PCI BRIDGE DESIGN MANUAL CHAPTER 8

TABLE OF CONTENTS
DESIGN THEORY AND PROCEDURE

NOTATION

8.0 AASHTO SPECIFICATION REFERENCES

8.1 PRINCIPLES AND ADVANTAGES OF PRESTRESSING

8.1.1 History
8.1.2 High Strength Steel
8.1.3 Prestressing Versus Conventional Reinforcing
8.1.4 Concrete to Steel Bond

8.2 FLEXURE
8.2.1 Allowable Stress Design (ASD)
8.2.1.1 Theory
8.2.1.1.1 Stage 1 Loading
8.2.1.1.2 Stage 2 Loading
8.2.1.1.3 Stage 3 Loading
8.2.1.1.4 Stage 4 Loading
8.2.1.1.5 Stage 5 Loading
8.2.1.1.5.1 Tensile Stresses - Normal Strength Concrete
8.2.1.1.5.2 Tensile Stresses - High Strength Concrete
8.2.1.1.5.3 Tensile Stresses -
8.2.1.2 Allowable Concrete Stresses
8.2.1.2.1
8.2.1.2.2
8.2.1.3 Design Procedure
8.2.1.4 Composite Section Properties
8.2.1.4.1 Theory
8.2.1.4.2 Procedure
8.2.1.5 Harped Strand Considerations
8.2.1.6 Debonded Strand Considerations
8.2.1.7 Minimum Strand Cover and Spacing
8.2.1.8 Design Example
8.2.1.8.1 Design Requirement 1
8.2.1.8.2 Design Requirement 2
8.2.1.8.3 Design Requirement 3
8.2.1.8.3.1 Strand Debonding
8.2.1.8.3.2 Harped Strands
8.2.1.8.3.3 Other Methods to Control Stresses
8.2.1.8.4 Design Requirement 4
8.2.1.9 Fatigue

JUL 03

7017 Bridge Man Ch 8 TOC 1 8/14/03, 10:58 AM

PCI BRIDGE DESIGN MANUAL CHAPTER 8
TABLE OF CONTENTS
DESIGN THEORY AND PROCEDURE

8.2.2 Flexural Strength Design

8.2.2.1 Theory
8.2.2.2
8.2.2.2.1 Ultimate Moment Capacity
8.2.2.2.1.1 Required Parameters

JUL 03
 PCI BRIDGE DESIGN MANUAL CHAPTER 8
TABLE OF CONTENTS
DESIGN THEORY AND PROCEDURE

8.3.2 Strand Development Length

8.3.2.1 Impact on Design
8.3.2.2 s
8.3.2.3
8.3.2.4 Factors Affecting Development Length
8.3.2.5 Bond Studies
8.3.2.6 Recommendations

8.4 SHEAR
8.4.1
8.4.1.1 Flexure-Shear Strength, Vci
8.4.1.2 Web-Shear Strength, Vcw
8.4.1.3 Web Reinforcement Contribution, Vs
8.4.1.3.1 Minimum Spacing Requirements
8.4.1.3.2 Minimum Shear Reinforcement
8.4.1.4 Application of to Continuous Spans
8.4.2 1979 Interim Revisions
8.4.3
8.4.3.1 Shear Design Provisions
8.4.3.1.1 Nominal Shear Resistance
c
8.4.3.1.3 Web Reinforcement Contribution, Vs
8.4.3.1.4 Values of β and θ
8.4.3.2 Design Procedure
8.4.3.3 Longitudinal Reinforcement Requirement
8.4.4 Comparison of Shear Design Methods

8.5 HORIZONTAL INTERFACE SHEAR

8.5.1 Theory
8.5.2
8.5.3
8.5.4 Comparison of Design Specifications

8.6 LOSS OF PRESTRESS

8.6.1 Introduction
8.6.2 Definition
8.6.3 Significance of Losses on Design
8.6.4 Effects of Estimation of Losses
8.6.4.1 Effects at Transfer
8.6.4.2 Effect on Production Costs
8.6.4.3 Effect on Camber
8.6.4.4 Effect of Underestimating Losses
JUL 03

7017 Bridge Man Ch 8 TOC 3 8/14/03, 10:58 AM

8.4312Concretibu,V
PCI BRIDGE DESIGN MANUAL CHAPTER 8
TABLE OF CONTENTS
DESIGN THEORY AND PROCEDURE

8.6.5 Prediction of Creep, Shrinkage and Relaxation Material Properties

8.6.5.1 Prediction of Creep Coefficient of Concrete
8.6.5.1.1 Creep Modification Factors

JUL 03
 PCI BRIDGE DESIGN MANUAL CHAPTER 8
TABLE OF CONTENTS
DESIGN THEORY AND PROCEDURE

8.8.2 Design of Bridge Decks Using Precast Panels

8.8.2.1 Determining Prestress Force
8.8.2.2 Service Load Stresses and Flexural Strength
8.8.2.3
8.8.2.3.1 Minimum Thickness
8.8.2.3.2 Live Load
8.8.2.3.3 Reinforcement Requirements
8.8.2.3.4 Shear Design
8.8.2.3.5 Crack Control
8.8.2.4
8.8.2.4.1 Refined Analysis
8.8.2.4.2 Strip Method
8.8.2.4.2.1 Minimum Thickness
8.8.2.4.2.2 Minimum Concrete Cover
8.8.2.4.2.3 Live Load
8.8.2.4.2.4 Location of Critical Sections
8.8.2.4.2.5 Design Criteria
8.8.2.4.2.6 Reinforcement Requirements
8.8.2.4.2.7 Shear Design
8.8.2.4.2.8 Crack Control
8.8.3 Other Precast Bridge Deck Systems
8.8.3.1 Continuous Precast Concrete SIP Panel System, NUDECK
8.8.3.1.1 Description of NUDECK
8.8.3.2 Full-Depth Precast Concrete Panels
8.8.4 Empirical Design Method

8.9 TRANSVERSE DESIGN OF ADJACENT BOX BEAM BRIDGES

8.9.1 Background
8.9.1.1 Current Practice
8.9.1.2 Ontario Bridge Design Code Procedure
8.9.2 Empirical Design
8.9.2.1 Tie System
8.9.2.2 Production
8.9.2.3 Installation
8.9.3 Suggested Design Procedure
8.9.3.1 Transverse Diaphragms
8.9.3.2 Longitudinal Joints Between Beams
8.9.3.3 Tendons
8.9.3.4 Modeling and Loads for Analysis
8.9.3.5 Post-Tensioning Design Chart
8.9.3.6 Design Method
8.9.3.7 Design Example
JUL 03

7017 Bridge Man Ch 8 TOC 5 8/14/03, 10:58 AM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
TABLE OF CONTENTS
DESIGN THEORY AND PROCEDURE

8.9.4 Lateral Post-Tensioning Detailing for Skewed Bridges

8.10 LATERAL STABILITY OF SLENDER MEMBERS

8.10.1 Introduction
8.10.1.1 Hanging Beams
8.10.1.2 Beams Supported from Beneath
8.10.2 Suggested Factors of Safety
8.10.2.1 Conditions Affecting FSc
8.10.2.2 Effects of Creep and Impact
8.10.2.3 Effects of Overhangs
8.10.2.4 Increasing the Factor of Safety
8.10.3 Measuring Roll Stiffness of Vehicles
8.10.4 Bearing Pads
8.10.5 Wind Loads
8.10.6 Temporary King-Post Bracing
8.10.7 Lateral Stability Examples
8.10.7.1 Hanging Beam Example
8.10.7.2 Supported Beam Example

8.11 BENDING MOMENTS AND SHEAR FORCES DUE TO

VEHICULAR LIVE LOADS
8.11.1 HS20 Truck Loading
8.11.2 Lane Loading, 0.640 kip/ft
8.11.3 Fatigue Truck Loading

8.12 STRUT-AND-TIE MODELING OF DISTURBED REGIONS

8.12.1 Introduction
8.12.2 Strut-and-Tie Models
8.12.2.1 Truss Geometry Layout
8.12.2.2 Nodal Zone and Member Dimensions
8.12.2.3 Strengths of Members
8.12.3 Provisions for Strut-and-Tie Models
8.12.3.1 Compression Struts
8.12.3.1.1 Unreinforced Concrete Struts
8.12.3.1.2 Reinforced Concrete Struts
8.12.3.2 Tension Ties
8.12.3.2.1 Tie Anchorage
8.12.3.3 Proportioning Node Regions
8.12.3.4 Crack Control Reinforcement
8.12.4 Steps for Developing Strut-and-Tie Models
8.12.4.1 Design Criteria
8.12.4.2 Summary of Steps
JUL 03

7017 Bridge Man Ch 8 TOC 6 8/14/03, 10:58 AM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
TABLE OF CONTENTS
DESIGN THEORY AND PROCEDURE

8.12.5 Pier Cap Example

8.12.5.1 Flow of Forces and Truss Geometry
8.12.5.2 Forces in Assumed Truss
8.12.5.3 Bearing Stresses
8.12.5.4 Reinforcement for Tension Tie DE
8.12.5.5 Strut Capacities
8.12.5.6 Nodal Zone at Pier
8.12.5.7 Minimum Reinforcement for Crack Control

8.13 DETAILED METHODS OF TIME-DEPENDENT ANALYSIS

8.13.1 Introduction
8.13.1.1 Properties of Concrete
8.13.1.1.1 Stress-Strain-Time Relationship
8.13.1.2 Effective Modulus
8.13.1.3 Age-Adjusted Effective Modulus
8.13.1.4 Properties of Prestressing Steel
8.13.1.5 Reduced Relaxation under Variable Strain
8.13.2 Analysis of Composite Cross-Sections
8.13.2.1 Initial Strains
8.13.2.2 Methods for Time-Dependent Cross-Section Analysis
8.13.2.2.1 Steps for Analysis
8.13.2.2.2 Example Calculations
8.13.3 Analysis of Composite Simple-Span Members
8.13.3.1 Relaxation of Strands Prior to Transfer
8.13.3.2 Transfer of Prestress Force
8.13.3.2.1 Example Calculation (at Transfer)
8.13.3.3 Creep, Shrinkage and Relaxation after Transfer
8.13.3.3.1 Example Calculation (after Transfer)
8.13.3.4 Placement of Cast-in-Place Deck
8.13.3.5 Creep, Shrinkage and Relaxation
8.13.3.6 Application of Superimposed Dead Load
8.13.3.7 Long-Term Behavior
8.13.4 Continuous Bridges
8.13.4.1 Effectiveness of Continuity
8.13.4.2 Applying Time-Dependent Effects
8.13.4.3 Methods of Analysis
8.13.4.3.1 General Method
8.13.4.3.2 Approximate Method
8.13.4.3.2.1 Restraint Moment Due to Creep
8.13.4.3.2.2 Restraint Moment Due to Differential Shrinkage

8.14 REFERENCES
JUL 03

7017 Bridge Man Ch 8 TOC 7 8/14/03, 10:58 AM

7017 Bridge Man Ch 8 TOC 8 8/14/03, 10:58 AM
 PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

A = area of cross-section of the precast beam [STD], [LRFD]

A = distance to pickup points from each end of the beam —
Ac = area of concrete on the flexural tension side of the member [LRFD]
Ac = area of beam cross-section —
Acv = area of concrete section resisting shear transfer [LRFD]
Acs = cross-sectional area of a concrete strut [LRFD]
Ag = gross area of section [LRFD]
Ak = area of cross-section of element k —
Ao = area enclosed by centerlines of the elements of the beam [LRFD]
Aps = area of pretensioning steel [LRFD]
As = area of non-pretensioning tension reinforcement [STD], [LRFD]
As = total area of vertical reinforcement located within a distance
(h/5) from the end of the beam [LRFD]
Asf = area of steel required to develop the ultimate compressive
strength of the overhanging portions of the flange [STD]
Asr = area of steel required to develop the compressive strength of the
web of a flanged section [STD]
Ass = area of reinforcement in strut [LRFD]
Ast = area of longitudinal mild steel reinforcement in tie [LRFD]
A*s = area of pretensioning steel [STD]
A´s = area of compression reinforcement [LRFD]
Av = area of web reinforcement [STD]
Av = area of transverse reinforcement within a distance s [LRFD]
Avf = area of shear-friction reinforcement [LRFD]
Avh = area of web reinforcement required for horizontal shear —
Av-min = minimum area of web reinforcement —
a = depth of the compression block [STD]
a = depth of the equivalent rectangular stress block [LRFD]
a = length of overhang —
b = effective flange width —
b = width of beam [STD]
b = width of top flange of beam —
b = width of the compression face of a member [LRFD]
b´ = width of web of a flanged member [STD]
bb = width of bottom flange of beam —
bv = width of cross-section at the contact surface being investigated
for horizontal shear [STD]
bv = effective web width [LRFD]
bv = width of interface [LRFD]
bw = web width [LRFD]
Ca = creep coefficient for deflection at time of erection due to loads
applied at release —

JUL 03

7017 Bridge Man Ch 8 TOC 9 8/14/03, 10:58 AM

PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

CRc = loss of pretension due to creep of concrete [STD]

CRs = loss of pretension due to relaxation of pretensioning steel [STD]
C(t,t0) — = creep coefficient of the concrete member at a certain age —
C(t,tj) = creep coefficient at time tj (j =

JUL 03
PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

E*c = age-adjusted, effective modulus of elasticity of concrete for a

gradually applied load at the time of transfer of prestressing —
E*cb = age-adjusted, effective modulus of elasticity of the beam —
E*cd = age-adjusted, effective modulus of elasticity of the deck —
E*c(t,t0) = effective modulus of elasticity at certain time —
E*ck = age-adjusted, effective modulus of element k —
e (

JUL 03
 PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

fpc = compressive stress in concrete after all prestress losses have occurred
either at the centroid of the cross-section resisting live load or at
the junction of the web and flange when the centroid lies in the
flange. In a composite section, fpc is the resultant compressive
stress at the centroid of the composite section, or at the junction
of the web and flange when the centroid lies within the flange,
due to both prestress and to the bending moments resisted by the
precast member acting alone. [LRFD]
fpe = compressive stress in concrete due to effective pretension
forces only (after allowance for all pretension losses) at
extreme fiber of section where tensile stress is caused by
externally applied loads [STD]
fpe = effective stress in the pretensioning steel after losses [LRFD]
fpi = initial stress immediately before transfer —
fpj = stress in the pretensioning steel at jacking [LRFD]
fpo = stress in the pretensioning steel when the stress in the
surrounding concrete is zero [LRFD]
fps = average stress in pretensioning steel at the time for which the
nominal resistance of member is required [LRFD]
fpu = specified tensile strength of pretensioning steel [LRFD]
fpy = yield strength of pretensioning steel [LRFD]
fr = modulus of rupture of concrete [STD], [LRFD]
fs = allowable stress in steel under service loads —
f ´s = ultimate stress of pretensioning reinforcement [STD]
fse = effective final pretension stress —
fsi = effective initial pretension stress —
f *su = average stress in pretensioning steel at ultimate load [STD]
f(tj) = stress at time tj —
fr(t,t0) = relaxation stress at a certain time —
f(t0) = tensile stress at the beginning of the interval —
fy = yield strength of reinforcing bars [STD]
fy = specified minimum yield strength of reinforcing bars [LRFD]
fy = yield stress of pretensioning reinforcement [STD]
f ´y = specified minimum yield strength of compression reinforcement [LRFD]
fyh = specified yield strength of transverse reinforcement [LRFD]
H = average annual ambient mean relative humidity [LRFD]
h = length of a single segment —
h = overall depth of precast beam [STD]
h = overall depth of a member [LRFD]
hcg = height of center of gravity of beam above road —
hd = deck thickness —
hf = compression flange depth [LRFD]
hr = height of roll center above road —
I = moment of inertia about the centroid of the non-composite
precast beam, major axis moment of inertia of beam [STD], [LRFD]
JUL 03

7017 Bridge Man Ch 8 TOC 12 8/14/03, 10:58 AM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

I = impact fraction [STD]

Ik = moment of inertia of element k —
IM = dynamic load allowance [LRFD]
Ieff = effective cracked section lateral (minor axis) moment of inertia —
Ig = gross lateral (minor axis) moment of inertia —
K = factor used for calculating time-dependent losses —
Kr = factor used for calculating relaxation loss in strand that occurs prior
to transfer —
Kθ = sum of rotational spring constants of supports —
k = factor used in calculation of average stress in pretensioning steel
for strength limit state; factor related to type of strand
[LRFD]
kc = product of applicable correction factors for creep = kla kh ks —
kcp = correction factor for curing period —
kla = correction factor for loading age —
kh = correction factor for relative humidity —
ks = correction factor for size of member —
ksh = product of applicable correction factors for shrinkage = kcp kh ks —
kst = correction factor for concrete strength —
L = live load [STD]
L = length in feet of the span under consideration for positive
moment and the average of two adjacent loaded spans for
negative moment [STD]
L = overall beam length or design span —
L = span length measured parallel to longitudinal beams [STD]
L = span length [LRFD]
LL = vehicular live load [LRFD]
Lr = intrinsic relaxation of the strand —
Lx = distance from end of prestressing strand to center of the panel [STD]
l = overall length of beam —
ld = development length —
lt = transfer length —
Mc = moment in concrete beam section —
Mcr = cracking moment [LRFD]
Mcr(t) = restraint moment due to creep at time t —
M*cr = cracking moment [STD]
Md/nc = moment due to non-composite dead loads [STD]
Mel = fictious elastic restraint moment at the supports —
Mg = unfactored bending moment due to beam self-weight —
Mg = self-weight bending moment of beam at harp point —
Mgmsp = self-weight bending moment at midspan —
Mk = element moment —

JUL 03

7017 Bridge Man Ch 8 TOC 13 8/14/03, 10:58 AM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

Mlat = lateral bending moment at cracking —

MLL = unfactored bending moment due to lane load per beam —
Mmax = maximum factored moment at section due to externally applied loads [STD]
Mn = nominal moment strength of a section [STD]
Mn = nominal flexural resistance [LRFD]
Mn/dc = non-composite dead load moment at the section —
Mr = factored flexural resistance of a section in bending [LRFD]
Msh = shrinkage moment —
Msr(t) = restraint moment due to differential shrinkage at time t —
Msw = moment at section of interest due to self-weight of the member
plus any permanent loads acting on the member at time of release —
Mu = factored bending moment at section [STD], [LRFD]
Mx = bending moment at a distance x from the support —
M0 = theoretical total moment in sections —
M0k = theoretical moment in section of element k —
m = stress ratio —
N = number of segments between nodes (must be even number) —
Nk = element normal force —
Nc = internal element force in concrete —
Ns = internal element force in steel —
Nu = applied factored axial force taken as positive if tensile [LRFD]
N0k = theoretical normal force in section of element k, positive when tensile —
N0 = theoretical total normal force in sections —
n = modular ratio between slab and beam materials [STD], [LRFD]
nk = modular ratio of element k —
ns = modular ratio of steel element —
PPR = partial prestress ratio [LRFD]
Pc = permanent net compression force [LRFD]
Pn = nominal axial resistance of strut or tie [LRFD]
Pr = factored axial resistance of strut or tie [LRFD]
Pse = effective pretension force after allowing for all losses —
Psi = effective pretension force after allowing for the initial losses —
Q = first moment of inertia of the area above the fiber being considered —
R = radius of curvature —
RH = relative humidity [STD]
Rn = strength design factor —
Ru = flexural resistance factor —
r = radius of gyration of the gross cross-section —
r = radius of stability
S = width of precast beam [STD]
S = spacing of beams [STD], [LRFD]

JUL 03

7017 Bridge Man Ch 8 TOC 14 8/14/03, 10:58 AM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

S = slab span [LRFD]

S = span between the inside faces of the beam webs [LRFD]
Sb = section modulus for the extreme bottom fiber of the
non-composite precast beam —
Sbc = composite section modulus for extreme bottom fiber of the
precast beam —
SH = loss of pretension due to concrete shrinkage [STD]
SN = the value of the integral —
S(t,t0) = shrinkage coefficient at a certain age —
St = section modulus for the extreme top fiber of the non-composite
precast beam —
Su = ultimate free shrinkage strain in the concrete adjusted for
member size and relative humidity —
s = longitudinal spacing of the web reinforcement [STD]
s = length of a side element [LRFD]
s = spacing of rows of ties [LRFD]
t = time, days; age of concrete at the time of determination of creep
effects, days; age of concrete at time of determination of shrinkage
effects, days; time after loading, days —
t = thickness of web —
t = thickness of an element of the beam —
tf = thickness of flange —
t0 = age of concrete when curing ends; age of concrete when load is
initially applied, days —
ts = cast-in-place concrete slab thickness —
ts = depth of concrete slab [LRFD]
Vc = nominal shear strength provided by concrete [STD]
Vc = nominal shear resistance provided by tensile stresses in the
concrete [LRFD]
Vci = nominal shear strength provided by concrete when diagonal
cracking results from combined shear and moment [STD]
Vcw = nominal shear strength provided by concrete when diagonal
cracking results from excessive principal tensile stress in web [STD]
Vd = shear force at section due to unfactored dead load [STD]
Vi = factored shear force at section due to externally applied loads
occurring simultaneously with Mmax [STD]
Vn = nominal shear resistance of the section considered [LRFD]
Vnh = nominal horizontal shear strength [STD]
Vp = vertical component of effective pretension force at section [STD]
Vp = component of the effective pretensioning force, in the
direction of the applied shear, positive if resisting the applied shear [LRFD]
Vs = nominal shear strength provided by web reinforcement [STD]
Vs = shear resistance provided by shear reinforcement [LRFD]
Vu = factored shear force at the section [STD], [LRFD]

JUL 03

7017 Bridge Man Ch 8 TOC 15 8/14/03, 10:58 AM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

Vuh = factored horizontal shear force per unit length of the beam [LRFD]
vu = average factored shear stress [LRFD]
W = total weight of beam —
w = a uniformly distributed load [LRFD]
w = width of clear roadway [LRFD]
w = weight per unit length of beam —
wc = unit weight of concrete [STD], [LRFD]
x = distance from the support to the section under question —
y = height of center of gravity of beam above roll axis
(beam supported from below) —
yb = distance from centroid to the extreme bottom fiber of the
non-composite beam —
ybc = distance from centroid to the bottom of beam of the composite section —
ybs = distance from the center of gravity of strands to the bottom
fiber of the beam —
yk = distance of the centroid of element k from edge —
yr = height of roll axis above center of gravity of beam (hanging beam) —
ys = height above soffit of centroid of prestressing force —
yt = distance from centroid to the extreme top fiber of the
non-composite beam —
ytc = distance from centroid to the top of deck of the composite section —
z = lateral deflection of center of gravity of beam —
zmax = distance from centerline of vehicle to center of dual tires —
z o = theoretical lateral deflection of center of gravity of beam with the
full dead weight applied laterally —
z ´o = theoretical lateral deflection of center of gravity of beam with the
full dead weight applied laterally, computed using Ieff for tilt angle
θ under consideration —
α = super-elevation angle or tilt angle of support in radians —
α = factor used in calculating elastic shortening loss —
α = coefficient defined by (Eq. 8.6.2.5.1-3) to account for interaction
between steel and concrete in pretensioning loss calculations —
αs = angle between compressive strut and adjoining tension tie [LRFD]
β = factor indicating ability of diagonally cracked concrete to
transmit tension (a value indicating concrete contribution) [LRFD]
β1 = factor for concrete strength [STD]
β1 = ratio of the depth of the equivalent uniformly stressed compression
zone assumed in the strength limit state to the depth of the actual
compression zone [LRFD]
δc = time-dependent multiplier —
∆ = deflection —
∆ = camber measured with respect to the beam-ends —

JUL 03

7017 Bridge Man Ch 8 TOC 16 8/14/03, 10:58 AM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

∆fcdp = change in concrete stress at center of gravity of pretensioning

steel due to dead loads except the dead load acting at the time
the pretensioning force is applied [LRFD]
∆fpCR = loss in pretensioning steel stress due to creep [LRFD]
∆fpES = loss in pretensioning steel stress due to elastic shortening [LRFD]
∆fpR = loss in pretensioning steel stress due to relaxation of steel [LRFD]
∆fpR1 = loss in pretensioning steel stress due to relaxation of steel at
transfer [LRFD]
∆fpR2 = loss in pretensioning steel stress due to relaxation of steel after
transfer [LRFD]
∆fpSR = loss in pretensioning steel stress due to shrinkage [LRFD]
∆fpT = total loss in pretensioning steel stress [LRFD]
∆fs = total loss of prestress —
ε = strain —
εc = strain in concrete beam —
εcr = the time dependent creep strain —
εf = the immediate strain due to the applied stress f —
εfc = elastic strain in concrete —
εfk = element strain —
εfs = elastic strain in steel —
εk = strain in element k —
εp = strain in prestressing steel —
εs = strain in mild steel —
εs = tensile strain in cracked concrete in direction of tensile tie [LRFD]
εsh = free shrinkage strain —
εshb(t,t2) = shrinkage strain of the beam from time t2 to time t —
εshb(t3,t2) = shrinkage strain of the beam from time t2 to time t3 —
εshd(t,t3) = shrinkage strain of the deck from time t3 to time t —
εshu = ultimate free shrinkage strain in the concrete, adjusted for member
size and relative humidity —
εsi = strain in tendons corresponding to initial effective pretension
stress —
εx = longitudinal strain in the web reinforcement on the flexural
tension side of the member [LRFD]
ε0c = initial strain in concrete —
ε1 = principal tensile strain in cracked concrete due to factored loads [LRFD]
γ*
= factor for type of pretensioning reinforcement [STD]
φ = strength reduction factor [STD]
φ = resistance factor [LRFD]
φ = curvature —
φc = curvature at midspan —
φcr = curvature due to creep —
φfk = element curvature —

JUL 03

7017 Bridge Man Ch 8 TOC 17 8/14/03, 10:58 AM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
NOTATION
DESIGN THEORY AND PROCEDURE

φk = curvature of element k —
φ0 = curvature at support —
λ = parameter used to determine friction coefficient µ [LRFD]
µ = Poisson’s ratio for beams [STD]
µ = coefficient of friction [LRFD]
θ = angle of inclination of diagonal compressive stresses [LRFD]
θ = roll angle of major axis of beam with respect to vertical —
θL = left end rotation of beam due to simple span loads —
θR = right end rotation of beam due to simple span loads —
θi = initial roll angle of a rigid beam —
θmax = tilt angle at which cracking begins, based on tension at the top corner
equal to the modulus of rupture —
θ´max = tilt angle at maximum factor of safety against failure —
ρb = reinforcement ratio producing balanced strain condition [STD]
ρ* = ratio of pretensioning reinforcement [STD]
ψ = a factor that reflects the fact that the actual relaxation is less than
the intrinsic relaxation —
χ = aging coefficient —
χ(t,t0) = aging coefficient at certain time —

JUL 03

7017 Bridge Man Ch 8 TOC 18 8/14/03, 10:58 AM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13 Detailed Methods of Time-Dependent Analysis/
8.13.1.1.1 Stress-Strain-Time Relationship

8.13 Section 8.6 in this chapter presents a variety of practical and relatively simple meth-
DETAILED METHODS ods to estimate time-dependent effects in prestressed concrete members. Those meth-
OF TIME-DEPENDENT ods are suitable for a very wide range of bridge projects but may not be applicable to
ANALYSIS certain special situations. More detailed methods are available when the designer feels
that a more rigorous estimate of time-dependent effects is warranted.

8.13.1 The following sections describe a method that can be used to perform time-depen-
Introduction dent analysis of a composite prestressed concrete bridge member of any cross-section.
This method is based on traditional composite section analysis, using transformed
elastic properties of steel elements and any cast-in-place concrete elements. Adjust-
ments are made to the elastic modulus of the concrete elements to reflect creep
characteristics. So-called initial strains are introduced in the analysis to account for
concrete shrinkage, steel relaxation and residual concrete creep. By analyzing discrete
cross-sections, and then performing the numerical integration described in Section
8.7.4, whole members may also be analyzed using the methods that follow.

8.13.1.1 The mechanical properties of concrete vary with time. As hydration progresses, com-
Properties of Concrete pressive strength and modulus of elasticity continually increase, but at a decreasing rate.
In addition, it has long been recognized that concrete exhibits creep, defined as the
time-dependent increase in strain that occurs while the material is subjected to constant
stress. Finally, concrete undergoes shrinkage caused by drying. Chapter 2 and Section
8.6.2.3 provide more detailed discussion of these time-dependent behaviors.

There exists a wide range of methods used to produce precast concrete bridge
components. Concrete mixes, aggregates, admixtures, and curing methods all have
significant effects on the time-dependent properties of concrete as a structural mate-
rial. Because of these variations, the recommendations in Chapter 2 and Section 8.6
should be used only as a starting point. For applications where it is critical to accu-
rately predict time-dependent behavior, the properties of the actual materials used
should be determined by testing.

8.13.1.1.1 In order to perform time-dependent analysis, it is necessary to establish the stress-

Stress-Strain-Time strain-time relationship for the concrete material. This relationship will predict the
Relationship total strain, ε, at a future time, t, that results from a stress increment applied at time,
to. The total concrete strain at any time, t, can be separated into three components:
• εf = the immediate strain due to the applied stress, f
• εcr = the time-dependent creep strain
• εsh = free shrinkage strain

It is important to recognize that both the modulus of elasticity, E, and the creep coef-
ficient, C, are functions of time. In addition, because concrete is an aging material,
C depends on the loading age, to, as well.

JUL 03

7017 Bridge Manual Ch. 8.13 1 8/14/03, 1:04 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.1.1.1 Stress-Strain-Time Relationship

a) Constant Stress
Total concrete strain is (εf + εcr + εsh) which is usually expressed as follows:
f (t 0 )
ε = [
E c (t 0 )
]
1 + C(t,t 0 ) + ε sh (Eq. 8.13.1.1.1-1)

where
E(t0) = modulus of elasticity at time, to, the beginning of the interval
C(t, t0) = creep coefficient over a time interval from t0 to t
Eq. (8.13.1.1.1-1) applies as long as stress, f, is a constant, sustained stress. Figure
8.13.1.1.1-1 shows the gradual development of creep strains with time under the
effects of a constant stress.

Figure 8.13.1.1.1-1
Concrete Strain vs. Time
Under Constant Stress ������

������ �

��
����

������ ������
�������������������
�� �
�������
������
�����������
�
�������

�� � ����

b) Variable Stress
Where the applied stress, f, is variable, Eq. (8.13.1.1.1-1) cannot be used directly.
Figure 8.13.1.1.1-2 depicts the development of creep strains under the effects of an
increasing applied stress.

At most stress levels experienced due to service loads, the principle of superposition
applies. Using superposition, the effects of a series of applied stress increments can
be determined individually, using the above equation, and then combined to give the
total time-dependent concrete strain. For a series of stress increments, fj, applied at
times, tj, the total concrete strain can be expressed as:

JUL 03

7017 Bridge Manual Ch. 8.13 2 8/14/03, 1:04 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.1.1.1 Stress-Strain-Time Relationship/8.13.1.2 Effective Modulus

Figure 8.13.1.1.1-2
Concrete Strain vs. Time ������
Under Variable Stress

����

�� � ����

������ ����
���������������������������������
�� � �
� ������
��

���� ����
�
� �����

�� � ����
f(t j )
ε = ∑
E c (t j )
[1 + C(t,t )] + ε
j sh (Eq. 8.13.1.1.1-2)

Therefore, a method for predicting concrete strain, ε, under conditions where stress is
not constant, is to break the time interval over which ‘f ’ is applied into many discrete
steps and perform a summation using Eq. (8.13.1.1.1-2). While this approach is
general and can be easily implemented on a computer, it is not effective for hand cal-
culations. However, an accurate, but simplified method exists and will be discussed
further in Section 8.13.1.3.

8.13.1.2 The effective-modulus concept is used frequently to simplify creep analysis. The
Effective Modulus effective modulus is defined as follows:

E c (t 0 )
E*c(t,t0) = (Eq 8.13.1.2-1)
1 + C(t,t 0 )

Comparison with Eq. (8.13.1.1.1-1) shows that E*c relates both the immediate strain,
εf, and the time-dependent creep strain, εcr, to the applied stress, f. Figure 8.13.1.2-1
illustrates the effective-modulus concept. Notice that the effective modulus, or the
slope of the stress vs. strain curve, depends on both the time of application of the
load, t0, and the time at which strains are to be determined, t1 or t2. The use of an
effective modulus allows a pseudo-elastic analysis to be performed within a given
time interval.

Eq. (8.13.1.1.1-1) can be rewritten to take advantage of the effective-modulus concept:

f(t)
ε= + ε sh (Eq. 8.13.1.2-2)
E *c (t,t 0 )

JUL 03

7017 Bridge Manual Ch. 8.13 3 8/14/03, 1:04 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.1.2 Effective Modulus/8.13.1.3 Age-Adjusted Effective Modulus

Figure 8.13.1.2-1 f(t0)

Stress vs. Strain for Constant ε(t1) = [1 + C(t 1 ,t 0 )]
Stress, F0 , Applied at time, t0 Ec(t0)
t = t1
Ec(t0)
1 E*c (t 1 ,t 0 ) =
stress [1 + C(t 1 ,t 0 )]
E*c (t 1 ,t 0 ) E*c (t 2 ,t 0 )
1 f(t0)
f(t0) ε(t 2 ) = [1 + C(t2 ,t0 )]
Ec(t0)
t = t2
Ec(t0)
E*c (t 2 ,t 0 ) =
[1 + C(t 2 ,t 0 )]
0

ε(t 1 ) ε(t 2 ) strain

8.13.1.3 Eqs. (8.13.1.1.1-1) and (8.13.1.2-2) are valid only when the stress, f, is constant. In
Age-Adjusted many situations, however, ‘f ’ will vary with time. Figure 8.13.1.1.1-2 graphically
Effective Modulus depicts this condition.

For instance, consider a hollow precast concrete cylinder that is filled with fresh
concrete shortly after the cylinder has been subjected to a constant axial compressive
force. When hardened, the cast-in-place concrete fill will be subjected to a load that
increases with time as creep strains develop in the surrounding precast cylinder. A
similar condition exists in a reinforced concrete member under sustained loads as the
reinforcing steel resists creep strains. Solutions of time-dependent problems such as
these require the ability to predict creep strains under varying load.

As discussed in Section 8.13.1.1.1, one approach would be to divide the problem

into many small time intervals. The stress increment during each interval could be
treated as a new load and, since superposition is valid, Eq. (8.13.1.1.1-2) could be
used to calculate the total response of the member.

An alternative approach, Bazant (1972), uses the aging coefficient, χ, to adjust the
creep coefficient. The aging coefficient accounts for three separate effects:
• When the applied stress, f(t), is increasing, the concrete experiences the maximum
force for only an instant at the end of the time interval (t0,t). At all other times, the
concrete experiences a load that is less than the maximum.
• The concrete is gaining strength, and therefore its modulus is increasing with time.
Portions of the time-varying load that occur earlier are acting on concrete which is
less stiff. Later in the interval, when the loads are larger, the concrete is also stiffer.
• As shown in Chapter 2, for a given concrete in a given environment, the total
creep potential for loads applied to young concrete is larger than for the same loads
applied to old concrete.

JUL 03

7017 Bridge Manual Ch. 8.13 4 8/14/03, 1:04 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.1.3 Age-Adjusted Effective Modulus/8.13.1.4 Properties of Prestressing Steel

Eq. (8.13.1.3-1) should be used when the stress varies over the interval (t0, t):

f(t)
ε=
E c (t 0 )
[ ]
1 + χ(t,t 0 ) C(t,t 0 ) + ε sh (Eq. 8.13.1.3-1)

The corresponding age-adjusted effective modulus is given by:

E c (t 0 )
E *c (t,t 0 ) = (Eq. 8.13.1.3-2)
1 + χ(t,t 0 ) C(t,t 0 )

From here on the effective-modulus will be referred to as defined by Eq. (8.13.1.3-2),

with the understanding that Eq. (8.13.1.2-1) represents the special case of an instan-
taneously applied load for which χ = 1.

There are methods available (Bazant, 1972) by which the aging coefficient can be
computed precisely for different ages at loading and for different concrete properties.
In most practical problems, however, it is sufficiently accurate to use a value of 0.7
or 0.8 for χ, depending on the age of concrete at the beginning of the time interval.
For loads applied at a relatively young concrete age, 0.7 should be used. For all other
situations, 0.8 is generally sufficiently accurate given all of the other uncertainties
present in this type of analysis.

8.13.1.4 Most prestressing materials, including steel bars and strand, exhibit relaxation.
Properties of Relaxation is similar to creep but is defined as the loss of stress in a stressed material
Prestressing Steel held at constant length. The following equation may be used to estimate the relax-
ation, fr, occurring in steel prestressing materials during the interval (t,t0):

f(t 0 )  f(t 0 )   24t + 1 

f r (t,t 0 ) =  − 0.55 log10   (Eq. 8.13.1.4-1)
K r  f y   24t 0 + 1
f(t 0 )
for ≥ 0.55
fy
where
f(t0) = tensile stress at the beginning of the interval
fy = yield strength of the strand
Kr = constant for the material. Values of Kr and fy for some prestressing strand
are provided in Table 8.13.1.4-1.

Table 8.13.1.4-1
Grade 270 Strand Kr fy, ksi
Values of Material Constant,
Kr and Yield Strength, fy Low-Relaxation 45* 243.0
Stress-Relieved 10 229.5
(Normal-Relaxation)

* Also, see Note accompanying Eq. (8.6.5.3-1)

JUL 03

7017 Bridge Manual Ch. 8.13 5 8/14/03, 1:04 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.1.5 Reduced Relaxation under Variable Strain/
8.13.2.2 Method for Time-Dependent Cross-Section Analysis

8.13.1.5 The relaxation predicted by Eq. (8.13.1.4-1) is the intrinsic relaxation, i.e., the relaxation
Reduced Relaxation under that occurs under the theoretical condition of constant strain. In an actual prestressed
Variable Strain concrete member, strain in the prestressing materials is not constant, and is usually
decreasing due to creep and shrinkage of the concrete. Under these circumstances, Eq.
(8.13.1.4-1) will somewhat over-predict relaxation. Various researchers, Ghali and
Travino (1985), Glodowski and Lorenzetti (1972), Hernandez and Gamble (1975),
and others have studied this problem and have proposed various methods of calculating
the reduced relaxation that occurs during intervals of decreasing strain.

With modern low-relaxation prestressing materials, however, relaxation effects are

very small compared to concrete creep and shrinkage. Therefore, it is sufficiently
accurate to adopt a single, standard reduction factor to adjust the intrinsic relaxation
during intervals in which the strain is decreasing. In most practical situations, a fac-
tor equal to 0.8 may be applied to that portion of the relaxation that occurs under
conditions of gradually reducing strain.

8.13.2 The method of analysis illustrated here is essentially no different than a conventional
Analysis of Composite elastic analysis of a prestressed concrete cross-section using transformed section proper-
Cross-Sections ties. Instead of a conventional modulus of elasticity, however, the age-adjusted, effective
modulus is used for all concrete elements in the section. In addition, initial strains must
be considered. The following sections will illustrate the procedure.

8.13.2.1 An initial strain is defined as one that is not directly due to an applied stress. Other
Initial Strains than time-dependent analysis, temperature strain may be the most familiar example
of an initial strain.
In time-dependent analysis of concrete members, the initial strains normally considered
are:
• free shrinkage of the concrete occurring during the interval in question
• creep strains of the concrete, occurring during the interval in question, that are
due to previously applied loads
• the apparent steel strain due to relaxation of prestressing steel during the time
interval in question

To incorporate initial strains into cross-section analysis, it is convenient to calculate

a fictitious restraining load which will restrain the initial strains described above. The
restraining load is then subtracted from any real loads applied to the section. Using
the net load, an analysis is performed in a manner similar to conventional transformed
section analysis. Finally, the internal forces are calculated using the two components.
The internal forces associated with the net load applied to the entire composite sec-
tion are calculated. These are then added to the individual element restraint forces to
give the total actual forces on an individual element of the cross-section. The follow-
ing section provides a detailed description of the procedure used.

8.13.2.2 Several researchers have published methods to perform the time-dependent analysis
Method for Time-Dependent of cross-sections of prestressed concrete members. Two approximate methods, suited
Cross-Section Analysis for manual calculations, and rigorous time-step methods suited only for computer-
ized solutions have been presented. References such as, Branson and Kripanarayanan
(1971), Tadros, et al (1975), Tadros, et al (1977A & B), Dilger (1982-both), Tadros,
et al (1985), and Collins and Mitchell (1991) can be consulted for additional infor-
mation.

JUL 03

7017 Bridge Manual Ch. 8.13 6 8/14/03, 1:04 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.2.2 Method for Time-Dependent Cross-Section Analysis

All of the methods in the cited references, as well as the method presented here, are
based on a pseudo-elastic analysis with the following assumptions and conditions:
• the superposition of creep strains from different stress increments is valid
• concrete members remain uncracked
• stress levels are low compared to the compressive strength of the concrete

It is necessary to consider the entire history of a cross-section in determining its time-

dependent behavior. This history is usually composed of time intervals of varying
lengths. Discrete events (such as the release of prestressing or the application of the
weight of a cast-in-place topping) mark the beginning and end of each time interval.
During the time between these discrete events there is continual creep, shrinkage and
relaxation, as well as internal redistribution of stresses. It is convenient to consider
each discrete event (such as release of prestressing) as though it occurred during a
time interval whose length is zero.

Within a given time interval an elastic analysis with initial strains is performed for the
cross-section being analyzed. Transformed composite section properties are recalculated
for the analysis in each time interval since the properties of the concrete are time-depen-
dent. A unique set of initial strains, dependent upon all of the stress increments applied
during the history of the member, is calculated for each time interval to be analyzed.

The most rigorous methods of time-dependent analysis reduce the time history into
many small steps. As the size of the time-step decreases, the accuracy of the analysis
increases. One such method is described by Tadros et al (1977B). A slightly less accurate,
but greatly simplified method, is presented by Dilger (1982-both) and will be used as
the basis for the procedure described here. That method uses creep-transformed section
properties based on the age-adjusted, effective modulus for a given time interval.

The sign convention for strain, curvature, and section forces in the following proce-
dure are shown in Figure 8.13.2.2-1.

Figure 8.13.2.2-1
Sign Conventions for
Centroid of
composite
Composite Section Analysis 0
section (-)
φ
M
N
ε
(+)
y
0
ycg Strain Section Force

JUL 03

7017 Bridge Manual Ch. 8.13 7 8/14/03, 1:04 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.2.2.1 Steps for Analysis

8.13.2.2.1 The following steps are repeated for each time interval that is to be analyzed over the
Steps for Analysis entire time history of a single cross-section:
1. Calculate the age-adjusted, effective modulus, E*ck, for the interval under consid-
eration for each element, k, comprising the composite section. (Take the effective
modulus, E*c, of the composite section to be that of the concrete beam).
2. Calculate the modular ratio, nk, for each element in the section.

E *ck
nk = (Eq. 8.13.2.2.1-1)
E *c

3. Calculate the transformed composite section area, A, center of gravity, y, and

moment of inertia, I.
A = ∑ A knk (Eq. 8.13.2.2.1-2)

1
y = ∑ y k A knk (Eq. 8.13.2.2.1-3)
A

I = ∑ I k + ( y − y k ) A k  n k
2
(Eq. 8.13.2.2.1-4)
 

4. Calculate the total initial strains, ε0k and φ0k, for each element in the composite
section. For concrete elements, the total initial strains will be those due to free
shrinkage plus those due to creep resulting from previously applied stresses. For
prestressed steel elements, the initial strain will be the apparent strain due to relax-
ation. Typically, non-prestressed steel will have no initial strain. Calculations of
initial strains will be presented in the examples that follow.

5. For each element, k, calculate N0k and M0k, the theoretical restraint forces. Sum
all of the N0k and M0k over the section to give N0 and M0.
N0k = −E*ε0kAk (Eq. 8.13.2.2.1-5)
N0 = ΣN0k (Eq. 8.13.2.2.1-6)
M0k = −E ckIkφ0k
*
(Eq. 8.13.2.2.1-7)
M0 = Σ[M0k − N0k(yk − y)] (Eq. 8.13.2.2.1-8)

6. Subtract the restraint forces, N0 and M0, from the real applied forces, N and M,
and calculate the total strains, ε and φ, in the section.

N − N0
ε = (Eq. 8.13.2.2.1-9)
E *c A
M − M0
φ = (Eq. 8.13.2.2.1-10)
E *c I

7. Calculate the element strains for each element in the composite section.
εk = ε − φ(yk − y) (Eq. 8.13.2.2.1-11)
φk = φ (Eq. 8.13.2.2.1-12)

JUL 03

7017 Bridge Manual Ch. 8.13 8 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.2.2.1 Steps for Analysis/8.13.2.2.2 Example Calculations

8. Calculate the element forces, Nk and Mk, and elastic strains, εfk and φfk, based on
the element strains and the effective modulus, E*ck, for each section element.
Nk = E*ckAkεk + N0k (Eq. 8.13.2.2.1-13)

Nk
εfk = (Eq. 8.13.2.2.1-14)
E *ck A k

Mk = E ck I k φ + M0k
*
(Eq. 8.13.2.2.1-15)

Mk
φfk = (Eq. 8.13.2.2.1-16)
E *ck I k

Steps 1 through 8 are repeated for each time interval to be analyzed over the time
history of the cross-section.

8.13.2.2.2 A 12 in.x12 in. concrete prism, reinforced with (4) # 9 reinforcing bars, was loaded
Example Calculations with a 216-kip axial compressive force immediately after being wet-cured for seven
days. Find the concrete and steel stresses 90 days after loading.

The creep coefficient, C(97,7), is 1.65 and the total free shrinkage strain, εshu, occur-
ring during this period is −400 x 10-6. The initial modulus of elasticity of the concrete,
Eci, is 3,500 ksi and the modulus of elasticity of the steel bars is 29,000 ksi. The strain
in the section immediately after initial loading was calculated to be −0.0003564. The
concrete and steel compressive stresses were 1.248 ksi and 10.34 ksi, respectively.

Step 1, Calculate the age-adjusted, effective modulus for the concrete:

3,500
E*c = = 1,624 ksi
1 + (0.7)(1.65)

Step 2, Calculate the modular ratio for the steel elements:

29,000
ns = = 17.86
1,624

Step 3, Calculate the transformed area of the composite section:

A = (140)(1.0) + (4.00)(17.86) = 211.4 in.2

Step 4, Calculate the initial strain in the concrete (due to both creep and shrinkage):
ε0c = −0.000400 + (1.65)(−0.0003564) = −0.0009881

Step 5, Calculate the restraint forces for the concrete element and the composite section:
N0 = N0c = −(1,624)(−0.0009881)(140) = 224.7 kips

Step 6, Calculate the composite section strain:

0 − 224.7
ε= = − 0.0006544
(1,624)(211.4)

JUL 03

7017 Bridge Manual Ch. 8.13 9 8/14/03, 1:05 PM

PCI BRIDGE DESIGN MANUAL CHAPTER 8

Step 7, Calculating the element strains is straightforward for this example. Both
the concrete and steel strains are equal to the composite section strain,
−0.0006544.

Step 8, Calculate the internal element forces and elastic strains on the concrete and steel:
Nc = (−0.0006544)(1,624)(140) + 224.7 = 75.9 kips (tension)
Ns = (−0.0006544)(29,000)(4.00) = −75.9 kips (compression)

−75.9
εfc = = − 0.000334
(140.0)(1,624)

75.9
fc = = 0.542 ksi (tension)
140

−75.9
εfs = = − 0.000654
(29,000)(4.00)

−75.9
fs = = − 19.0 ksi (compression)
4.00

Therefore, the total concrete and steel stresses at 90 days are 0.706 ksi and 29.3 ksi,
respectively. In this example, the initial strains at 90 days were the result of gradual
changes that had occurred during the preceding time interval. The aging coefficient
in this example was taken to be 0.7 since loading occurred when the concrete was
still relatively young.
Figure 8.13.3-1
Beam Used for Example
Calculations

JUL 03
 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.3 Analysis of Composite Simple-Span Members/
8.13.3.2.1 Example Calculation (at Transfer)

8.13.3 A typical simple-span prestressed concrete bridge beam is shown in Figure 8.13.3-1.
Analysis of Composite Detailed information about this bridge is given in Section 8.13.3.2.1. During its his-
Simple-Span Members tory, this beam will experience several different discrete events. Creep, shrinkage and
relaxation will continue between these discrete events, accompanied by an internal
redistribution of stresses. Table 8.13.3-1 summarizes the significant time intervals
during the life of this typical beam.

Table 8.13.3-1 Interval Event Typical Duration

Beam Lifetime Intervals
1 Strand relaxation before transfer 12 to 24 hours
2 Transfer of prestress 0
3 Creep, shrinkage and relaxation of beam after transfer 30 days to 1 year
4 Placement of cast-in-place deck 0
5 Creep, shrinkage and relaxation of composite deck 7 days to 6 months
and beam
6 Application of superimposed dead load on the 0
composite deck and beam
7 Creep, shrinkage and relaxation of composite 25 years or more
deck and beam

The following sections will describe the analyses performed for each time interval dur-
ing the life of the beam. This is an incremental analysis and the state of stress or strain
in the system at any point in time is equal to the sum of the previous intervals.

8.13.3.1 Eq. (8.13.1.4-1) may be used without adjustment to calculate the intrinsic relaxation
Relaxation of Strands of the strands prior to release or transfer. While the strands are anchored at the ends
Prior to Transfer of the casting bed, the strain is constant, so the intrinsic relaxation is the correct
quantity in this situation.

8.13.3.2 The method described in Section 8.13.2.2 is used to calculate the effects of transfer-
Transfer of Prestress Force ring the prestressing force. Transformed section properties, including the strands,
the concrete, and any additional mild steel, are calculated as described. Because this
is considered to be an interval of zero duration, the creep coefficient is zero for the
concrete. The effective modulus of elasticity of the concrete, therefore, will be equal
to the modulus of elasticity at the time of transfer.

The total prestress force in the strands is treated as an external compressive load
applied to the transformed section at the centroid of the strands.

8.13.3.2.1 Analyze the midspan section of the beam shown in Figure 8.13.3-1 immediately
Example Calculation after the transfer of prestress (note that this is the same beam used in several examples
(at Transfer) in Section 8.6). To simplify, assume that the strands were tensioned and the concrete
cast 18 hours prior to transfer. In practice, it is more likely that strands might be
tensioned 18 hours, and concrete cast 12 hours before release. The beam self-weight
moment at midspan is 3,694 in.-kips.

JUL 03

7017 Bridge Manual Ch. 8.13 11 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.3.2.1 Example Calculation (at Transfer)

Beam data:
Strands: 1/2-in.-diameter strand, Grade 270, low-
relaxation
Eccentricity at midspan = 16.413 in.
Eccentricity at end of beam = 11.556 in.
Reinforcing Bars: #6, Grade 60
Beam concrete: f ´ci = 4,950 psi at 18 hours, Eci = 4,054 ksi
f ´c = 6,750 psi at 28 days, Ec = 4,734 ksi
εshu = −0.0004 in./in.
Cu = 1.4
Beam section properties: AASH4O-PCI 4ype II
A = 560 in.2
I = 125,390 in.4
yb = 20.27 in.
Composite section properties: Ic = 382,372 in.4
ybc = 36.02 in.
Deck properties: Width = 104 in.
Thickness = 8 in.
f ´c = 4,500 psi at 28 days
εshu = –0.0004 in./in.
Cu = 1.4
Dead loads: Self-weight = 583 plf
Deck weight = 867 plf
Haunch weight = 40 plf
Diaphragm weight = 100 plf
Superimposed dead load = 360 plf

The total force in the strands, prior to release, is equal to the jacking force less relax-
ation losses occurring prior to release:
fpj = (0.75)(270.0) = 202.5 ksi
202.5  202.5   (24)(0.75) + 1 
fr =  − 0.55 log10   = 1.63 ksi
45  243.0   (24)(0.0) + 1 
The force applied to the transformed section at release is:
Pi = (202.5 − 1.63)(28)(0.153) = 860.5 kips = –N

Calculation of transformed composite section properties, i.e. steps 1 through 3 from

Section 8.13.2.2.1, is shown in Table 8.13.3.2.1-1. The modulus of elasticity of
the beam is based on the concrete strength at release. (The section properties of the
bare beam have been adjusted in this example to remove the concrete area occupied
by strands and mild steel bars. In practice, this refinement may be omitted with no
significant loss of accuracy).

JUL 03

7017 Bridge Manual Ch. 8.13 12 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.3.2.1 mple lculation at nsfe

Table 8.13.3.2.1-1
Calculation of Transformed Composite Section Properties at Transfer

Item Area y cg Moment Modulus Modular (1)x(5) (1)x(5)x(2) (3)x(5) [y-(2)] 2x(1) (8)+(9)
in.2 in. of Inertia ksi Ratio, n in.2 in.3 in.4 in.4 in.4
(1) (2) in.4 (4) (5) (6) (7) (8) (9) (10)
(3)
Beam 554.836 20.362 123,805 4,054 1.000 554.8 11,298 123,805 207 124,011
Strands 4.284 3.857 10.93 28,500 7.030 30.1 116 77 7,609 7,686
Mild steel 0.880 42.000 0.0 29,000 7.154 6.3 264 3,116 3,116
Composite
Section 19.752 591.2 11,678 134,813

Because this is a zero-length interval, there are no initial strains. Steps 4 and 5 may
be omitted for zero-length intervals.

Use Eqs. (8.13.2.2.1-9) and (8.13.2.2.1-10) to calculate the strain and curvature of
the composite section immediately after release:
−860.5
ε= = − 0.000359
(4,054)(591.2)

3,694 − (860.5)(19.752 − 3.857)

φ= = − 1.83 x 10 −5 in.-1
(4,054)(134,813)

Next, calculate the individual element strains. The strain due to transfer of prestress
at the centroid of the strands is:
εp = −0.000359 − (−1.83 x 10-5)(3.857 − 19.752) = −0.000649
The strain at the centroid of the mild steel bars is:
εs = −0.000359 − (−1.83 x 10-5)(42.00 − 19.752) = 0.0000474
The strain at the centroid of the concrete beam section is:
εc = −0.000359 − (−1.83 x 10-5)(20.362 − 19.752) = −0.000348

Finally, calculate the element forces and elastic strains. For this example, since there
are no initial strains, the elastic strains are equal to the total strains that were calcu-
lated above. The force on the strands:
Np = (28,500)(4.284)(−0.000649) = −79.3 kips

With this information, the remaining stress in the prestressing strands can be calculated:
860.5 − 79.3
fp = = 182.4 ksi
4.284

The force in the mild steel bars:

Ns = (29,000)(0.88)(0.0000474) = 1.2 kip
The axial force and moment on the concrete beam section:
Nc = (4,054)(554.8)(−0.000348) = −782.7 kips
Mc = (4,054)(123,805)(−1.83 x 10−5) = −9,185 in.-kips

JUL 03

7017 Bridge Manual Ch. 8.13 13 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.3.3 Creep, Shrinkage and Relaxation after Transfer/
8.13.3.3.1 Example Calculation (after Transfer)

8.13.3.3 Following the transfer of prestress, but before casting the deck, the beam will undergo
Creep, Shrinkage and gradual changes due to creep and shrinkage of the concrete and relaxation of the pre-
Relaxation after Transfer stressing steel. The procedure of Section 8.13.2.2 can be used to analyze these gradual
changes. Initial strains due to concrete creep and shrinkage, as well as the apparent
strain due to strand relaxation, are included in the analysis. Since the changes occur
gradually over this interval, the age-adjusted modulus is used.

8.13.3.3.1 Analyze the midspan section of the beam in Figure 8.13.3-1, using the results of
Example Calculation Example 8.13.3.2.1. Perform the analysis for a time 90 days after casting of the
(after Transfer) beam.

First, calculate the age-adjusted, effective modulus for the concrete beam. The creep
coefficient is:
(90 − 0.75 )0.6
C(90,0.75) = (1.4) = 0.836
10 + (90 − 0.75 )0.6

Using an aging coefficient of 0.7, the age-adjusted, effective modulus for the concrete
is:
4,054
E*c = = 2,558 ksi
1 + (0.7)(0.836)

Calculation of the transformed composite section properties is similar to the proce-

dure in Example 8.13.3.2.1 (see Table 8.13.3.3.1-1). The modulus of elasticity of
the beam concrete is based on the concrete strength at the beginning of the interval,
i.e., 4,950 psi.
Table 8.13.3.3.1-1
Calculation of Transformed Composite Section Properties After Transfer
Item Area y cg Moment of Modulus Modular (1)x(5) (1)x(5)x(2) (3)x(5) [y-(2)] 2x(1) (8)+(9)
in.2 in. Inertia ksi Ratio, n in.2 in.3 in.4 in.4 in.4
(1) (2) in.4 (4) (5) (6) (7) (8) (9) (10)
(3)
Beam 554.836 20.362 123,805 2,558 1.000 554.8 11,298 123,805 484 124,289
Strands 4.284 3.857 10.93 28,500 11.142 47.7 184 122 11,574 11,696
Mild Steel 0.880 42.000 0.0 29,000 11.338 10.0 419 5,083 5,083
Composite
Section 19.429 612.5 11,901 141,067

Unlike the previous example, there are initial strains to consider in association with
the current time interval. First, calculate the initial strain due to shrinkage of the
beam concrete:
90 − 0.75
εsh = S(90,0.75) = ( −0.000400) = − 0.000247
55 + (90 − 0.75)

Next, calculate the creep strain in the beam for this interval. The creep coefficient has
already been computed. The elastic strain and curvature from the previous example
will be used to compute the creep strains occurring during the current interval:
εcr = (0.836)(−0.000348) = −0.000291
φcr = (0.836)(−1.83 x 10-5) = −1.53 x 10-5 in.-1

JUL 03

7017 Bridge Manual Ch. 8.13 14 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.3.3.1 Example Calculation (after Transfer)

The theoretical restraint forces for the concrete are calculated next:
Nc = −(2,558)(−0.000247 − 0.000291)(554.8) = 763.7 kips
M0c = −(2,558)(123,805)(−1.53 x 10−5) = 4,845 in.-kips

The theoretical restraint force for the strands is due to the apparent strain due to
relaxation. Eq. (8.13.1.4-1) is used, along with a reduction factor of 0.8, to compute
the reduced relaxation occurring in the strands during the interval:
182.4  182.4   (24)(90) + 1 
fr(90,0.75) = (0.8)  − 0.55 log10   = 1.34 ksi
45  243.0   (24)(0.75) + 1

The relaxation of stress in the strand is treated as an apparent positive initial strain in
the strand, i.e., an apparent increase in strain without a change in stress. Using Eq.
(8.13.2.2.1-5), with a positive value for ε0p gives the following value for N0p:
 1.34 
N0p = −(28,500)  (4.284) = − 5.7 kips
 28,500 

Summing the individual restraint forces gives the theoretical restraint forces on the
composite transformed section [Eqs. (8.13.2.2.1-6) and (8.13.2.2.1-8)]:
N0 = 763.7 + (−5.7) = 758.0 kips
M0 = 4,845 − (763.7)(20.362 − 19.429) − (−5.7)(3.857 − 19.429) = 4,044 in.-kips

Eqs (8.13.2.2.1-9) and (8.13.2.2.1-10) are used to compute section strain and curvature:
(0) − (758.0)
ε= = − 0.000484
(2,558)(612.5)

(0.0) − (4,044)
φ= = − 1.12 x10 −5 in.-1
(2,558)(141,067)

The element strains in the concrete beam, strands, and mild steel (Eqs. 8.13.2.2.1-11
and 8.13.2.2.1-12) are:
εc = −0.000484 − (−1.12 x 10-5)(20.362 − 19.429) = −0.000474
εp = −0.000484 − (−1.12 x 10-5)(3.857 − 19.429) = −0.000658
εs = −0.000484 − (−1.12 x 10-5)(42.0 − 19.429) = −0.000231

The element forces (Eqs 8.13.2.2.1-13 and 8.12.2.2.1-14) are:

Nc = (2,558)(554.8)(−0.000474) + 763.7 = 91.0 kips
Np = (28,500)(4.284)(−0.000658) + (−5.7) = −86.0 kips
Ns = (29,000)(0.88)(−0.000231) + 0.0 = −5.9 kips
Mc = (2,558)(123,805)(−1.12 x 10-5) + 4,845 = 1,298 in.-kips

Finally, calculate the elastic strains, i.e. the strains due to stress, in the concrete that
occurred during this time interval. These strains will be used to compute creep strains
during future time intervals:

JUL 03

7017 Bridge Manual Ch. 8.13 15 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.3.3.1 Example Calculation (after Transfer)/8.13.3.7 Long-Term Behavior

91.0
εfc = = 6.41 x 10 −5
(2,558)(554.8)
1,298
φfc = = 4.10 x 10 −6 in.−1
(2,558)(123,805)

8.13.3.4 In a typical prestressed concrete beam bridge, the dead weight of a cast-in-place deck,
Placement of plus intermediate diaphragms (where required) will be carried by the bare precast
Cast-in-Place Deck beam. The placement of these loads on the beam is assumed to occur during an
interval of zero length. Analysis of this interval is essentially an elastic analysis using
transformed composite section properties calculated on the basis of the modulus of
elasticity of the concrete at the time of deck placement.

8.13.3.5 Following casting and curing, shrinkage of the deck concrete becomes a significant
Creep, Shrinkage item affecting the state of stress and strain in the composite system. Since the beam
and Relaxation will typically have undergone 40% to 60% of its ultimate shrinkage by the time the
slab is cast, the ongoing shrinkage of the slab will usually be larger than the com-
bination of the ongoing shrinkage and creep of the beam. This produces positive
curvatures and moments, i.e. tending to cause tension in the bottom of the beam,
that gradually diminish with time.

8.13.3.6 Application of the superimposed dead load on the composite deck/beam system may
Application of occur within 14 days of placement of deck concrete or may be delayed for several
Superimposed Dead Load months. Concrete barriers and wearing surfaces are the most common instances of
superimposed dead loads. Usually, it is assumed that these loads are applied during
an interval of zero length, resulting in the performance of an elastic analysis for this
interval. The appropriate values of concrete modulus of elasticity for both the deck
and the beam, based on their respective ages, are used in calculating the transformed
composite section properties.

8.13.3.7 Following the application of superimposed dead loads, the bridge will typically
Long-Term Behavior remain in a constant configuration for several years. During this period, shrinkage
of both the deck and beam will continue, but at a steadily decreasing rate. Similarly,
creep strains in both the deck and beam will continue to develop. Total creep strains
during this interval will be the sum of the creep strains caused by each stress incre-
ment applied during the preceding intervals.

By the time the superimposed dead load is applied, 60% or more of the creep due to
transfer of prestress probably will have occurred. However, only a small percentage
of the creep due to the dead weight of the deck will have occurred at the start of this
final interval.

Usually the largest stress increments on the beam are associated with transfer of pre-
stress, application of deck dead weight, and application of superimposed dead loads.
It is these large, sudden stress increments which produce the majority of the creep
strains. In addition to these stresses, however, we must consider the gradually devel-
oping stresses that occur between the major events in the life of the member. These
gradually developing stresses are due to restrained or differential shrinkage, relaxation
of the strands, and restrained or differential creep. For the purpose of calculating
future creep strains, it is customary to assume that these gradually developing stresses

JUL 03

7017 Bridge Manual Ch. 8.13 16 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.3.7 Long-Term Behavior/8.13.4.1 Effectiveness of Continuity

can be represented by a sudden stress increment applied at the midpoint of the inter-
val during which they occur. As long as these gradually developing stresses are small
compared to the stresses associated with discrete events, the error is small.

8.13.4 In simple-span bridges there will be little or no change in the distribution of forces
Continuous Bridges and moments within the structure as a result of time-dependent deformations.
However, multiple-span bridges which are made continuous for live loads and super-
imposed dead loads, become statically indeterminate after the deck has cured. As a
result, any time-dependent deformations that occur after the time that the deck is
cured will generally induce forces and moments in the beams (Freyermuth, 1969).

Creep of the beams under the net effects of prestressing, self-weight, deck weight,
and superimposed dead loads will tend to produce additional upward camber with
time. Shrinkage of the deck concrete will tend to produce downward camber of the
composite system with time. In addition, loss of prestress due to creep, shrinkage
and relaxation will result in downward camber. Depending on the properties of the
concrete materials and the age at which the beams are erected and subsequently made
continuous, either positive or negative moments may occur over continuous supports
(Oesterle, et al, 1989).

In the situation where beams are made continuous at a relatively young age it is more
likely that positive moments will develop with time at the supports. These positive
restraint moments are the result of the tendency of the beams to continue to camber
upwards as a result of ongoing creep strains associated with the transfer of prestress.
Shrinkage of the deck concrete, loss of prestress, and creep strains due to self-weight,
deck weight, and superimposed dead loads all have a tendency to reduce this positive
moment.

The alternate situation, i.e. where mature beams are erected and made continuous,
could result in negative moments at the supports. In this situation, the time-depen-
dent creep strains associated with the transfer of prestress have diminished to the
point where the effects that produce downward deflection are more significant. This
will induce negative moments as the end rotations that are associated with this sag-
ging at midspan are restrained over the supports.

For the more typical condition of positive moments developing at the piers it is
recommended that reinforcing steel be provided to minimize the detrimental effects
of cracking at the bottom of the concrete diaphragm. This reinforcement may be
accomplished by extending and bending strands from the bottom rows of the beam
into the pier diaphragms. Alternatively, mild steel reinforcing bars protruding from
the ends of the beams can be extended and bent into the diaphragms. If mild steel
bars are utilized, it is essential that these bars extend far enough into the beam to
adequately develop the bars. In addition, different length bars should be used to
avoid the situation where all the bars terminate at one location.

8.13.4.1 The effects of positive moments, and associated diaphragm cracking, on bridge
Effectiveness of Continuity performance continues to be a hotly debated subject. An argument can be made
(Oesterle, et al, 1989) that continuity for live loads becomes unreliable after a small
crack has opened near the bottom of the diaphragm. It is pointed out that a finite end
rotation is required to close this crack, forcing the beam to carry live loads as a simple-

JUL 03

7017 Bridge Manual Ch. 8.13 17 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.4.1 Effectiveness of Continuity/8.13.4.3.1 General Method

span member. Theoretically, this simple-span action results in live load moments that
are significantly higher than those predicted by the design calculations that assumes
full continuity.

Countering this argument, however, is the successful experience of the many agencies
that routinely design precast, prestressed concrete bridges under the assumption of
full continuity for live loads. Distress in the midspan regions of these bridges, predict-
ed by the preceding discussion, has not been reported. In addition, only service load
behavior is significantly affected. Under ultimate loads, end rotations of the beams
will be large enough to close any crack that may have opened, restoring full continu-
ity. Ultimate capacity, therefore, is relatively unaffected by this phenomenon.

It is unlikely that this issue will be settled completely in the near future. In the
meantime, on the basis of the excellent performance of structures of this type, it is
recommended that designers continue to rely on continuous action for the design of
routine bridges and use details at the piers that have proven to be successful.

8.13.4.2 For unusual or special bridges, a time-dependent analysis to predict restraint

Applying Time- moments at the piers may be performed according to the procedure in the follow-
Dependent Effects ing section. Construction sequence restrictions, special pier details, and beam design
modifications are alternatives the designer may consider should such an analysis pre-
dict excessive positive moments.

Specifically, the designer may wish to consider such an analysis when one or more of
the following conditions are present:
• Spans 140 feet and longer in humid climates (shorter span lengths should be con-
sidered for analysis in arid climates due to increased creep and shrinkage)
• Concrete materials whose creep properties are either unknown, i.e. the mix has not
been used previouse whose creep behavior is known to be poor
• Situations where thermal movements due to daily heating and cooling of the deck
are expected to be unusually high

For more information, the reader is advised to consult the references by Mattock
(1961), Fre e m% h (1969), O e# e le e al, (1989), and Dilge (1982-bo h) rega d
analysis of the effects of creep movements in continuous bridges.

8.13.4.3 The following sections describe two methods to evaluate restraint moments in con-
Methods of Analysis tinuous bridges. The first is a general method and the second is a simplification of
the first.

8.13.4.3.1 The analysis of restraint moments in continuous bridges is a relatively straightfor-

General Method ward extension of the methods described in previous sections. Specifically, the fol-
lowing procedure is used:

1. Calculate the time-dependent beam end rotations that would occur under the
effects of prestressing, self-weight, and deck weight acting on the simple-span
beam using the methods described in Section 8.13.3. Consider only the portions
of time-dependent end rotations that occur after the system is made continuous.

JUL 03

7017 Bridge Manual Ch. 8.13 18 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.4.3.1 General Method/8.13.4.3.2.1 Restraint Moment Due to Creep

2. Using the age-adjusted, effective-modulus method, calculate the rotational stiffness

of the beams by conventional stiffness analysis methods. If the beam is prismatic,
use of the gross section properties including the deck, is sufficiently accurate. The
stiffness factors are given by Eqs. (8.13.4.3.1-1) and (8.13.4.3.1-2) for interior
spans and end spans, respectively.
4E *c I 2E * I
ML = θL + c θR (Eq. 8.13.4.3.1-1)
L L
2E *c I 4E * I
MR = θL + c θR (Eq. 8.13.4.3.1-2)
L L

where
θL = left end rotation of beam due to simple span loads
θR = right end rotation of beam due to simple span loads
E*c = age-adjusted modulus of concrete
I = moment of inertia of the gross concrete section
L = span length measured center-to-center of the supports for the continuous
structure

3. Calculate restraint moments equal to the product of the time-dependent end rota-
tions calculated in step 1 and the rotational stiffness calculated in step 2. Any sign
convention may be used, as long as it is consistent.

4. Perform moment distribution analysis for the continuous structure, using the
restraint moments as the fixed end moments and the stiffness properties calcu-
lated in step 2.

The age-adjusted, effective modulus used in step 2 must be based on gradually varying
loads, i.e., use a value of χ that is appropriate for the age and creep coefficient of the
concrete. A value of 0.8 may be used with relatively little error.

The effects of permanent loads applied to the structure after continuity is achieved
may be computed using a similar analysis. Theoretically, the age-adjusted, effective
modulus for this second analysis, however, should be based on a value of χ equal to
unity since the application of the load is considered to be instantaneous. As a practi-
cal matter, however, this distinction will only affect the distribution of moments in
the structure when different creep coefficients are used in different spans. For almost
all situations, it would be sufficiently accurate to incorporate the effects of the super-
imposed dead loads directly into the continuity analysis described above.

8.13.4.3.2 The above general steps can be further simplified into the following approximate
Approximate Method procedures of calculating the restraint moment due to the time-dependent effects.

8.13.4.3.2.1 Only loads introduced before continuity can cause time-dependent restraint moment
Restraint Moment due to creep. Typically, there are pretensioning forces, member self-weight and pos-
Due to Creep sibly deck weight. Each loading case is considered separately. The total effect is
obtained by simple superposition.

JUL 03

7017 Bridge Manual Ch. 8.13 19 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.4.3.2.1 Restraint Moment Due to Creep/
8.13.4.3.2.2 Restraint Moment Due to Differential Shrinkage

The following assumptions are made. The load is introduced at time, t0, and the
modulus of elasticity of concrete at this time is E(t0). The continuity is made at time,
t1, and the modulus of elasticity of the concrete at this time is E(t1). Specifically, the
following procedure is used for each load:
1. Calculate time-dependent material properties:
C(t,t0) is creep at time, t, for concrete loaded at time, t0
C(t,t1) is creep at time, t, for concrete loaded at time, t1
C(t1,t0) is creep at time, t1, for concrete loaded at time, t0
Age-adjusted effective modulus for concrete subjected to gradual loading:
E c (t1 )
E*c(t,t1) = (Eq. 8.13.4.3.2.1-1)
1 + 0.7C(t,t1 )
E c (t 0 )
E*c(t,t0) = (Eq. 8.13.4.3.2.1-2)
C(t,t 0 ) − C(t1 ,t 0 )

2. Perform elastic analysis, assuming that the load was introduced to a continuous
member. Determine the fictitious elastic restraint moments at the supports, Mel

3. Determine the time-dependent multiplier, δc, corresponding to the load:

E *c (t,t1 )
δc = (Eq. 8.13.4.3.2.1-3)
E *c (t,t 0 )

4. Determine the restraining moment, Mcr(t):

Mcr(t) = δcMel (Eq. 8.13.4.3.2.1-4)

8.13.4.3.2.2 The following assumptions are made: (1) The curing of the beam concludes at time,
Restraint Moment Due to t2. (2) The curing of the deck ends at time, t3. Specifically, the following procedure is
Differential Shrinkage used for calculating the restraint moment due to differential shrinkage:
1. Calculate time-dependent material properties
Deck:
Cd(t,t3) is the creep at time, t, for deck concrete loaded at time, t3
εshd(t,t3) is the shrinkage strain of the deck from time t3 to time, t
Ecd(t3) is the modulus of elasticity for deck concrete at time, t3
Beam:
Cb(t,t3) is the creep at time, t, for beam concrete loaded at time, t3
εshb(t,t2) is the shrinkage strain of the beam from time t2 to time, t
εshb(t3,t2) is the shrinkage strain of the beam from time t2 to time, t3
Ecb(t3) is the modulus of elasticity for beam concrete at time, t3
The age-adjusted, effective modulus for concrete subjected to gradual loading:
E cd (t 3 )
E*cd = (Eq. 8.13.4.3.2.2-1)
1 + 0.7Cd (t,t 3 )
E cb (t 3 )
E*cb = (Eq. 8.13.4.3.2.2-2)
1 + 0.7Cb (t,t 3 )

JUL 03

7017 Bridge Manual Ch. 8.13 20 8/14/03, 1:05 PM

 PCI BRIDGE DESIGN MANUAL CHAPTER 8
DESIGN THEORY AND PROCEDURE
8.13.4.3.2.2 Restraint Moment Due to Differential Shrinkage

2. Calculate the shrinkage moment, Msh:

 h 
Msh = Sh d E cd
*
ε shd (t,t 3 ) y tc − d  − AE *cb [ ε shb (t,t 2 ) − ε shb (t 3 ,t 2 )](y bc − y b )
 2
(Eq. 8.13.4.3.2.2-3)

where
S = beam spacing
hd = deck thickness
ytc = distance from centroidal axis of the composite section to the top of the
deck
A = gross area of the non-composite beam
ybc = distance from centroidal axis of composite section to the bottom of the
beam
yb = distance from centroidal axis of non-composite section to the bottom of the
beam

3. Perform moment distribution analysis for the continuous structure, using the
shrinkage moments as the fixed end moments and the stiffness properties cal-
culated from the composite section. The moment at the supports after moment
distribution is the restraint moment, Msr(t), due to the differential shrinkage.

Should this analysis predict net positive moments at the piers, the results should
probably be treated as an upper bound to the actual moments in the structure. It
is likely that the non-prestressed section of the diaphragm between the ends of the
beams would experience some cracking at relatively low moments. This would have
the effect of introducing a slightly “softer” joint than the fully continuous joint that
is assumed by this analysis.

JUL 03

7017 Bridge Manual Ch. 8.13 21 8/14/03, 1:05 PM