
1301
Understanding Combustion Promotion by Halogenated Fire Suppressants INTERNATIONAL AIRCRAFT SYSTEMS FIRE PROTECTION WORKING GROUP MEETING May 11-12, 2011 Greg Linteris, NIST Fire Research Jeff Manion / Wing Tsang / Don Burgess, NIST Chemical Kinetics Harsha Chelliah, Univ. of Virginia Vish Katta, Innovative Scientific Solutions, Inc. Fumi Takahashi, Case Western Reserve Univ. Oliver Meier, The Boeing Company
Innovative Scientific Solutions Incorporated
National Institute of Standards and Technology Building and Fire Research Laboratory
Acknowledgements
Wing Tsang, Don Burgess,
NIST
John Reinhardt, Dave Blake FAA Technical Center Med Colket, Ken Smith,
UTRC
The work was supported by The Boeing Company.
Approach Physics in FAA test is too complicated to examine with detailed kinetics, so simplify. Fuel discharge port (propane/ethanol/water) Arcing ignitor
Droplet evaporation, turbulent pre-mixing Ignition induction period (PFR)
Partially premixed fuelrich reactants (PREMIX), or distributed reaction region (PSR).
Air and agent mixture
High strain (shear), partially premixed diffusion flame region (OFDF).
Partially premixed diffusion flame with ancillary burning of agent (UNICORN)
Progress
Reviewed previous work Thermodynamic Equilibrium Calculations Kinetic Mechanism Development Measurement of 2-BTP Decomposition Perfectly-Stirred Reactor (PSR) Calculations Diffusion Flame Calculations (Cup Burner) Homogeneous Auto-Ignition (PFR) Calculations Diffusion Flame Calculations (Counterflow) Premixed Flame Calculations (PREMIX)
Background:
Previous findings
~ Of the 65 relevant papers collected and assimilated, these are highlights (in which enhanced combustion has been discussed): Researchers
Fuel
Agents
Experiment
Phenomena
Explanation
Grosshandler and Gmurczyk
Propane, ethylene
CF3I, CF3Br, HFCs
Detonation Deflagratoin Tube
Higher Ma, flame speed, pressure ratio
None
Shebeko et al.
methane, hydrogen C2HF5, C4F10
Deflagration
Higer pressure rise and dP/dt
Moriwaki et al.
methane, ethane
CH3Cl, CH3I, CH3, Br
Shock tube
Shorter ignition delay
Added heat release from agent None
Ikeda and Mackie
ethane
C3HF7
Shock tube
Shorter ignition delay
None
Mawhinney et al.
heptane
water mist
Heptane pool fi Higher heat release
Enhanced fluid-dynamic mixing
Hamins et al
hydrocarbons
HFCs, water mist, N2, powders
Full-scale testsHigher pressure, visual flames
Enhanced fluid-dynamic mixing
Holmstedt et al.
propane
C3HF7, C2H2F4, CF3Br,
Diffusion flame Higher heat release
None
Katta et al.
methane
CF3H
Cup burner
Agent reaction
Ural
none
C3HF7, C2H2F4, CHClF2
Flammability Visual observation tube/chamber
Higher heat release
Heat loss/ gain
Background:
Flame Extinction Chemical Flow Time (s) Time (s)
Flames go out when:
10-1
10-1
10-2
10-2
10-3
10-3
10-4
10-4
10-5
10-5
τchem > τflow
Background:
Flame Extinction
The flow-field influences the extinction process:
D ≡ τr /τc Chemical Time:
τ c ≡ ρ / w = ρ c c A exp( E / RT ). −n −m F O
Flow Time:
τr = l / v
-1
Background:
Flame Extinction
A measure of the overall chemical reaction rate can be obtained with: Perfectly-Stirred Reactor (PSR) Calculations Diffusion Flame Calculations (Counterflow) Premixed Flame Calculations
=> Concentrate on R-125 => Why is it surprising that R-125 did not put out the ACT at 11.3 %
Background:
Flame Extinction, Propane-Air with R-125, Counterflow
R-125
~ jet velocity
fuel
flame
oxidizer
Cup burner “Suppression of Nonpremixed Flames by Fluorinated Ethanes and Propanes,” E. J. P. Zegers, et al. CNF 121:471-487 (2000)
Flame Extinction, Heptane-Air with R-125, Counterflow
Agent Volume Fraction
Background:
R-125
Extinction Strain Rate /s
~ jet velocity
Trees et al. NIST SP-861
Background:
Flame Extinction
To understand why R-125 does not extinguish the FAA ACT, we must understand : - the fuel reaction chemistry - agent reaction chemistry - mixing - flame characteristics.
Thermodynamic Equilibrium Calculations
What do equilibrium calculations tell us about the general behavior of the system? To do an equilibrium calculation, one must know the initial reactant mix (fuel, air, agent, water vapor, etc.). We don’t really know them for the ACT, so keep them all as variables, and find the equilibrium conditions for a wide range of initial mixtures.
HFC-125: Adiabatic Flame Temperature (Taft)
Taft / K
Taft,max / K
Xi 0
2000
Thermodynamic Equilibrium Calculations
1.2
2000
1.0
0.6 1500
1500
0.144
0.4 0.2 1000 0.0
a.)
0.2
0.4
0.6
0.8
1.0
Fraction of Chamber Volume Involved in Combustion, η
1000
b.)
0.0 0
5
10
Inhibitor Fraction in Oxidizer, Inhibitor Volume Volume Fraction in Oxidizer, Xi (%)Xi (%)
15
- Taft is high for all η. - Change in behavior at [X]/[H]=1 (about 7.5 % HFC-125, red curve above). - With large amounts of agent, a wide range of η gives nearly equivalent Taft. - As agent is added, more and more chamber volume is necessary to achieve stoichiometric combustion. - Where flame goes out (Xi=13.5 %), all the chamber volume is involved in combustion (i.e., η=1).
η (Taft,max )
0.8 0.072
2-BTP: Adiabatic Flame Temperature (Taft)
Thermodynamic Equilibrium Calculations
BTP simulant
Taft / K
Taft / K
1.2
Taft,max / K
2000 2000
2000
1.0
0 0 1500 1500
0.6 1500
0.025 0.03 0.040 0.044 0.06
0.4 0.2
1000 1000 0.0 0.0
c.)
0.2 0.2
0.4 0.4
0.6 0.6
0.8 0.8
ηη in Combustion, η Fraction of Vessel Air Involved
1.0 1.0
1000
d.)
0.0 0
1
2
3
4
5
6
X i (%) Inhibitor Volume Fraction in Oxidizer, Xi (%)
- Taft is high for all η. - most of the plot is below [X]/[H]=1 (about 6 % 2-BTP), so can’t see change at [X]/[H]=1 . - With large amounts of agent, a wide range of η gives equivalent Taft. - As agent is added, more and more chamber volume is necessary to achieve stoichiometric combustion. - Where flame goes out (Xi=6 %), all the chamber volume is involved in combustion (i.e., η=1).
η (Taft,max )
0.8
Xi Xi
Taft / K
Xi
Thermodynamic Equilibrium Calculations
Taft,max / K
0 2000
0.5
2000
0.118
0.4
0.3 1500
1500
0.2
0.1
1000 0.0
e.)
0.2
0.4
η
0.6
0.8
Fraction of Vessel Air Involved in Combustion, η
1.0
1000
f.)
0.0 0
1
2
3
4
X I (%) Inhibitor Volume Fraction in Oxidizer, Xi (%)
- Taft is high for all η., but decreases somewhat as agent is added. - most of the plot is below [X]/[H]=1 (about 11 % CF3Br), so can’t see change at [X]/[H]=1 . - The amount of chamber volume for peak Taft does not change with Xi. -Why? =>
CF3Br + 2H2O = 3HF + HBr + CO2 ,
-i.e., there’s always enough H and O in the system to oxidize the CF3Br without more air! - The Taft is very sensitive to η.
5
6
η ( Taft,max )
Halon 1301: Adiabatic Flame Temperature (Taft)
Thermodynamic Equilibrium Calculations
What do they tell us about the maximum pressure rise?
Thermodynamic Equilibrium Calculations
HFC-125: Predicted Pressure Rise
Pressure Rise / bar 6
Pressure Rise at η {ypeak} / bar 0.144 0.072
5
6
Tmax CO2
5
4
4
3
0
2
Xi
3
2 FAA Experiment
1
1
0
0
a.)
0.0
0.2 0.4 0.6 0.8 η Fraction of Vessel Air Involved in Combustion, η
1.0
0
5
10
15
Inhibitor Volume Fraction in Oxidizer, X I (%)
- The higher η, the greater ΔP (more reactants, more heat release, more expansion of hot products—since the oxidizer also includes a “fuel” species). - The actual fraction of chamber volume (oxidizer) which can react has a large influence on ΔP. - Equilibrium thermodynamics predicts the final pressure quite well. - Why does the agent not reduce the extent of reaction?
Thermodynamic Equilibrium Calculations
2-BTP: Predicted Pressure Rise
Pressure Rise / bar
0.06
7
Pressure Rise at η {ypeak} / bar 7
6
0.03
5
4
4
0
2
Xi
3 2 FAA Experiment
1
1
0
0
c.)
0.0
0.2
0.4
η
0.6
0.8
1.0
Fraction of Vessel Air Involved in Combustion, η
CO2
6
5
3
Tmax
0
1
2
3
X i (%)
4
Inhibitor Volume Fraction in Oxidizer, Xi (%)
- Same basic behavior as R-125, but greater ΔP. - The actual fraction of chamber volume (oxidizer) which can react has a large influence on ΔP. - Equilibrium thermodynamics predicts the final pressure quite well. - Why does the agent not reduce the extent of reaction?
5
6
Thermodynamic Equilibrium Calculations
Halon 1301: Predicted Pressure Rise
Pressure Rise / bar
Pressure Rise at η {ypeak} / bar
7
7
6
6
5
5
4
4
3
3
2
2
1
1
0
0
CO2 Tmax FAA Experiment
0.0
e.)
0.2 0.4 0.8 η 0.6 Fraction of Vessel Air Involved in Combustion, η
1.0
f.) 0
1
2
3
X i (%)
4
5
6
Inhibitor Volume Fraction in Oxidizer, Xi (%)
- Higher η has very little effect on ΔP. - At η of peak Taft, or CO2, the ΔP is constant! => can’t use pressure rise to determine η. - Actual ΔP is always less than predicted. This is due to a chemical kinetic effect, but is it from Br or from reduced temperature (i.e., from mixing-induced dilution)? => MUST LOOK AT THE KINETICS TO FIND OUT!
Halon 1301: Predicted Pressure Rise
Thermodynamic Equilibrium Calculations
Pressure Rise at η {y {Tpeak } / }bar / bar aft,peak 7
2-BTP 6
R-125
5 4 3 2
1301 1 0 0
5
X i (%)
10
15
- As Xi of agent goes up, ΔP can increase for R-125 and 2-BTP, but not for 1301. => MUST LOOK AT THE KINETICS TO FIND OUT WHY!
Kinetic Mechanism Development CH4-air premixed flame, 0, 4, and 6 % R-125 Currently developing these charts for HFC125 with propane and ACT.
Kinetic Mechanism Development
Sub-Mechanisms
Aersol Can Test Mechanism: Species Reactions C4 hydrocarbon mechanism from Wang
111
784
Ethanol mechanism of Dryer
5
36
HFC mechanism from NIST1,2
51
600
CF3Br mechanism of Babushok (NIST)2
10
122
-------177
-------1494
1 Updated rates from more recent literature, additional rates of fuel radical reaction with R-125. 2 Validation: CH -air and CH OH systems (with CHF , C H F , C HF , CF Br, C HF ): 4 3 3 2 2 4 2 5 3 3 7
- premixed flame speed, - species profiles in low-pressure premixed flames, - extinction strain rate for counterflow diffusion flames, - cup-burner extinction.
Kinetic Mechanism Development :
Measurements of 2-BTP Decomposition
- Can’t do calculations yet for 2-BTP because there’s no mechanism for its initial decomposition. - Once we have its decomposition to HFC and HBrC fragments, it will feed into the overall NIST HFC mechanism. - So, we must first estimate/measure/calculate its decomposition => CSTL. -But! For now, use a 2-BTP simulant: 1 mole CF3Br, 1 mole C2H2, 3 moles N2 (gives the right Taft, and has the right number of molecules).
Perfectly-Stirred Reactor (PSR) Calculations - Used to estimate the overall chemical reaction rate. - Performed for R-125, 1301, and 1301 with N2.
m& Yi ,in hi ,in Yi T P V
m& Yi ,out
Assumptions:
hi ,out
- specified premixed inlet conditions. - adiabatic (no heat losses), no species reaction at the walls. - perfectly stirred (outlet conditions are the same as the reactor conditions). - steady-state operation.
Perfectly-Stirred Reactor (PSR) Calculations
Calculation method
1. We want a measure of τchem 2. At the blow-out condition, τchem=τflow 3. To find the blow-out condition, calculate Tpsr at decreasing values of the residence time, τflow, until the time is too short for reaction to occur (Tpsr drops to inlet temperature (blow-out). From Colket and co-workers, 2010
Perfectly-Stirred Reactor (PSR) ω psr / s-1 10000
Overall Chemical Rate with R-125
HFC-125 Xi 0
1000
100
0.135 10
1 0.0
0.2
0.4
0.6
0.8
1.0
Fraction of Vessel Air Involved in Combustion, η
- Adding R-125 lowers ωchem for rich mixtures (low η), but raises (then lowers) it for lean mixtures (high η). −η has a big effect on overall chemical rate at low Xi, less effect at high Xi (follows temperature results). - i.e., for higher Xi, these curves flatten ( ωchem is insensitive to η for η > 0.4 ).
Perfectly-Stirred Reactor (PSR)
Overall Chemical Rate with 1301
ω psr / s-1
1301
10000
1000 Xi 0
100
0.045
10
1 0.0
0.2
0.4
0.6
0.8
Fraction of Vessel Air Involved in Combustion, η
- Adding 1301 always lowers ωchem (for all η) - ωchem falls off very steeply with η (for all Xinh; follows temperature results).
1.0
Perfectly-Stirred Reactor (PSR) Overall Chemical Rate with 2-BTP simulant BTP simulant
-1 ω psr / s
10000
1000
Xi (%) 0 100
2.5 10
4 4.4
1 0.0
0.2
0.4
0.6
0.8
1.0
Fraction of Chamber Volume Involved in Combustion, η
- Adding 2-BTP simulant lowers or raises ωchem (depending upon η) - variation of ωchem with η is mild at high Xinh, but strong without agent.
Perfectly-Stirred Reactor (PSR) Calculations
R-125 vs. 1301
-1 -1 Overall Chemical ωpsr / s Rate / s 10000
Inhibitor Volume Fraction in Air % 1000
XR-125 = 7.5 %
100
11.2 %
13.5 % 10
Fraction of Vessel Air Involved in Combustion
1 0.0
0.2
0.4
0.6
- Top two curves do not put the flame out; bottom one does.
0.8
1.0
Perfectly-Stirred Reactor (PSR) Calculations
R-125 vs. 1301
-1 -1 Overall Chemical ωpsr / s Rate / s 10000
1000
XR-125 = 7.5 %
100
X1301 = 2.2 %
13.5 %
X1301 = 3.9 %
10
Fraction of Vessel Air Involved in Combustion
1 0.0
0.2
0.4
0.6
0.8
1.0
- For R-125, we can use pressure rise data with equilibrium calculations to estimate η. - For 1301, can’t use pressure rise, so we don’t really know η.
Perfectly-Stirred Reactor (PSR) Calculations
R-125, 2-BTPsim, and 1301
ω psr / s-1 10000 Xinh = 0 %
1000 X1301 = 2.2 % XR-125 = 7.5 % 100
XR-125 = 13.5 % 10
X1301 = 3.9 %
1 0.0
0.2 0.4 0.6 0.8 Fraction of Vessel Air Involved in Combustion, η
1.0
- For R-125, we can use pressure rise data with equilibrium calculations to estimate η (solid dots). - For 1301, can’t use pressure rise, so we don’t really know η. =>BUT for 1301 ωchem is very sensitive to η.
Perfectly-Stirred Reactor (PSR) Calculations
R-125, 2-BTPsim, and 1301
ω psr / s-1 10000 Xi = 0 %
1000
XBTPsim = 2.5 %
XBTPsim = 4.8 % X1301 = 2.2 % XR-125 = 7.5 %
100
XR-125 = 13.5 % 10
X1301 = 3.9 %
1 0.0
0.2 0.4 0.6 0.8 Fraction of Vessel Air Involved in Combustion
1.0
- For R-125, or 2-BTP simulant, we can use pressure rise data with equilibrium calculations to estimate η(solid dots). - For 1301, can’t use pressure rise, so we don’t really know η. =>BUT for 1301 ωchem is very sensitive to η.
Perfectly-Stirred Reactor (PSR) Calculations 10000
0 0 00
PSR Overall Chemical Rate / s
-1
1
1000
1
100
1
0 0 0
1301, η = 0.33
HFC-125, η= 0.33 HFC-125, η(Taft,max) 1301, η = 0.47
1
0
50
10
1.
More agent generally reduces wpsr , for all assumed values of η.
2.
For the case η=0.47, there is little change in wpsr for the curve for HFC125 up to 30 %.
3.
For HFC-125 (blue curves), the reduction in wpsr with addition of agent is similar regardless of the value of h; i.e., for η=0.33, η=0.47, or η(Taft|peak).
4.
The effectiveness of the agent CF3Br is very sensitive to the value of η.
5.
For CF3Br to be more effective than HFC-125, η must be greater than about 0.4.
0 0
HFC-125, η= 0.47
10
R-125 vs. 1301
0
100
Inhibitor Volume Fraction / Inhibitor Volume Fraction at Suppression (% )
Perfectly-Stirred Reactor (PSR) Calculations 10000
1
0 00 0
PSR Chemical Rate / s
-1
BTPsim 1000
1
100
1
R-125 vs. 1301
0 00
1.
More agent generally reduces wpsr , for all assumed values of η.
2.
For the case η=0.47, there is little change in wpsr for the curve for HFC125 up to 30 %, and for the curve for 2BTP simulant up to 50 %.
3.
For HFC-125 (blue curves), the reduction in wpsr with addition of agent is similar regardless of the value of h; i.e., for η=0.33, η=0.47, or η(Taft|peak).
4.
The 2-BTP simulant (green curves) is much less effective than HFC-125.
5.
The effectiveness of the agents CF3Br and 2-BTP simulant (which differ primarily in the addition of C2H2 to CF3Br in the 2-BTP simulant) are both very sensitive to the value of h, but the influence is in the opposite direction: increasing h reduces the effectiveness of 2-BTP simulant, but increases the effectiveness of CF3Br.
6.
For CF3Br to be more effective than HFC-125, η must be greater than about 0.4.
η(Taft,max)
1301, η = 0.33 BTPsim
η=0.47
BTPsim
η=0.33
0 0
HFC-125, η= 0.47 HFC-125, η= 0.33 HFC-125, η(Taft,max) 1301, η = 0.47 10
1
0
50
10
0
100
Inhibitor Volume Fraction / Inhibitor Volume Fraction at Suppression
Perfectly-Stirred Reactor (PSR) Calculations
Current Understanding
Equilibrium and PSR Calculations Indicate: 1. In the FAA ACT with R-125 or 2-BTP, to achieve the observed pressure rise, a large fraction of the chamber volume (with the agent) must be involved in the combustion. 2. Thus, the agents are not inert, but rather, act like poorly-burning fuels. 3. Unlike in other flames, very little kinetic inhibition is occurring with R-125 and 2BTP; whereas, CF3Br does inhibit the flame, as expected. 4. The amount of chamber volume involved in the combustion, η, appears to be a key parameter controlling the kinetic behavior (i.e., the kinetic inhibition by CF3Br is very sensitive to η, but R-125 is not). 5. Simulations with 2-BTPsimulant imply that the fuel portion of the 2-BTP molecule causes it to be much less effective than CF3Br.
Equilibrium Calculations Indicate: => With R-125 or 2-BTP: - the peak Taft does not drop much with added agent. - more agent requires higher η. - at high Xi, Taft is nearly the same regardless of the equivalence ratio.
=> With CF3Br, - Taft drops off away from stoichiometric. - η is not changed with added agent. R-125
2-BTP
CF3Br
Taft / K
Taft / K
Taft / K
Xi 0
Xi 0
2000
2000
2000
0.118
Xi
0.072
0 1500
1500
0.144
1500
0.03 0.06
1000
a.)
1000
1000 0.0
0.2
0.4
0.6
0.8
1.0
Fraction of Chamber Volume Involved in Combustion, η
0.0
c.)
0.2
0.4
η
0.6
0.8
1.0
0.0
e.)
0.2
0.4
η
0.6
0.8
1.0
Equilibrium Calculations Indicate (pressure rise): => With R-125 or 2-BTP, complete reaction of the agent at the equivalence ratio giving peak Taft can predict the pressure rise (except at the extinction point).
=> With CF3Br, adding agent will never increase the pressure rise, even without kinetic
inhibition.
Pressure Rise at η {y {Tpeak } / }bar / bar aft,peak 7
2-BTP 6
R-125
5 4 3 2
1301 1 0 0
5
X i (%)
10
15
Perfectly-Stirred Reactor (PSR) Calculations Indicate: => With R-125 or 2-BTP simulant, at high Xi, the overall reaction rate is relatively unchanged as equivalence ratio changes.
=> With CF3Br, the overall reaction rate is very sensitive to the equivalence ratio. ω psr / s-1 10000 Xi = 0 %
1000
XBTPsim = 2.5 %
XBTPsim = 4.8 % X1301 = 2.2 % XR-125 = 7.5 %
100
XR-125 = 13.5 % 10
X1301 = 3.9 %
1 0.0
0.2 0.4 0.6 0.8 Fraction of Vessel Air Involved in Combustion
1.0
Key Questions Still to Answer
0. Are the results for 2-BTP itself the same as for 2-BTP simulant? •
Is the amount of involved oxidizer the key feature?
•
Does the agent reaction rate affect the strain conditions in the FAA ACT?
•
Why are the kinetics with R-125 not slower (i.e., slow enough for extinguishment)?
•
Does Br help slow the kinetics with 2-BTP?
•
Is the overpressure due to a pressure enhancement of the agent flammability?
•
Is the inerting concentration required for suppression?
•
Is there any way around the undesired results?
Other Ongoing Work
1. Cup burner flame simulations. 2. Premixed flame simulations. 3. Counterflow diffusion flame simulations. 4. Homogeneous ignition simulations.
Future Plans 1. - Perform further analysis of simulations in progress to understand reasons for lack of kinetic inhibition with R-125. - Perform 2-D, axi-symmetric, unsteady simulations for a turbulent fuel jet to understand the effects of mixing on the extinction. - Repeat existing calculations at higher pressure. 2. Perform large-scale tests in cooperation with the FAA Technical Center to test our understanding. 3. 2-BTP: - measure and estimate decomposition rate - develop kinetic mechanism - perform calculations - analyze results to understand lack of kinetic effect with 2-BTP 4. Develop a new laboratory-scale experiment to: - validate our understanding (e.g., η, pressure effects), and the mechanisms. - explore range of conditions for which inhibition/enhancement occurs - rapidly screen new agents.
New Constant-Volume Combustion Device*
*Photo Courtesy of Prof. Li, Purdue
Questions ?
Extra Material
Perfectly-Stirred Reactor (PSR) Calculations
1301 with N2
5 31
4.5
CF3Br Mole Fraction in Air / %
4
non-Flammable
3.5
y = 0.0615x2 - 1.6472x + 11.057 0
0
18
0
3
0
2.5
0 1
0
0
2
0 0 0
0
12 0
0
0.5
0 8
0
0
10
12
18 0 0 12 18 12 3 00 0 18 18 12100188 1 0 25 18 0
18
8 25
0
Flammable
3
00
14
18
25
0
8 0 01 12 0
0 0
0
18
0 10 0 1 0 0 3 0 8
1.5 1
1
0
0
0
25
16
O2 Mole Fraction in Air / %
18
20
Kinetic Mechanism Development : Measurements of 2-BTP Decomposition
Kinetic Mechanism Development : Measurements of 2-BTP Decomposition
Single Pulse Shock Tube Driver Section Diaphragm
Sample Port
Sample Tanks
Dump Tank
Driver Gas
Driven Gas (Sample) Valve & Loop Sampling System
GC Column C1 - C4
GC Column > C4
Characteristics: System heated to 100 °C
τ = (500 ± 50) μs (monitored with
pressure transducers) Typical shock conditions: 2‐6 bar, 900 – 1250 K
Splitter Data
FID 1 FID 2
MSD
Advantages of Shock Tube for Gas Kinetic Studies: Essentially a pulse heater, τ = (500 ± 50) μs
No surface induced chemistry (diffusion slow compared with τ)
Use of dilute conditions, radical chain inhibitors, sensitive GC/MS analysis isolation of initial processes, observation of multiple channels
Studies of 2-BTP Decomposition Unimolecular (initial studies):
Initial kinetic studies show some interference from radical induced decomposition (work in progress) Slow rate suggests importance of radical processes in practical systems
log k(HBr elimination) s-1
HBr elimination from 2‐BTP ca. 100x slower than unfluorinated analog
Kinetics of HBr Elimination 3.5 3
2‐bromopropene
2.5 2 1.5 1 0.5 0 -0.5
2‐BTP (333‐trifluoro‐2‐bromopropene)
-1 -1.5 8
8.5
9 10000/T K
Bimolecular decomposition induced by reactive radicals (e.g. H atoms): H + 2‐BTP Products Initial studies show products indicating displacement and abstraction of Br as major channels. But ‐ product spectrum more complex than expected with some as yet unidentified species. Work in progress to determine mechanism and kinetics.
9.5