5/29/2013

Behavior and Design of 
Concrete‐Filled Composite 
Columns

Roberto T. Leon
Virginia Tech, Blacksburg, VA
Jerome F. Hajjar
Northeastern University, Boston, MA
Larry Griffis
Walter P. Moore, Austin, TX

Scope
• Brief introduction to composite columns (LG)
• Research motivation and experimental results (RL)
• Analytical modeling and system studies (JH)
• Conclusions and design recommendations (LG)

In‐Kind:

Work is based on the dissertations of:
Tiziano Perea, UAM, Mexico City (MX) – Georgia Tech
Mark Denavit, SDL, Atlanta (GA) – UIUC

1
 5/29/2013

Composite or hybrid system (concrete & steel)
System which combines the advantages of concrete and structural steel

Concrete Structural steel
* Rigid * Economic * High strength * Ductile
* Fire resistant * Durable * Easy to assembly * Fast to erect

Frames with CFT columns

• Steel tube confines concrete
• Concrete restricts the buckling of the steel tube
• Increase in strength & deformation of the concrete 
• Delay in the buckling of the steel tube

Frames with SRC columns

• Steel element supports the construction loads
• The concrete gives final stiffness and fire resistant
• Shear connections become FR once concrete is cast
• System fast to erect & build (redundancy)

Uses for Composite Columns
• Extra capacity in concrete column for no increase in 
dimension
• Large unbraced lengths in tall open spaces
– Lower story in high rise buildings
– Airport terminals, convention centers
• Corrosion, fireproof protection in steel buildings
• Composite frame – high rise construction
• Transition column between steel, concrete systems
• Toughness, redundancy as for blast, impact

2
 5/29/2013

Composite Systems
• Perimeter moment frames for 
stiffness in hurricane zones.
• Extension to seismic based on 
Japanese experience.
• Distributed systems vs. 
supercolumns

Buildings with SRC Columns (Martinez‐Romero, 1999 & 2003)

3
 5/29/2013

Composite Braced Frame

Bank of China
Hong Kong

Composite Column

Bank of China
Hong Kong

4
 5/29/2013

Composite Moment Frame

“Tube” Design

3 Houston Center
Houston, Texas

Composite Column Forming

5
 5/29/2013

“Tree Columns”
Composite Columns

3 Houston Center
Houston, Texas

Composite “Erection Columns”

6
 5/29/2013

Composite Columns
Reinforcement Cage

Composite Shear Walls

7
 5/29/2013

Composite Braced Frame

2 Union Square
Seattle, Washington

Composite Frame Construction

Dallas, Texas

8
 5/29/2013

Composite Frame Construction

Possible configurations in composite columns

a) SRC b) Circular and Rectangular CFT

c) Combinations between SRC and CFT

9
 5/29/2013

Flexibility
Sizes and Shapes

Filled Composite Column
(Covered in this Webinar)

Round HSS Square or Rectangular HSS

10
 5/29/2013

Encased Composite Column

Motivation for Research
• Lack of design information for the stiffness of 
columns to be used for buckling and lateral 
rigidity calculations
• Lack of knowledge on the interaction between 
axial load and bending at ultimate (2D and 3D)
• Lack of knowledge on system factors (force 
reduction and deflection amplification for seismic 
design)
• Gaps in data for slender columns (local and 
overall buckling)

11
 5/29/2013

(1) Flexural rigidity for lateral forces
for calculating 
• Concrete‐only or Steel‐only column 
capacity, not 
• Semi‐empirical : EIeff  sEI
s s cEI
c c
for lateral 
analysis

• Advanced computational analysis:
HSS
Section

Fiber element  Finite element 
analysis analysis

(2) Behavior factors for seismic design?
ASCE/SEI 7‐10, Table12‐2‐1

Selected Systems R  Cd
S‐SMF (Steel Special Moment Frames): 8.0 3.0 5.5
C‐SMF (Composite Special Moment Frames): 8.0 3.0 5.5
S‐IMF (Steel Intermediate Moment Frames; SDC B, C, D): 4.5 3.0 4.0
C‐IMF (Composite Intermediate Moment Frames; SDC B, C): 5.0 3.0 4.5
S‐OMF (Steel Ordinary Moment Frames; SDC B, C, D): 3.5 3.0 3.0
C‐OMF (Composite Ordinary Moment Frames; SDC B!!): 3.0 3.0 2.5
SCBF (Steel Concentrically Braced Frames): 6.0 2.0 5.0
C‐SBF (Composite Special Braced Frames): 5.0 2.0 4.5
OCBF (Composite Ordinary Conc. Braced Frames; SDC B‐F): 3.25 2.0 3.25
C‐OBF (Composite Ordinary Braced Frames; SDC B, C!!): 3.0 2.0 3.0

12
 5/29/2013

(3) Lack of Slender Experimental Tests
Databases compiled by León et al., 2005 and Goode et al., 2007
2.5
Pexp/Po
1375 Circular CFT
2.0 Pn/Po
• 912 columns
• 463 beam‐columns AISC
1.5
798 Rectangular CFT

P/Po
• 524 columns 1.0
• 274 beam‐columns
267 Encased SRC 0.5
• 119 columns
• 148 beam‐column 0.0 
0.00 0.50 1.00 1.50 2.00 2.50 3.00
CCFT columns database

(4) Interaction Equations

How do we get a simplified expression 
that is close to the design strength?

13
 5/29/2013

(5) Biaxial Interaction Surface
Analytical vs. Experimental Data

(6) Local Buckling

Theoretical difference of 
1.73 between two cases 
not reflected in code 
provisions

14
 5/29/2013

Project Objectives
• Obtain and evaluate experimental response: 
– Critical load (Pcr)
– P‐M interaction diagram (uniaxial and biaxial bending)
– Cyclic lateral force (uniaxial and biaxial bending)
– Torsion (torsional strength and rigidity)
– Wet concrete pressure due to the pouring
– Flexural rigidity (EIeff)
– Steel local buckling and concrete confinement
• Develop new computational formulations for 
complete frame analysis of composite systems
• Provide recommendations on construction, analysis, 
and design of CFTs.

NEES – UMN MAST Lab
MAST capabilities:
• 6 DOFs
• Pz = 1320 kip 
• Px, Py = 880 kips
• Ux=Uy=+/‐16” 
• 14’ < L < 28’ 
Databases gaps: 
• L = 18 ft. and 26 ft.
• ,  <  2.7
• D/t  86 (CCFT) 
• B/t  67 (RCFT)
• fc’ = 5 ksi and 12 ksi

15
 5/29/2013

CFT Test  Matrix (18 specimens)
Specimen L Steel section Fy fc’ D/t
name (ft) HSS D x t (ksi) (ksi)
1-C5-18-5 18 HSS5.563x0.134 42 5 45
2-C12-18-5 18 HSS12.75X0.25 42 5 55
CCFT
3-C20-18-5 18 HSS20x0.25 42 5 86 103
52 (S)
4-Rw-18-5 18 HSS20x12x0.25 46 5 67
5-Rs-18-5 18 HSS20x12x0.25 46 5 67
6-C12-18-12 18 HSS12.75X0.25 42 12 55
7-C20-18-12 18 HSS20x0.25 42 12 86 RCFT
8-Rw-18-12 18 HSS20x12x0.25 46 12 67 56
34 (S)
9-Rs-18-12 18 HSS20x12x0.25 46 12 67
Similar for specimens 10‐18 but at 26 ft. 

Setup and Instrumentation
• Strain Gages
Uniaxial and rosettes distributed 
along height
Measurements during concrete 
pouring and testing 
• LVDTs
Sets of three for biaxial curvature 
measurement
• String Pots
Distributed along height
• Krypton Coordinate 
Measurement Machine
• Video and Still Images
Four towers for images of whole 
specimen as well as base 

16
 5/29/2013

Hydrostatic Pressures on Slender RCFT

≈2’

FE Analysis:
max ≈  Stiffeners to reduce expansion in the 
max ≈ 36.1 ksi
RCFTs during the concrete pouring
max ≈ ¼ in

Surveyed Initial Imperfections
Length (ft) Length (ft)

18 14 10 15 12 13 17
25 25
11
16
20 20

15 15

10 10
 o=L/500=0.63
 o=L/500=0.63

5 5

0 0
0 0.5 1 1.5 2 0 0.5 1 1.5
Initial imperfection Initial imperfection
CCFTs, L=26ft RCFTs, L=26ft

17
 5/29/2013

Load protocol

0, PA
LC1
Stability Effects  ME, PE

Pcr 0, PA PA, PA

LC1 MB, PC

MD, PC/2
MB, 0

LC 1 – Axial load only

Load protocol
0, PA

LC2

P Stability Effects  ME, PE

0, PA PA, PA

Fmax LC1 MB, PC

MLC2a, 2PA
LC2a
unidirectional
MD, PC/2

LC2b MLC2b, PA
unidirectional

MB, 0

LC 2 – Axial load plus lateral displacement along X
at two different axial load levels

18
 5/29/2013

Load protocol
LC3

0, PA
P Stability Effects  ME, PE

0, PA PA, PA

Fmax y LC1 MB, PC

LC3a
bidirectional MLC2a, 2PA
x LC2a
unidirectional
LC3b
bidirectional

MD, PC/2
LC2b MLC2b, PA
unidirectional

LC3c
bidirectional

MB, 0

LC 3A – Axial load at three levels plus lateral displacement 
along both X and y in a diamond‐spike configuration

Load protocol
LC3
0, PA

Lateral Drift (%)

-6 -4 -2 0 2 4 6
30
ME, PE 1
P
Cracking of concrete
Steel yielding in compression
PA, PA 0.8
0, P
20A Steel yielding in tension

y Crushing of concrete 0.6
Fmax Steel local buckling LC1
10
MB, PC 0.4
Lateral Force (kip)

LC3a
bidirectional MLC2a, 2PA 0.2
x 0
LC2a
unidirectional 0

LC3b -0.2
-10 bidirectional
-0.4
MD, PC/2

LC2b MLC2b, PA
unidirectional -0.6
-20
-0.8
LC3c
bidirectional -1
-30
-10 -5 0 5 10
Lateral Displacement (in) MB, 0

LC 3B – Axial load at three levels plus lateral displacement 
along both X and y in a “figure eight” configuration

19
 5/29/2013

Load protocol
CCFT20x0.25‐18ft‐5ksi

0, PA
Lateral Drift (%)
LC4 600 -6 -4 -2 0 2 4 6
30
Cracking of concrete
P=0 ME, PE 1

PTcr 400
0, P
Steel yielding in compression
PA, PA 0.8
20A Steel yielding in tension

Torsional Moment (kip‐ft)
P=0.2Po
Crushing of concrete 0.6
T 200 Steel local buckling LC1
10
MB, PC 0.4

Lateral Force (kip)

LC3a
bidirectional MLC2a, 2PA 0.2
0 LC2a
0 unidirectional 0

 LC3b -0.2
-200 bidirectional
-10 -0.4

MD, PC/2
LC2b MLC2b, PA
unidirectional -0.6
-400
-20
-0.8
LC3c
bidirectional -1
-600
-30
-10 -10 -5 -5 0 0 5 510 10
Lateral Displacement (in)
Angle of twist (deg) MB, 0

LC 4 – Torsion at two levels of axial load

20
 5/29/2013

Load protocol: LC1 – Pure compression
P (kip) Specimen 17‐Rs‐26‐12
P 3000
Cross-section
Beam-column

Stability Effects
2500
Experimental

2000

1500

1000

500

0
0 200 400 600 800 1000
M M (kip‐ft)

Load protocol: LC2 – Uniaxial bending
Specimen 3‐C20‐18‐5

21
 5/29/2013

Load protocol: LC3 – Biaxial bending
CCFT Specimen
20x0.25
AISC Beam Column Strength (K=2)
All Load Cases
1500
Fy = 42 ksi Probe
Experimental Interaction Points

f’c = 5 ksi
L = 18 feet
KL = 36 feet 1000
Z Force (k)

500

0 500
-500 0
0 -500
500
Y Moment (k-ft)
X Moment (k-ft)

Corrected Column Strengths (LC1)

MAST capacity reached: 3, 5, 7, 9
Large imperfection: 1, 8, 11, 17

22
 5/29/2013

Local Buckling ‐ 2010

Composite Members Subject to Axial Compression

Description of Width- p r Max.
Element Thickness Compact/ Noncompact/ Permitted
Ratio Noncompact Slender
Sides of rectangular
box and hollow E E E
structural sections b/t 2.26 Fy
3.00 Fy
5.00 Fy
of uniform thickness
Round filled sections
D/t 0.15 E/Fy 0.19 E/Fy 0.35 E/Fy

23
 5/29/2013

Extraction of EI from the experimental M‐ curves

M (kip-ft) M (kip‐ft)
600
EI =21081046 kip-in2
eff
EI =21865004 kip-in2
expL
500
EI /EI =1.0372
expL eff
EI =21868261 kip-in2
expU
400
EI /EI =1.0373
expU eff

300

200

100

0
0 1 2 3 4 5
 (10-4/in)  (1/in)
Specimen 4-Rw-18-5 Specimen 13Rs‐26‐5, LC2

Load protocol: LC4 –Torsion
P 600
Specimen 3‐C20‐18‐5

T P=0
400
P=0.2Po
200
T (kip-ft)

-200

-400

-600
-10 -5 0 5 10
z (deg)

24
 5/29/2013

Load protocol: LC4 –Torsion
RCFTs, P=0 to 0.2Po 4‐Rw‐18‐5, P=0 kip
T (kip‐ft) T (kip‐ft) GJ eff  Gs J s  T Gc J c
500 500
GJ =17430909 kip-in2
400 exp
400
G J =11003678 kip-in2
s s
300 300
G J =31370173 kip-in2
200 c c
200
 T=0.2049
100 100

0 0
-100 -100

-200 -200
-300 -300

-400 -400

-500 -500
-10 -5 0 5 10 -10 -5 0 5 10

z (deg) z (deg)

Summary  of 
Experimental Results
A comprehensive and unique data for:
• Slender CCFTs and RCFTs
• Axial strength and  beam‐column 
strength for CFTs
• Complex cyclic loadings
• Initial imperfections
• Construction stresses/deformations
• Local buckling
• Ductility
Current AISC equations predict 
strength well for these specimens

25
 5/29/2013

Analysis of Composite Frames:

Mixed Beam‐Column Element
• Mixed beam finite element 
formulation was developed using 
both displacement and force shape 
functions
• Distributed plasticity fiber 
formulation:  stress and strain 
modeled explicitly at each fiber of  Shape Functions
1

Displacement
cross section

Transverse
• Perfect composite action assumed 
(i.e., slip neglected) 0
• Total‐Lagrangian corotational 0 L
formulation 1
• Implemented in the OpenSees 

Bending
Moment
framework
0
0 L

Constitutive Relations
• Constitutive formulations, calibration, and validation developed for five 
separate steel and steel‐concrete composite cross sections plus 
connections
– CCFT, RCFT, and SRC beam‐columns
– WF beams
– WF and Rect. HSS braces
– Moment frame and braced frame connections
• “Proposed for Behavior” constitutive model
– Aims to capture the behavior as accurately as possible
• “Proposed for Design” constitutive model
– Follows typical assumptions common in the development of design 
recommendations (e.g., no steel strain hardening, no concrete tension)

• Calibrated and validated against detailed results of over 100 
monotonically‐ and cyclically‐loaded experiments of composite beam‐
columns, connections, and frames

26
 5/29/2013

Uniaxial Cyclic Concrete Constitutive

Relations for CFTs and SRCs
• “Proposed for Behavior” constitutive relation:
– Based on the rule‐based model of Chang and Mander (1994)
– Backbone stress‐strain curve for the concrete is based on Tsai’s Equation, which is 
defined by:
• Initial stiffness Ec
• Peak coordinate (´cc, f´cc)
• r, which acts as a shape factor for Tsai’s equation and enables calibration for 
confinement in CFTs, between the flanges in SRCs, etc.
• “Proposed for Design” constitutive relation: simplified version of PB
1 1

0 0

-1 -1
Stress (ksi)

Stress (ksi)
-2 -2

-3 -3

-4 -4

-5 -5
-10000 -9000 -8000 -7000 -6000 -5000 -4000 -3000 -2000 -1000 0 1000 -10000 -9000 -8000 -7000 -6000 -5000 -4000 -3000 -2000 -1000 0 1000
Strain (strain) Strain (strain)

Uniaxial Cyclic Steel Constitutive

Relations for CFTs, SRCs, WFs, Rebar
• For the “Proposed for 
Behavior” model, based on the 
bounding‐surface plasticity 
model of Shen et al. (1995).
• Modifications for the analysis 
of composite members
– Local buckling
1.4
– Residual stress defined with  Corner Et3
Et2
initial plastic strain 1.2
Fy
Flat
• For the “Proposed for Design” 
Et1
Normalized Stress (/Fy,flat)

1 Fym
Et3
Et2
Fp
model, either elastic‐perfectly  0.8 Et1

plastic (SRC WFs; rebar) or  0.6
Elastic Unloading

based on the model of Abdel‐ 0.4 Es

Rahman & Sivakumaran 1997  0.2 Fp = 0.75 Fy

Et1 = Es/2
Et2 = Es/10
(CFTs) 0
Fym = 0.875 Fy Et3 = Es/200

0 2 4 6 8 10
Normalized Strain (/y,flat)

27
 5/29/2013

SRC Beam‐Column Validation

Ricles and Paboojian 1994
400 500

400
300

300
200
200

100
Lateral Load (kN)

Lateral Load (kN)

100

0 0

-100
-100

-200
-200
-300

-300
Expt. -400 Expt.
PfB PfB
-400 -500
-150 -100 -50 0 50 100 150 -150 -100 -50 0 50 100 150
Lateral Displacement (mm) Lateral Displacement (mm)
Test #4: 4 (Ricles and Paboojian 1994) Test #8: 8 (Ricles and Paboojian 1994)

H = 406 mm; B = 406 mm H = 406 mm; B = 406 mm

W8x40 W8x40
Fy = 372 MPa Fy = 372 MPa
4 #9; Fyr = 448 MPa 12 #7; Fyr = 434 MPa
f′c = 31 MPa f′c = 63 MPa
P/Pno = 0.19 P/Pno = 0.11 
L/H =  4.8 L/H =  4.8

RCFT Beam‐Column Validation

Varma 2000
500 500

400 400

300 300

200 200
Lateral Load (kN)

Lateral Load (kN)

100 100

0 0

-100 -100

-200 -200

-300 -300

-400 Expt. -400 Expt.

PfB PfB
-500 -500
-100 -80 -60 -40 -20 0 20 40 60 80 100 -80 -60 -40 -20 0 20 40 60 80
Lateral Displacement (mm) Lateral Displacement (mm)
Test #5: CBC-32-46-10 (Varma 2000) Test #8: CBC-48-46-20 (Varma 2000)

H/t = B/t = 35 H/t = B/t = 53 

Fy = 269 MPa Fy = 471 MPa
f′c = 110 MPa f′c = 110 MPa
P/Pno = 0.11 P/Pno = 0.18
L/H = 4.9 L/H = 4.9

28
 5/29/2013

CCFT Beam‐Column Validation

Specimen 11 – Load Case 3a

L = 7.9 m; D = 508 mm.; t = 5.9 mm.; D/t = 85.8; Fy = 305 MPa;  f′c = 55.9 MPa

Benchmark Frame Studies for

Composite Frames: Schematic
P P P
M
H
6 EIgross
ktop =
Gg,top L

EIgross
EIelastic EIelastic L = oe1g 
Pno,gross

6 EIgross
kbot = x
Gg,bot L

M EIgross = EsIs + EsIsr + EcIc

Pno,gross = AsFy + AsrFysr + Acf′c

Initial Imperfections:
Out-of-plumbness o = L/500
Out-of-straightness o = L/1000 (sinusoidal)

29
 5/29/2013

Selected Sections
CCFT RCFT
Index D t s Index H B t s
A 7 0.500 24.82% A 6 6 1/2 27.63%
B 10 0.500 17.70% B 9 9 1/2 19.06%
C 12.75 0.375 10.65% C 8 8 1/4 11.13%
D 16 0.250 5.72% D 9 9 1/8 5.05%
E 24 0.125 1.93% E 14 14 1/8 3.27%
Fy = 42 ksi; f′c = 4, 8, 16 ksi Fy = 46 ksi; f′c = 4, 8, 16 ksi
SRC
Index Steel Shape s
Index Rebar sr
A W14x311 11.66%
A 20 #11 3.98%
B W14x233 8.74%
B 12 #10 1.94%
C W12x120 4.49%
C 4 #8 0.40%
D W8x31 1.16%
Gross dimensions of all SRC sections = 28″ x 28″
Fy = 50 ksi; Fyr = 60 ksi; ; f′c = 4, 8, 16 ksi

Elastic Flexural Rigidity in

Composite Beam‐Columns
• EIeff – used to determine the axial compressive strength 
of columns in AISC 360‐10
• EIelastic – used in a 1st or 2nd order static, dynamic, or 
eigenvalue analysis
– in conjunction with Direct Analyses stiffness reductions to 
perform strength checks
– to compute story drifts used in interstory drift checks
– to compute fundamental periods and mode shapes 
(including for response spectrum analysis)
– as the elastic component of a concentrated plasticity 
beam‐column element
• EIDA – used in the Direct Analysis method
For Structural Steel: EIeff = EIelastic = EsIs

30
 5/29/2013

AISC 360‐10 Section I2: Calculation

of Axial Compressive Strength: EIeff
EI eff  Es I s  0.5Es I sr  C1 Ec I c (SRC)

 As 
C1  0.1  2    0.3
 Ac  As 

EI eff  Es I s  Es I sr  C3 Ec I c (CFT)

 As 
C3  0.6  2    0.9
 Ac  As 
/ 2

0/

Composite Axial Compressive Strength

from Benchmark Study

CCFT RCFT

SRC  SRC 
(strong axis) (weak axis)

31
 5/29/2013

Proposed Formula for Axial

Compressive Strength of SRCs
EI eff , proposed  Es I s  Es I sr  C1, proposed Ec I c (SRC)

2 As
C1, proposed  0.60   0.75
Ag

SRC  SRC 
(strong axis) (weak axis)

Axial Compressive Strength of SRC

Columns: Experimental Validation
EI eff , proposed  Es I s  Es I sr  C1, proposed Ec I c (SRC)

2 As
C1, proposed  0.60   0.75
Ag
1.5

Column Curve
Pexp/Pno,proposed

1 Anslijn & Janss 1974

Chen, Astaneh-Asl, & Moehle 1992
Han & Kim 1995
0.5 Han, Kim, & Kim 1992
Roderick & Loke 1975

0
0 0.5 1 1.5
oe,proposed

32
 5/29/2013

Benchmark Study Results:

Secant Values of EIelastic for Elastic Analysis
0.8 0.8
First-Order 1 “Serviceability” Level 1
Normalized Axial Compression (P/Pno)

Normalized Axial Compression (P/Pno)

0.7 0.7
Applied Load Strength/1.6
Interaction 0.9
0.6 0.6
EI elastic EI elastic 0.8
0.5 0.8 0.5
Es I s  Ec I c Es I s  Ec I c
0.4 0.7 0.4
0.6
0.3 0.6 0.3

0.2 0.5 0.2 0.4

0.1 0.1
0.4
0 0
0 0.5 1 1.5 0 0.5 1 1.5
Normalized Bending Moment (M/Mn) Normalized Bending Moment (M/Mn)
Section 4: RCFT-B-4, Frame 37: UA-67-g1 Section 13: RCFT-E-4, Frame 37: UA-67-g1
EIelastic value provides comparable deflection to fully nonlinear
analysis for forces shown

AISC 360‐10 Direct Analysis Method

Chapter C
Calculation of Required Strengths
• Analysis Requirements cPn,K=1
Second‐Order Elastic  Analysis Effective Length 
Factor Method Distributed 
• Consideration of Initial Imperfections cPn,K=K Plasticity 
Analysis
N i  0.002Yi Pr
• Adjustments to Stiffness
EI DA  0.8 b EI elastic Direct Analysis 
Method
EADA  0.8EAelastic Mr
Calculation of Available Strengths
• Chapters D though K without further 
consideration of overall structure stability
K 1

33
 5/29/2013

Direct Analysis
• From a practical standpoint it is best to 
maintain a stiffness reduction of 0.8b
EI DA  0.8 b EI elastic

 1.0 for Pr Pno  0.5

b  
 4  Pr Pno 1  Pr Pno  for Pr Pno  0.5
• Thus, differences between composite and 
steel may be embodied in proposed EIelastic:
EI elastic  Es I s  Es I sr  0.75C1 Ec I c (SRC)

EI elastic  Es I s  0.75C3 Ec I c (CFT)

Composite Interaction Strength

 = Pn/Pno  = Pn/Pno
Nominal Nominal
P P
(PA,0) Section (PA,0) Section
Strength Strength

(PA,0) (PA,0)

(PC,MC) (PC,MC)

(PC,MC)
(CPA,0.9BMB)
Nominal
Beam-Column Nominal
Strength Beam-Column
Strength
(0,MB) (0, BMB) (0,MB)

AISC 2010 M Proposed M

 1 for oe  1  PC PA for oe  0.5

 
 B  1  0.2  oe  1 for 1  oe  2  C   PC PA   PC PA  0.2  oe  0.5  for 0.5  oe  1.5
 for oe  2  for oe  1.5
 0.8  0.2

34
 5/29/2013

Variation of the Composite Interaction

Diagram with Slenderness

1.2
Normalized Axial Load (P/Pno)

0.8

0.6

0.4

0.2
0
1
0 2
0 0.5 1 1.5 3
Normalized Bending Moment (M/Mn)

CFT Bond Provisions in AISC 360‐10
For CCFT:
Rn = 0.25πD2CinFin
 = 0.45
For RCFT:  = 3.33
Rn = B2CinFin
where,
Rn = nominal bond strength, kips
Cin = 2 if the CFT extends to one side of the point of force transfer
= 4 if the CFT extends to both sides of the point of force transfer
Fin = nominal bond stress = 60 psi
B  = overall width of rectangular steel section along face transferring load, in.
D = outside diameter of the round steel section, in.

35
 5/29/2013

Experimental Setups for 
Assessing Bond Strength

Air Gap

Air Gap

(b) Push-out test (c) Push-out test (d) Typical CFT

(a) Push-off test
without shear tabs with shear tabs connection

Proposed Design Provisions
For CCFT: For RCFT:

Rn = πDLbondFin Rn = 2(B+H)LbondFin
Lbond = CinD Lbond = CinH
Fin = 30.9(t/D2) ≤ 0.2 Fin = 12.8(t/H2) ≤ 0.1
where, For RCFT: Both Lbond and Fin are based 
Rn = nominal bond strength, kips on the larger lateral dimension of the 
Fin = nominal bond stress, ksi tube (H ≥ B)
t = design wall thickness of steel section, in.
B = overall width of rectangular steel section (B ≤ H), in.
H = overall height of rectangular steel section (H ≥ B), in.  = 0.50,  = 3.00
D = outside diameter of round steel section, in.
Lbond = length of the bond region (the bond region of adjacent connections shall not overlap), in.
Cin = 4 if load is applied to the steel tube and the CFT extends to both sides of the point of force transfer
= 2 otherwise

36
 5/29/2013

Seismic Performance Factors:

FEMA P695 Archetype Frame Study:
Selection and Design of Archetype Frames

= Fully Restrained Connections

= Location of Braced Frame
= Shear Connections

Moment Frames Braced Frames

Selected Composite Archetype Frames

Moment Frames Braced Frames
Design  Design  Conc.
Bay 
Gravity  Seismic  Strength Index RCFT RCFT SRC RCFT‐Cd CCFT CCFT
Width
Load Load (f′c)
3 Stories 9 Stories 3 Stories 3 Stories 3 Stories 9 Stories
High 20’ Dmax 4 ksi 1      
High 20’ Dmax 12 ksi 2   
High 20’ Dmin 4 ksi 3      
High 20’ Dmin 12 ksi 4   
High 30’ Dmax 4 ksi 5    
High 30’ Dmax 12 ksi 6  
High 30’ Dmin 4 ksi 7    
High 30’ Dmin 12 ksi 8  
Low 20’ Dmax 4 ksi 9      
Low 20’ Dmax 12 ksi 10   
Low 20’ Dmin 4 ksi 11      
Low 20’ Dmin 12 ksi 12   
Low 30’ Dmax 4 ksi 13    
Low 30’ Dmax 12 ksi 14  
Low 30’ Dmin 4 ksi 15    
Low 30’ Dmin 12 ksi 16  

37
 5/29/2013

Typical Composite Connection Region Modeling:

Validated Against Tests
Nonlinear Moment
Nonlinear
Column Release
Column Zero Length Spring Element
Element Representing the
Nonlinear
Panel Zone Shear
Rigid Links Beam
Behavior
Element

Rigid
Nonlinear Links Nonlinear
Elastic
Beam Brace
Beam
Element Element
Element

Nonlinear stress‐resultant‐space multi‐surface  Modeling assumptions established 
kinematic hardening model used for rotational  by Hsiao et al. (2012)
spring formulation (after Muhummud 2003)

Evaluation of
Seismic Performance Factors
Archetype frames are categorized into performance 
groups based on basic structural characteristics
Design Design
Group  Period  Number of  Number of 
Gravity Load  Seismic Load 
Number Domain C‐SMFs C‐SCBFs
Level Level
PG‐1 High Dmax Short 6 4
PG‐2 High Dmax Long 2 2
PG‐3 High Dmin Short 6 4
PG‐4 High Dmin Long 2 2
PG‐5 Low Dmax Short 6 4
PG‐6 Low Dmax Long 2 2
PG‐7 Low Dmin Short 6 4
PG‐8 Low Dmin Long 2 2

38
 5/29/2013

Typical Static Pushover Analysis

1000

900 V max = 879.3 kips

800

700
V 80 = 703.4 kips
Base Shear (kips)

600

500

u = 50.8 in
400

300

200
V = 153.9 kips

100

0
0 10 20 30 40 50 60
Roof Displacement (in)
SFRS: C-SMF, Frame: RCFT-3-1

System Overstrength Factor, Ωo

• By the FEMA P695 methodology, Ωo 
should be taken as the largest  Average Ω
Group 
average value of Ω from any  Number C‐SMF C‐SCBF
performance group
– Rounded to nearest 0.5 PG‐1 5.9 2.1
– Upper limits of 1.5R and 3.0 PG‐2 5.3 1.9
• High overstrength for C‐SMFs PG‐3 7.6 2.8
– Displacement controlled design PG‐4 9.9 2.7
– Current value (Ωo = 3.0) is upper limit  PG‐5 6.2 1.8
and is acceptable PG‐6 5.5 1.7
• Overstrength for C‐SCBFs near  PG‐7 7.5 2.3
current value (Ωo = 2.0)
PG‐8 6.5 2.2
– Higher for PG‐3 and PG‐4 (High gravity 
load, SDC Dmin)

39
 5/29/2013

Typical Dynamic Time History Analyses:

Incremental Dynamic Analysis
18

16

14

12
ST = SMTSF2 (g)

10

6
S
ˆ  5.72
CT
g
4

  1.50
SMT  
 g
0
0% 5% 10% 15%
Maximum Story Drift
SFRS: C-SMF, Frame: RCFT-3-1

Response Modification Factor, R

• ACMR10% = Acceptable value of the Adjusted 
Collapse Margin Ratio for 10% collapse 
probability
Group  ACMR
• ACMR10% = 1.96 for both C‐SMF and C‐SCBF  Number
C‐SMF C‐SCBF
and are less than the ACMR shown for each 
PG‐1 4.8 3.3
performance group in the table
PG‐2 3.7 2.3
• Similarly positive results for ACMR20% per 
frame PG‐3 7.5 5.1

• ACMR values show correlation with the  PG‐4 8.5 5.4

overstrength PG‐5 4.9 2.6
• C‐SMFs PG‐6 3.9 2.9
– Current value (R = 8.0) is acceptable PG‐7 7.1 3.8

• C‐SCBFs PG‐8 6.9 3.7

– Current value (R = 5.0) is acceptable 

40
 5/29/2013

Deflection Amplification Factor, Cd

• By the FEMA P695 methodology, Cd = R for these 
systems
• Would represent a minor change for C‐SCBF 
– Current values: Cd = 4.5, R = 5.0
– Typically strength controlled design
• Would represent a significant change for C‐SMF
– Current values: Cd = 5.5, R = 8.0
– Typically already displacement controlled design
• Four C‐SMF archetype frames designed with the 
current Cd value 
– Lower overstrength with current Cd (average 4.9 vs. 6.4 
with Cd = R)
– Acceptable performance with current Cd

Key Conclusions from the Research
Experimental Research
• A comprehensive and unique data set for axial strength and beam‐column 
strength has been generated for slender CCFTs and RCFTs.
• CFTs demonstrated great toughness under complex cyclic loadings.
• Local buckling did not lead to substantial strength or stiffness losses.

Computational Research
• New mixed element analysis formulation developed for composite beam‐
columns 
• Composite beam‐columns exhibit robust performance under severe cyclic 
loading
• Analysis formulation enables benchmark studies of stability and strength 
of composite frames (non‐seismic and seismic)

41
 5/29/2013

Proposals for AISC 360‐16 (2016)
Specification for Structural Steel Buildings
• New commentary on addressing wet weight of concrete during concrete 
pour for CFTs
• New EIeff value for calculating column strength of SRCs to better reflect 
computational data
• New recommendations for EIelastic value to use for calculating elastic 
stiffness of CFTs and SRCs for use in elastic analysis and use in Direct 
Analysis
• New interaction equation that addresses possible unconservative errors 
for very slender composite members
• New CFT bond provisions that more accurately reflect the change in bond 
strength with CFT diameter and that clarify how to compute bond strength 
in load transfer regions
• Validation of current seismic performance factors in ASCE 7‐10 and 
recommendation to consider increasing the deflection criteria for C‐SMFs 
if Cd = R

Future Work

• Finalize recommendations for AISC 360‐16  
• Prequalified composite connections
• Incorporate creep and shrinkage effects into design of 
composite systems
• Effects of elevated temperature in composite systems, and 
effects of internal reinforcement
• Innovative composite framing systems:  
– Prefabricated composite construction systems
– Integration of new materials, including higher strength 
materials
– Etc.

42
 5/29/2013

Thank You
NEES Project Warehouse: https://nees.org/warehouse/project/440
440 – System Behavior Factors for Composite and Mixed Structural System
Roberto T. Leon, Jerome F. Hajjar, Nakin Suksawang

References and a list of papers and publications for this work are available at the NEES 
site for this webinar:  https://nees.org/events/details/190

In‐Kind:

The work described here is part of a NEESR project supported by the National 
Science Foundation under Grant No. CMMI‐0619047, the American Institute of Steel 
Construction, the Georgia Institute of Technology, and the University of Illinois at 
Urbana‐Champaign.  These experiments were conducted at the Multi‐axial 
Subassemablage Testing System (MAST) at the University of Minnesota.

43