1. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 1 Fire Dynamics II Lecture # 8 Flame Spread & Burning Rates Jim Mehaffey 82.583

2. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 2 Flame Spread & Burning Rates Outline Models for flame spread on solids (review)Models for flame spread on solids (review) –wind-aided vs opposed-flow flame spread –in the absence or presence of external radiation Burning rates of common itemsBurning rates of common items –in the open (review) –limited by ventilation –enhanced by radiation

3. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 3 Factors Affecting Rate of Spread of Flame Material Factors Chemical Composition of fuel Presence of fire retardants Physical Initial temperature Surface orientation Direction of propagation Thickness Thermal conductivity Specific heat Density Geometry Continuity

4. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 4 Factors Affecting Rate of Spread of Flame Environmental Factors Composition of atmosphere Temperature Imposed heat flux Air velocity

5. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 5 Spread of Flame over Wall Linings

6. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 6 Spread of Flame over Wall Linings

7. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 7 Room Fire Test - Apparatus ISO 9705 “Fire tests: Full scale room fire tests for surface products”ISO 9705 “Fire tests: Full scale room fire tests for surface products”

8. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 8 Room Fire Test - Procedure Line walls and ceiling with productLine walls and ceiling with product Burner in back cornerBurner in back corner –First 10 min: = 100 kW (large wastepaper basket) –Last 10 min: = 300 kW (small upholstered chair) Observe time to flashoverObserve time to flashover Room experiences flashover when  1,000 kWRoom experiences flashover when  1,000 kW

9. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 9 Room Fire Test - Results

10. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 10 Flame Spread Models: Concepts Flame spread = an advancing ignition frontFlame spread = an advancing ignition front Leading edge of flame is heat source (raising fuel to ignition temp) and the pilotLeading edge of flame is heat source (raising fuel to ignition temp) and the pilot Visually flame spread is advancing flame close to solidVisually flame spread is advancing flame close to solid Two interacting advancing frontsTwo interacting advancing fronts –flame front in gas phase –pyrolysis front along solid surface Heat transfer from flame  pyrolysis front to advanceHeat transfer from flame  pyrolysis front to advance Advance of pyrolysis front  increased release of volatiles  advance of flame frontAdvance of pyrolysis front  increased release of volatiles  advance of flame front Flame-spread velocity  rate of advance of pyrolysis frontFlame-spread velocity  rate of advance of pyrolysis front

11. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 11 Wind-aided Spread

12. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 12 Wind-aided Spread  = region of heat transfer from flame & smoke  = region of heat transfer from flame & smoke For wind-aided spread: 0.1 m    10 mFor wind-aided spread: 0.1 m    10 m For opposed-flow spread: 1 mm    3 mmFor opposed-flow spread: 1 mm    3 mm Surface temp in control volume drops from T ig to T sSurface temp in control volume drops from T ig to T s Pyrolysis front moves at speedPyrolysis front moves at speed Model for wind-aided flame spread:Model for wind-aided flame spread:

13. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 13 Example of Accelerating Flame Spread Upward turbulent spread on thick PMMAUpward turbulent spread on thick PMMA x b = 0 and n = 0.94 ~ 1x b = 0 and n = 0.94 ~ 1 Eqn (8-1) Experiment finding:Experiment finding: When x p ~ 1.0 m, v ~ 5.0 mm s -1

14. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 14 Example of Constant Flame Spread Upward turbulent spread on thin textilesUpward turbulent spread on thin textiles n ~ 0.6n ~ 0.6 Eqn (8-2) After some time, (x p - x b ) and v become constantAfter some time, (x p - x b ) and v become constant

15. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 15 Apartment Fire: Hiroshima, Japan (1996) Building- reinforced concrete structureBuilding- reinforced concrete structure - 20 storeys - height of each storey = 3 m - each apartment had a balcony Balcony- PMMA glazingBalcony- PMMA glazing - height of glazing = 1 m

16. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 16 Chronology of Fire 00:00 Fire commences within apartment 96500:00 Fire commences within apartment 965 13:00 Outer surface of PMMA glazing (9th storey) ignites13:00 Outer surface of PMMA glazing (9th storey) ignites 18:00 Outer surface of PMMA glazing (10th storey) ignites18:00 Outer surface of PMMA glazing (10th storey) ignites 20:00 Outer surface of PMMA glazing (11th storey) ignites20:00 Outer surface of PMMA glazing (11th storey) ignites 22:00 Outer surface of PMMA glazing (12th storey) ignites22:00 Outer surface of PMMA glazing (12th storey) ignites 23:00 Outer surface of PMMA glazing (13th storey) ignites23:00 Outer surface of PMMA glazing (13th storey) ignites 23:30 Outer surface of PMMA glazing (14th storey) ignites23:30 Outer surface of PMMA glazing (14th storey) ignites 24:00 Outer surface of PMMA glazing (15th storey) ignites24:00 Outer surface of PMMA glazing (15th storey) ignites 24:20 Outer surface of PMMA glazing (16th storey) ignites24:20 Outer surface of PMMA glazing (16th storey) ignites 24:40 Outer surface of PMMA glazing (17th storey) ignites24:40 Outer surface of PMMA glazing (17th storey) ignites 25:00 Outer surface of PMMA glazing (18th storey) ignites25:00 Outer surface of PMMA glazing (18th storey) ignites 25:15 Outer surface of PMMA glazing (19th storey) ignites25:15 Outer surface of PMMA glazing (19th storey) ignites 25:30 Outer surface of PMMA glazing (20th storey) ignites25:30 Outer surface of PMMA glazing (20th storey) ignites

17. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 17 - - - - inner surface burning ——— outer surface burning  Ignition of outer side of PMMA O Burn-out of PMMA

18. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 18 Problem Set 3: Problem 4 5. 5. In 1975, FMRC studied upward turbulent flame spread on thick PMMA and found the process obeyed the model for wind-aided flame spread presented in class with x b = 0 and n ~ 1. They found that when the flame extension x p = 1 m, the upward flame spread velocity V = 5 mm/s. Calculate the flame extension 2, 4, 6, 8, 10 and 12 minutes later. Compare your predictions with the observed flame extensions in the Hiroshima fire by plotting your predictions on the graph on page 8-17.

19. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 19 Opposed Flow Flame Spread Absence of External Radiation Compared to PMMA, a very slow processCompared to PMMA, a very slow process Not accelerating, but roughly constant velocityNot accelerating, but roughly constant velocity Speed of downward flame spread on PMMASpeed of downward flame spread on PMMA v ~ 0.04 mm s -1 v ~ 2.4 mm min -1

20. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 20 Opposed Flow Flame Spread In Presence of External Radiation (1) Effect of preheating time on rate of downward flame spread on PMMA exposed to radiant flux (kW m -2 )Effect of preheating time on rate of downward flame spread on PMMA exposed to radiant flux (kW m -2 ) CHF (PMMA) ~ 11 kW m -2CHF (PMMA) ~ 11 kW m -2

21. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 21 Opposed Flow Spread: Model for Thick Materials Quintiere and Harkleroad, 1985Quintiere and Harkleroad, 1985 Eqn (8-3)  = flame-heating parameter (kW 2 m -3 ) {material property}  = flame-heating parameter (kW 2 m -3 ) {material property} Provided no dripping, this model holds forProvided no dripping, this model holds for –downward flame spread (wall) –lateral flame spread (wall) –horizontal flame spread (floor) , k  c and T ig - measured (LIFT apparatus) , k  c and T ig - measured (LIFT apparatus) T s - depends on scenario (external flux)T s - depends on scenario (external flux)

22. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 22 LIFT Apparatus - Standard Tests ASTM E1321, “Standard test method for determining material ignition and flame spread propertiesASTM E1321, “Standard test method for determining material ignition and flame spread properties ISO 5668, “Fire tests: Reaction to fire: surface spread of flame on building products”ISO 5668, “Fire tests: Reaction to fire: surface spread of flame on building products”

23. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 23 LIFT Apparatus

24. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 24 LIFT Apparatus - Results

25. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 25 Estimating the Surface Temperature T S To employ Eqn (8-3) one must estimate T STo employ Eqn (8-3) one must estimate T S Assume the surface is heated by a radiant flux and cools by convection h (T S -T o )Assume the surface is heated by a radiant flux and cools by convection h (T S -T o ) Following pages 5-35 to 5-38 in Fire Dynamics IFollowing pages 5-35 to 5-38 in Fire Dynamics I Eqn (8-4)

26. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 26

27. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 27

28. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 28 Thermal Properties for Ignition, Flame Spread & Pre-flashover Fires (1)

29. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 29 Problem Set 3: Problem 3 3.Consider a pre-flashover fire in a room 2.4 m x 3.6 m x 2.4 m (height). The door to the room (0.8 m x 2.0 m (height)) is open and the interface between the hot layer and cool air is at the mid-height of the door. The fuel in the room is a mixture of wood and plastics and the mean extinction (absorption) coefficient of the upper layer is K m ~ 1.0 m -1. What is the emissivity of the upper layer? Calculate the radiant flux at the centre of the floor when the layer temperature is 300°C, 400°C, 500°C and 600°C.

30. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 30 Problem Set 3: Problem 4 4. 4.Calculate the time to piloted ignition of a wood floor and a polyurethane cushion at floor level for the four upper layer temperatures considered in Problem 3. Use Tewarson’s model assuming for the wooden floor that CHF = 10 kW m -2 and TRP = 134 kW s 1/2 m -2, and for the polyurethane cushion CHF = 11 kW m -2 and TRP = 55 kW s 1/2 m -2.

31. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 31 Problem Set 3: Problem 6 6. 6.Consider the room of Problem 3. For upper layer temperatures of 300°C and 400°C, calculate the flame velocity on a wooden floor and on a polyurethane cushion 30 seconds and 1 minute after the flux is applied. (Assume that the convective cooling is governed by h = 9.0 W m -2 K -1 ).

32. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 32 Burning Rates of Common Items * In the open (review) * Limited by ventilation * Enhanced by radiation

33. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 33 Wooden Cribs (2)

34. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 34 Wooden Cribs D = stick thickness (m)D = stick thickness (m) S = spacing between sticks (m)S = spacing between sticks (m) h c = height of crib (m)h c = height of crib (m) N = number of rows = h c / DN = number of rows = h c / D n = number of sticks per rown = number of sticks per row L = length of each stick (m) {L >> D}L = length of each stick (m) {L >> D}  = density of sticks (kg m -3 )  = density of sticks (kg m -3 ) m o = initial mass of crib (kg) = N n  D 2 Lm o = initial mass of crib (kg) = N n  D 2 L

35. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 35 Steady-State Burning of Wooden Cribs Fuel surface controlled burning: Stick surfaces burn freely {S >> D}Fuel surface controlled burning: Stick surfaces burn freely {S >> D} Eqn (8-5) Eqn (8-5) = mass loss rate of crib (kg s -1 ) = mass loss rate of crib (kg s -1 ) t o = time at which steady burning is established (s)t o = time at which steady burning is established (s) v p = surface regression rate (m s -1 )v p = surface regression rate (m s -1 )

36. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 36 Steady-State Burning of Wooden Cribs Crib porosity controlled burning: Burning controlled by rate of flow of air & combustion products through holes in crib {S << D}Crib porosity controlled burning: Burning controlled by rate of flow of air & combustion products through holes in crib {S << D} Eqn (8-6) Eqn (8-6) for t > t o : is given by lesser of Eqns (8-5) & (8-6)for t > t o : is given by lesser of Eqns (8-5) & (8-6)

37. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 37 Growth Rates - Burning of Wooden Cribs Assume crib is ignited at bottom / centreAssume crib is ignited at bottom / centre t o = time at which steady burning is established (s)t o = time at which steady burning is established (s) For t < t oFor t < t o Eqn (8-7) Eqn (8-7) t o = time Eqn (8-7) yields lesser of Eqns (8-5) & (8-6)t o = time Eqn (8-7) yields lesser of Eqns (8-5) & (8-6)**************************************************************** For a crib ignited at bottom / centre and whose steady- state burning is fuel-surface controlled: t o ~ 15.7 n (s)For a crib ignited at bottom / centre and whose steady- state burning is fuel-surface controlled: t o ~ 15.7 n (s)

38. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 38 Wooden Cribs - Heat Release Rate The heat release rate is given byThe heat release rate is given by Eqn (8-8) with H ch = 12.4 kJ g -1with H ch = 12.4 kJ g -1 Knowing one can also calculate, radiative and convective components of heat release rate, and rates of generation of CO and soot.Knowing one can also calculate, radiative and convective components of heat release rate, and rates of generation of CO and soot.

39. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 39 Wooden Cribs in an Enclosure Radiation from upper layer has little impact on because fire is largely “self-contained” with many surfaces “seeing” each other.Radiation from upper layer has little impact on because fire is largely “self-contained” with many surfaces “seeing” each other. If fire is limited by ventilation, will be reduced because H ch and are both reduced.If fire is limited by ventilation, will be reduced because H ch and are both reduced.

40. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 40 Post-flashover Fires Involving Wooden Cribs Harmathy (1972) identified two burning regimes for room fires involving wooden cribs: ventilation-controlled & fuel-surface controlledHarmathy (1972) identified two burning regimes for room fires involving wooden cribs: ventilation-controlled & fuel-surface controlled = mass loss rate of fuel (kg s -1 ) = mass loss rate of fuel (kg s -1 )  = ventilation parameter (kg s -1 )  = ventilation parameter (kg s -1 ) = A f = exposed surface area of fuel (m 2 )A f = exposed surface area of fuel (m 2 )

41. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 41 Post-flashover Fires Involving Wooden Cribs

42. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 42 Post-flashover Fires Involving Wooden Cribs Post-flashover fire is ventilation-controlled ifPost-flashover fire is ventilation-controlled if  / A f < 0.63 kg m -2 s -1 Eqn (8-9) Fuel mass loss rate isFuel mass loss rate is Eqn (8-10) Eqn (8-10) Least of Eqns (8-5), (8-6), (8-7) or (8-10) appliesLeast of Eqns (8-5), (8-6), (8-7) or (8-10) applies

43. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 43 Wooden Pallets (1)

44. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 44 Wooden Pallets

45. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 45 Wooden Pallets - Peak Burning (in the open) Eqn (8-11) Eqn (8-11)

46. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 46 Wooden Pallets - Theory vs. Experiment

47. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 47 Wooden Pallets For non-standard pallet sizes,For non-standard pallet sizes, Eqn (8-12) Heat release rate per unit floor area covered by pallet stackHeat release rate per unit floor area covered by pallet stack

48. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 48 Wooden Pallets - Mass Loss Rate The heat release rate & mass loss rate are related byThe heat release rate & mass loss rate are related by Eqn (8-13) Implicitly assumed that H ch = 12 kJ g -1Implicitly assumed that H ch = 12 kJ g -1 Knowing can calculate, radiative and convective components of heat release rate, and rates of generation of CO and soot.Knowing can calculate, radiative and convective components of heat release rate, and rates of generation of CO and soot.

49. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 49 Wooden Pallets in an Enclosure Radiation from upper layer has little impact on because fire is largely “self-contained” with many surfaces “seeing” each other.Radiation from upper layer has little impact on because fire is largely “self-contained” with many surfaces “seeing” each other. If fire is limited by ventilation, will be reduced.If fire is limited by ventilation, will be reduced. Fuel mass loss rate isFuel mass loss rate is Eqn (8-10) Eqn (8-10) Smaller of Eqns (8-11) or (8-10) appliesSmaller of Eqns (8-11) or (8-10) applies

50. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 50 Unusual Nature of Wooden Cribs & Pallets Early work on enclosure fires used wood cribs to achieve reproducible firesEarly work on enclosure fires used wood cribs to achieve reproducible fires However, burning surfaces of wooden cribs & pallets are shielded from environment within the enclosureHowever, burning surfaces of wooden cribs & pallets are shielded from environment within the enclosure Consequently rate of burning is relatively insenstive to the thermal environmentConsequently rate of burning is relatively insenstive to the thermal environment When wood is present as wall lining, however, there is a large exposed area that is sensitive to the thermal environmentWhen wood is present as wall lining, however, there is a large exposed area that is sensitive to the thermal environment

51. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 51 Diffusion Flames (in the open)

52. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 52 Rate of Burning (in the open) Eqn (8-14) Eqn (8-14)

53. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 53 Heat Release Rates (in the open) Fires burning in the open are well-ventilatedFires burning in the open are well-ventilated Actual (chemical) heat release rate / unit area isActual (chemical) heat release rate / unit area is (kW m -2 ) Eqn (8-15) (kW m -2 ) Eqn (8-15) H ch = Actual ( chemical ) heat of combustion (kJ / g) H ch = Actual ( chemical ) heat of combustion (kJ / g)

54. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 54 Consider a material burning in an enclosure but getting sufficient air for combustion?

55. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 55 Rate of Burning (Mass Loss Rate) Eqn (8-3) Eqn (8-4) 2nd term can be estimated by open burning models2nd term can be estimated by open burning models

56. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 56 Example: Study of effect of trapping heat on rate of burning of slab of PMMA (0.76m x 0.76 m) (*)

57. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 57Observations Trapping of heat (radiation from hot layer) increases steady-state burning rate of PMMATrapping of heat (radiation from hot layer) increases steady-state burning rate of PMMA Trapping of heat (radiation from hot layer) reduces time to steady-state burning  rate of flame spread across PMMA also increasesTrapping of heat (radiation from hot layer) reduces time to steady-state burning  rate of flame spread across PMMA also increases

58. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 58 Burning rate in post-flashover fires involving fuels with exposed surfaces is enhanced by radiationBurning rate in post-flashover fires involving fuels with exposed surfaces is enhanced by radiation Large burning rates inhibit inflow of air so increase equivalence ratio  reduced heat release (inside)Large burning rates inhibit inflow of air so increase equivalence ratio  reduced heat release (inside) Heat release rate still can be ventilation-controlledHeat release rate still can be ventilation-controlled

59. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 59 Burning rate as function of radiant intensity at ceilingBurning rate as function of radiant intensity at ceiling

60. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 60 Burning rate as function of radiant intensity at ceiling  ethanol (L V = 850 J g -1 )  PMMA pool (L V = 1,600 J g -1 )  polyethylene (L V = 22,00 J g -1 )  wood (L V = 1,340 J g -1 )  PMMA crib — ethanol in open

61. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 61 Pool fire burning rates in the open & in enclosuresPool fire burning rates in the open & in enclosures f ex = 1 - 1/  (excess fuel factor)(some fuel burns outside)f ex = 1 - 1/  (excess fuel factor)(some fuel burns outside)

62. Carleton University, 82.583, Fire Dynamics II, Winter 2003, Lecture # 8 62References D. Drysdale, An Introduction to Fire Dynamics,Wiley, 1999, Chap 1 F.W. Billmeyer, Textbook of Polymer Science, Wiley, 1984, Chap 1 Donald R. Askeland, Science and Engineering of Materials, Chapman & Hall, 1990, Chapter 15 C.F. Cullis and M.M. Hirschler, The Combustion of Organic Polymers, Oxford Science Publications, 1981, Chapter 1 C.L. Beyler and M.M. Hirschler, "Thermal Decomposition of Polymers" Section 1 / Chapter 7, SFPE Handbook, 2nd Ed. (1995) C.F. Cullis and M.M. Hirschler, The Combustion of Organic Polymers, Oxford Science Publications, 1981, Chapter 1