MODELING OF CONCRETE MATERIALS AND STRUCTURES

Kaspar Willam, University of Colorado at Boulder

With contributions by Dr. Inkyu Rhee, Dr. Jaesung Lee and Keun Lee, UC-Boulder

Class Meeting #7: Discrete Interface Formulations

Zero-Thickness Interface Models:
Cohesive Damage and Elasto-Plastic Traction-Separation Models
Inherent Length Scale:
Fracture Energy Basis GIf and GII
f of Softening
Example Problems:
- Thermal Softening of Axial Strength due Mismatch
- High Temperature Bond: Pull-Out Experiment vs FE Model
Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007

MULTI-SCALE ASPECTS OF CONCRETE MATERIALS

Four Scales of Observation, Continuum vs Particle Models-Size Effects?
Meter Level
(Real Structure)
Millimeter Level
(Concrete)

Micrometer Level
(Cement Paste)
X1000
Nanometer Level
(Calcium Silicate Hydr
X1000

X1000

HIGH TEMPERATURE EFFECTS IN CONCRETE MATERIALS

Concrete Spalling: Temperature gradient vs pore pressure effects.
Transient Effects: Load-induced thermal strains-LITS.
Andersen and Thelandersen [1984]: Thermal sweep experiments.

10

1
TEST 1 C / min
TEST 5 C / min
0.8

a = -0.225 f'c

11

/ f'c

Strain [%]

0.6
0

0.4

a = -0.675 f'c

-5
0.2

= -0.45 f'c

-10

0
0

200

400

600

Temperature, C

800

1000

200

400

600

Temperature, C

800

1000

THERMAL MISMATCH OF CONCRETE MATERIALS

Heterogeneity of Mesostructure:
Aggregate vs Cement Paste and Bond of ITZ

Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007

ZERO-THICKNESS INTERFACE FORMULATION

Weak ITZ-Cohesive Traction-Separation Relations:
Softening Damage and Plasticity Models
Fracture Energy Basis:
Coupling of Mode I and Mode II Softening Damage/Plasticity Models
(softening of tensile bond vs tangential shear strength).

]2

|
[|u s

|]
[|u n

1
|]
[|u s

Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007

|]
[|u n

BIMATERIAL INTERFACE CONDITIONS

Perfect Bond:
[|uN |] = uaN ucN = 0 and [|tN |] = taN tcN = 0
Weak Discontinuities: all strain components exhibit jumps across interface
except for aT T = cT T restraint.
Note: Jump of tangential normal stress, Ta T 6= Tc T .
Imperfect Contact:
[|uN |] = uaN ucN 6= 0 whereas [|tN |] = taN tcN = 0
Strong Discontinuities: all displacement components exhibit jumps across
interface.
Note: FE Displacement method enforces traction continuity in weak sense
only, hence [|tN |] 6= 0.
Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007

ZERO-THICKNESS INTERFACE FORMULATION

Deformed FE Mesh after Thermal Sweep:
Diagonal separation of RVE due to slip.

Axial Thermal Stress due to Temperature Sweep:

Experimental observations vs computational results.
1

-40
o

11 = 0

TEST 1 C/min
o
TEST 5 C/min

0.8
T
0.6

0.4

0.2

200

400
o
Temperature [ C]

600

800

Stress_YY[Mpa] at Outer Edge of Confined Surfaces

- 11 / fc

ZERO-THICKNESS INTERFACE FORMULATION

Continuum
Discrete, b=0.1
Discrete, b=0.5
Discrete, b=1.0
-30

-20

-10

200

400
Temperature [C]

Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007

600

800

HIGH TEMPERATURE BOND

Pull-Out Test Setup: Residual Experiments after Cooling.
Test Variables:
-

Coated-Uncoated Rebars
Slow-Rapid Heating Rates
Target Temperatures: 20oC, 200oC, 400oC, 600oC
Natural-Water Cooling

HIGH TEMPERATURE BOND

Pull-Out Test Program: Residual Experiments after Cooling.

HIGH TEMPERATURE BOND

Pull-Out Test Results:
Bond-Slip Relationships for T=20, 200, 400, 600C (rapid heating)

2.0
Uncoated ref.

1.8

UR15_200N

Bond strength (ksi)

1.6

UR15_400N
UR15_600N

1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0

0.01

0.02

0.03
Slip (inch)

0.04

0.05

0.06

HIGH TEMPERATURE BOND

Pull-Out Test Results:
Bond Strength of slow heating tests up to T = 20o, 200o, 400o, 600oC

1.80
1.60

1.67

1.40
Bond strength (ksi)

UR2 test series (Natural cooling)

CR2 test series (Natural cooling)

1.29

1.44

1.20
1.11
1.00

0.99

0.80
0.60
0.60
0.40

0.36
0.29

0.20
0.00
25

200

400
Temperature (C)

600

0.00
800

HIGH TEMPERATURE BOND

Pull-Out Test Results:
Bond Strength of rapid heating tests up to T = 20o, 200o, 400o, 600oC

1.80
1.60

Bond strength (ksi)

1.40

1.44
1.29

1.20
1.00

UR15 test series (Natural cooling)

CR15 test series (Natural cooling)

1.27
0.99

0.94

0.80
0.60
0.40

0.35
0.33

0.20

0.16
0.00
25

200

400
Temperature (C)

600

0.00
800

PULL-OUT PROBLEM
Finite Element Model of Pull-Out Test Specimen

BIMATERIAL INTERFACE CONDITIONS

Perfect Bond:
co
st
co
[|uN |] = ust
N uN = 0 and [|tN |] = tN tN = 0
Weak Discontinuities: all strain components exhibit jumps across interface
co
except for st
T T = T T restraint.

Note: Jump of tangential normal stress, TstT 6= TcoT .

Imperfect Bond:
co
st
co
[|uN |] = ust

u
=
6
0
whereas
[|t
|]
=
t

t
N
N
N
N
N =0
Strong Discontinuities: all displacement components exhibit jumps across
interface.

Note: FE Displacement method enforces traction continuity in weak sense

only, hence [|tN |] 6= 0.
Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007

MISMATCH OF THERMAL EXPANSION

3-D Thermoelastic Response:
co + coT
co = E 1
co :

and

st + stT
st = E 1
st :

co
Interface Bond Kinematics: st
z = z

Interface Bond Traction: rst = rco

Radial Contact Stress:
r =

EcoEst
[co st]T + HO
Estcon Ecost
250

0.0000225

200

0.00002

150

0.0000175

Radial stress@MPaD 100

0.000015

50

0.0000125

alpha

0.00001
0

100

200
300
Temperature

400

500

100

200
300
Temperature

400

500

TRIAXIAL CONCRETE MODEL

Damage-Plasticity Model in ABAQUS: Lee & Fenves [1998]

PLANE STRESS FAILURE ENVELOPE

Damage-Plasticity Model: Lee & Fenves [1998]

PULL-OUT PROBLEM
Evolution of Tensile Damage at Different Outside Temperatures
T = 100, 300, 580oC

PULL-OUT PROBLEM
Evolution of Compression Damage at Different Outside Temperatures
T = 100, 300, 580oC

PULL-OUT PROBLEM
Circumferential Stress Distribution at Outside Temperatures
T = 100, 300, 580oC

0
0

10

20

30

Stress [MPa]

-5
stress_t at T=108

-10

stress_t at T=300
stress_t at T=580

-15

-20

-25
Radius [mm]

40

50

60

PULL-OUT PROBLEM
Circumferential Stress Contours at Outside Temperature T = 580oC

Bond stress b in MPa

PULL-OUT PROBLEM
Bond Stress-Axial Slip Response (cold vs hot conditions)
25
cold
warm
20
cooled
cold exp. Kurz
15
10
5
0
0

Displacement in mm
Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007

PULL-OUT PROBLEM
Axial Stress Contours after Heating to T = 580oC
S, S22

Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007

PULL-OUT PROBLEM
Compression Damage after Heating to T = 580oC
DAMAGEC

Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007

CONCLUDING REMARKS

Thermal Mismatch of Steel and Concrete:

Full Bond vs Cohesive Interface Formulation - Loss of Confinement.
Mesoscale Analysis of Concrete RVE:
Interaction of tensile cracking and shear debonding along weak
zero-thickness interface layers-ITZ.
High Temperature Pull-Out:
Residual experiments and axisymmetric FE simulation of
steel-concrete contact separation.

Class #7 Concrete Modeling, UNICAMP, Campinas, Brazil, August 20-28, 2007