Validation and Usage of Digital Image Correlation for High Temperature Deformation Measurement in Modern Prestressing Steel

*Speaker

John Gales* PhD candidate

Supervisor:

Dr Luke Bisby 2nd Supervisor: Tech. Supervisor

Dr Martin Gillie Dr Tim Stratford

What is Prestressing Steel (PS)?

Conventional steel rebar ‘Unbonded’ PS steel

• Advantages of post-tensioning concrete with PS steel

- Rapid construction

- Shallow floors (high ceilings)

- Increased span lengths

- Reduces building materials

Highly

Optimized

UPT 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 than typical 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 UPT Slab Behaviour in Fire is Unknown

• UPT optimization increases susceptibility to fire:

- Prestressing steel more sensitive to high temperature than mild steel

- Spalling of concrete cover (HS concrete, precompression can contribute to this)

- Tendons run continuous, local damage effects the entire floor

• Only standard furnace tests of simple span slabs are available:

- modern construction?, building

materials?, real fires?

2008 standard fire test (Kelly and Purkiss)

Can standard fire tests predict true fire behaviour of these UPT buildings?

• Key Biscayne, USA, 2000: – UPT tendons continuous across 7 interior bays (> 50m) – Localized fire on 2nd floor spread vertically to 7th floor over 2

bays – “triggered progressive failure of the UPT slab well beyond

the zone of visible damage”

Po

st a

nd

Ko

rman

, En

gin

eeri

ng

New

s R

eco

rd J

un

e 1

2th

20

00

Tendons released by fire

Visible fire damage

Columns

UPT slabs

Localized Fire Damage

• Localized fires may be due to spalling, travelling, ceiling jets…

• Queens university tests in 2009 showed that tendon rupture is more probable under localized heating

• Complex stress relaxation / strength interaction - influenced by a permanent time dependent damage called, creep

Localized UPT tendon tests (strong back tests) conducted at Queens university in Canada

High Temperature Creep

•Creep is a time, temperature and load dependent deformation

•Possible to express creep through combined temperature time variable in a heated tensile test at constant load

‘Dorn-Harmathy creep model’

•Creep parameters for PS steel are from dated (1970) and different material compositions than modern counterparts- Errors in creep modelling?

Temperature compensated time (θ)

Cre

ep

str

ain

(e

cr)

Δθ

Δecr Secondary Creep rate=Z= Δecr

Δθ

ecr,0

Pri

ma

ry

Cre

ep

Se

co

nd

ary

Cre

ep

Te

rtia

ry

Cre

ep

t

RTH dte0

/

)2(cosh2ln

0,0 cre

Z

cr

cr

ee

Creep in Prestressing Steel

Temperature Compensated Time

Predicting Creep for Stress Relaxation?

• Over estimation of (creep)

• May be due to parameter extrapolation, older steel derivations, inaccuracies in derivation

• Typical Queens university high temperature stress relaxation test

Regimes

•Transient heating

•Steady state hold

•Cooling

How Can We Derive Modern Creep Parameters?:

Digital Image Correlation (DIC) in transient and steady state uniaxial tensile tests to measure deformation

What is Digital Image Correlation?

Previously validated Image correlation algorithm (by Dr Andy Take, Queens) are used to measure virtual deformation (strain) through a series of sequenced test photos

Two photos showing deformation from a

uniaxial ambient strength test

DIC Validation (1 of 2)

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0 500 1000 1500 2000

Stra

in (

mm

/mm

)

Time (s)

MSG1MSG2MSG3PIV strain

All strain gauges debonded

Elas

tic

ran

ge

Pla

stic

ra

nge

• Uniaxial ambient strength test with comparison of DIC and bonded foil strain gauges

0

500

1000

1500

2000

2500

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

Str

ess

(M

Pa

)

Strain

MSG1

MSG2

MSG3

PIV strain

DIC Validation (2 of 2)

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0 100 200 300 400 500

Therm

al

str

ain

(m

m/m

m)

Temperature (°C)

EC2 mild steel DIC mild steel

EC2 PS steel DIC PS steel

• Uniaxial transient thermal expansion test (unloaded) with comparison of DIC to theoretical calculation(EC2)

Steady and Transient State Creep Tests

• Steady and transient state testing are suppose to be equivalent in the ‘Harmathy Dorn creep model’ based on activation energy

Is this really valid and to what extent?

Temperature compensated time (θ)

Cre

ep

str

ain

(e

cr)

Δθ

Δecr Secondary Creep rate=Z= Δecr

Δθ

ecr,0

Pri

ma

ry

Cre

ep

Se

co

nd

ary

Cre

ep

Te

rtia

ry

Cre

ep

t

RTH dte0

/

)2(cosh2ln

0,0 cre

Z

cr

cr

ee

Creep in Prestressing Steel

Temperature Compensated Time

Harmathy-Dorn Creep Model

Creep Test Equivalency

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

0.045

0.050

0 5E-20 1E-19 1.5E-19 2E-19 2.5E-19 3E-19 3.5E-19 4E-19

Temperature compensated time (θ) (hr)

To

tal c

ree

p s

tra

in (

ec

r)

Steady State

Steady State Harmathy Prediction

Transient

Transient Harmathy Equation Prediction

0.000

0.002

0.004

0.006

0 2.5E-20

θ(hr)

ecr

0.000

0.002

0.004

0.006

0 2.5E-20

θ (hr)

ecr

• Steady state and transient state test at 700 MPa

Transient Steady State

Varying Stress and Repeatability

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0 5E-20 1E-19 1.5E-19 2E-19 2.5E-19 3E-19 3.5E-19 4E-19

Temperature compensated time (θ)

To

tal

creep

str

ain

(ecr)

690 Mpa Steady State (427C)

690 Mpa 2C/min

800 Mpa 2c/min

1000 MPA 2c/min (Test 1)

1000 Mpa 2c/min (Test 2)

1000MPa 10c/min

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0 5E-21 1E-20 1.5E-20 2E-20 2.5E-20

Temperature Compensated Time (θ)

To

tal creep

strain

(ecr)

1000 MPA 2c/min (Test 1)

1000 Mpa 2c/min (Test 2)

1000MPa 10c/min

Varied stress levels

1000MPa

800MPa

700MPa

All tests 1000MPa 2C/min 2C/min

10C/min

Repeatability

Parameter Updates

• Secondary creep parameter updates (old vs new)

• New creep parameters do not match old, suggest less creep deformation in modern PS steel

Preliminary Modelling with New Parameters

• New creep parameters are conservative end

Key Insights:

• DIC deformation measurement is adequate when shown with traditional strain prediction and instrumentation

• Less limitations for usage (SGs fall off, extensometers can break in the tertiary phase of creep)

• New creep model params. more accurate for high temperature PS stress relaxation (creep) predictions

• More repeat testing needed at both transient and steady state in order to truly understand and quantify variability and validation using the Dorn-Harmathy creep model for prestressing steel

Impacts of Research

• Modern UPT concrete techniques continue to promote innovation in construction; however....

- These innovations potentially sacrifice safety using out of date prescriptive testing and guidance

- Research being conducted will allow us to rationally design for real fires in real buildings

One Museum Park

Thank you

For additional information

Email: j.gales@ed.ac.uk

Web: www.see.ed.ac.uk/fire

The Ove Arup Foundation

mailto:j.gales@ed.ac.ukhttp://www.see.ed.ac.uk/fire