Fire Engineering Research: Key Issues for the Future Post‐tensioned Concrete Structures in Fire John Gales Supervision: Luke Bisby, Co supervision: Martin Gillie‐ modelling, Phase 1 and 3 Tim Stratford‐ experimentation, Phase 2 and 3
Fire Engineering Research: Key Issues for the Future Post‐tensioned Concrete Structures in Fire
John Gales
Supervision: Luke Bisby, Co supervision: Martin Gillie‐modelling, Phase 1 and 3
Tim Stratford‐ experimentation, Phase 2 and 3
What are post‐tensioned buildings?
Conventional steel rebar Prestressing (PS) steel
• Advantages of post‐tensioning concrete with PS steel for load balancing
‐ Thin floors (high ceilings)‐ Increased span lengths ‐ Reduces building materials ‐ Rapid construction
Highly optimized
Typical post‐tensioned buildings
Modern BPT building, UK
Modern UPT building, USA
Antiquated (1960s) UPT building, USA
Novel building optimization
“ Today’s flat‐slab post‐tensioned buildings, for example, with columns spaced (12 m) on center and span‐depth ratios of 40 are more complex and require more engineering attention thantypical flat‐slab buildings of 40 years ago, with columns spaced at (6 m) on center and span‐depth ratios of 20. ” ‐Randall Poston (chair ACI 318)
• Current guidance is dated and has not kept up with modern optimization trends
Real PT slab behaviour in fire is debatable
• PT optimization increases susceptibility to fire:
‐ PS steel more sensitive to strength loss in high temperature
‐ Spalling of concrete cover (HS concrete, precompression of slab)
‐ Unbonded tendons run continuous, local damage WILLeffect the entire floor (Key Biscayne demolition)
• Code guidance is based on (often dated) standard furnace tests of simple span slabs: ‐ modern construction?, building materials?, real fires?
PT Standard fire test (Kelly and Purkiss, 2008)
The PhD
• Phase 1 Fire code assessment for unbonded PS steel rupture (spalling, and variable heating length)
• Phase 2 High temperature mechanical behaviour of modern PS steel (softening, strength and creep)
• Phase 3 three large‐scale continuous PT slab tests under localised heating
• Side projects while I wait for Phase 3 to begin (curing time delayed)
Temperature compensated time (θ)
Cre
ep s
train
(ecr
)
∆θ
∆ecr Secondary Creep rate=Z= ∆ecr ∆θ
ecr,0
Prim
ary
Cre
ep
Sec
onda
r yC
reep
Terti
ary
Cre
ep
Phase 1: Localized fire damage to unbonded PS steel
• 2009 Tests demonstrated unbonded PS steel rupture is more probable under localized heating ‐ influenced by creep
• Localized fires may be due to spalling, travelling, ceiling jets…
Localized heated UPT tendon tests (strong back tests) conducted in my masters
Lower ratio of heating, failed tendons at equivalent temperatures
Phase 1: Localized fire damage to unbonded PS steel
• IBC, and EC2 analyzed with simple tendon rupture modelling with creep (time, temp, load dependent) relation and heat transfer ( ASTM E119 curve)
Parametric analysis: Heated length ratio, spalling, specified concrete cover
Phase 1 results Performance based guidance not
clearly specified in codes with respect to losing unbonded PS steel in a fire
Considerations to made; restraint, bonded reinforcing, spalling mitigation
American IBC code was unconservative Real unbonded PS steel behaviour more
severe than Phase 1 modelling, new modelling parameters needed (Phase 2)
Results have tied in directly or inspired related PhD projects at Edinburgh(spalling, concrete cover influence using FEM)
0.00
100.00
200.00
300.00
400.00
500.00
600.00
700.00
800.00
900.00
1000.00
0 0.2 0.4 0.6 0.8 1
Time (hr)
Tem
pera
ture
(°C)
Cover (0mm)Cover (1mm)Cover (2mm)Cover (3mm)Cover (4mm)Cover (5mm)Cover (6mm)Cover (7mm)Cover (8mm)Cover (9mm)Cover (10mm)Cover (11mm)Cover (12mm)Cover (13mm)Cover (14mm)F|IR|E
0
100
200
300
400
500
600
700
800
900
1000
0 0.2 0.4 0.6 0.8 1
Time (hr)
Tem
pera
ture
(°C)
Cover (0mm)Cover (1mm)Cover (2mm)Cover (3mm)Cover (4mm)Cover (5mm)Cover (6mm)Cover (7mm)Cover (8mm)Cover (9mm)Cover (10mm)Cover (11mm)Cover (12mm)Cover (13mm)Cover (14mm)F|IR|E
Example heat transfer compensated for spalling input (200mm slab)
Phase 2: Modern PS steel behaviour in high temperature
Used Digital Image Correlation (DIC) in uniaxial tensile tests to measure deformation and cross section reduction
Phase 2: Modern PS steel behaviour in high temperature
• DIC patch correlations based on HT paint speckle pattern Patch A 750Px gauge length = 37.5mm Patch B
Pixel (y)
Pixel (x)
Stress = 0 MPa
Stress = 1950 MPa
Patch A Patch B
Method needed validation for current use……………….
Phase 2: Modern PS steel behaviour in high temperature
• DIC to bonded foil strain gauges and extensometer • DIC cross section to Poisson constant volume theory• DIC to theoretical thermal expansion calculation (EC2)
Phase 2: Modern PS steel behaviour in high temperature
• Creep behaviour using temperature compensated time.• PS steel types considered; ASTM 421‐1970, ASTM 416‐2008, and BS 5896‐2011 (all of different composition, but considered structurally equivalent)
Phase 2 results• Uniaxial creep tests at Steady state and Transientinvestigating equivalency
Results appeared similar (creep parameters were identical magnitudes; at 690MPa and 1000MPa stress levels)
Change in transient test heating rate had same magnitudes
Phase 2 results• Tertiary creep as manifestation of localized yielding
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0 1E-19 2E-19 3E-19 4E-19 5E-19 6E-19
εcr
Virtual creepstrain (steadystate)
Harmathyequationw/ASTM 416params and areareduction
690 MPa
12
13
14
15
16
0 1E-19 2E-19 3E-19 4E-19 5E-19 6E-19
Temperature compensated time, θ (x 10-19 hrs)
Are
a (m
m2 )
Necking region
3300Px distance
1 2 3 4 5 6
Creep curve initiates runaway (tertiary) failure when a local necking region develops
Result appears in transient test
Possible to model, but relations produce error
Phase 2 results• Strength tests with true stress in steady state; Implicit creep strength tests comparison underway (post peak softening).
Reduction ratios matched well to Eurocode
Loading rate decrease, decreased yield point
True strength retention at elevated temperature better than EC2 until post peak softening occurs
0
500
1000
1500
2000
2500
0 0.05 0.1 0.15 0.2 0.25 0.3
Stre
ss (M
Pa)
Strain
100ºC (true)
300ºC (true)
400ºC (true)
500ºC(true)
100
500ºC
400ºC
300ºC
Phase 2 results• Creep models were compared with the results of the locally heated strong back tests (varied transient and steady state heating with cooling)
600
650
700
750
800
850
900
950
1000
1050
0 1 2 3 4 5 6
Test duration (hrs)
Unb
onde
d PS
stre
ss le
vel
(MPa
)
ASTM 416 (2008)actual stress ASTM 416 (2008)modelASTM 421 (1970)modelBS 5896 (2011)
Transient heating (2°C/min) Steady heating (400°C)
Cooling (natural)
Creep model accuracy function of heating rate and metallurgy
Error at 2% for2⁰C/min growing to 7% error at 30⁰C/min
Third creep phase not considered yet
Phase 3: Continuous post‐tensioned concrete slabs under localized fire
• Two UPT and One BPT , 1‐hour rated EC2 slabs
Tests planned for this summer (6+ months, low MC%) Restraining forces measured from steel columns (stiffness based on
representative concrete columns Applied loading Realistic span to depth ratio (>40) Bonded steel provided Thermocouples (x24), Linear Potentiometers (x8), Load cells (x2) Radiant panel heating (locally heated)
Phase 3: Continuous post‐tensioned concrete slabs under localized fire
Issues and problems with Phase 3:• What do we want to do with the results…..‐ Apriori and Aposteriori round robin modelling?‐ In house modelling (FEM packages)?
• Instrumentation‐What should we be measuring and what does it mean?‐Motion imaging? (2D DIC, 3D tracking?)
• Pretesting‐ Ambient tests before heating?
• Intangibles; prestressing the slabs?
Current collaborative side projects• Project 1: The History of Fire Safety
Engineering (The full story is not recorded) Traditional and non traditional construction Large scale testing (Modern and antiquated)
ICEM15 conference this July in Porto Fire behaviour, dynamics and design philosophy
• Project 2: Axis distance vs. clear cover of miniature PS slabs exposed to ISO 834. Should this design rule change?
• Project 3: Open access repositories for historical fire engineering photographs and articles
Thank you
For additional information
Email: [email protected] reading:
•http://www.eng.ed.ac.uk/fire/2009‐phd‐john.html
•Results of Phase 1 can be consulted in the Journal of Structural Fire Engineering and
Fire Safety Journal (see web link for references)
•Some preliminary results of Phase 2 will be presented at SIF 2012 conference in Zurich
•Phase 3 is currently in progress targeting 2013 for completion.