TRB2003 001620

Effects of Time-Dependent Loss Models for Spliced-Girder Bridges
David A. Tomley, P.E., Manager, Engineering Services, LEAP Software, Inc.

LEAP Software, Inc.
P.O. Box 16827
Tampa, FL 33687-6827
Telephone: (813) 985-9170
FAX: (813) 980-3642
e-mail address: david@leapsoft.com

Lee D. Tanase, President/CEO, LEAP Software, Inc.

LEAP Software, Inc.
P.O. Box 16827
Tampa, FL 33687-6827
Telephone: (813) 985-9170
FAX: (813) 980-3642
e-mail address: lee@leapsoft.com

TRB 2003 Annual Meeting CD-ROM Paper revised from original submittal.

Effects of Time-Dependent Loss Models for Spliced-Girder Concrete Bridges
Lee D. Tanase, President/CEO, LEAP Software, Inc. and

ABSTRACT (WORD COUNT 201)
In order to compare prestress losses over time, a time-dependent finite element analysis was performed on two
precast spliced-girder bridges using the industries currently recommended loss models (ACI-209, CEB-FIP, and
AASHTO LRFD). The same time-dependent analysis was also performed on a simple span conventional
precast/prestressed concrete bridge to compare the results using the current AASHTO prestress loss equations
compared to the prestress loss predictions using the ACI-209, CEB-FIP, and AASHTO LRFD loss models. The
CONSPLICE PT software program was used for the design studies. Key construction stages and the comparison
between prestress and post-tensioning losses over time and the critical beam stresses are highlighted using each of
the three loss models independently.
Simple Span Conventional Precast/Prestressed Bridge
40 ft. single span
prestressing only (16-0.5 270 ksi strands)
precast double-T beam
HS20 Live Load
Spliced-Girder Project Details
Example 1: Twisp River Bridge
Location: State of Washington

192 ft. single span
3 individual precast segments
Temporary supports
Combination prestressing and post-tensioning
Option for 1 or 2 stage stressing sequence for post-tensioning
Example 2: Black Warrior Parkway Bridge
Location: State of Alabama

3-span (208-260-208 ft) continuous with drop-in segment
Variable depth 10 ft. pier segments using constant web height
Lee D. Tanase and David A. Tomley, LEAP Software, Inc.
Page 1
Effects of Time-Dependent Loss Models for Spliced-Girder Concrete Bridges
INTRODUCTION
A finite element-based, time-dependent analysis using either of the three code specified loss models (e.g., ACI-209,
CEB-FIP, and AASHTO LRFD) allows engineers to compare the analysis and design effects over time and more
accurately take into account the variation in material behavior (e.g., increases in concrete strength and modulus of
elasticity, creep and shrinkage effects, and prestress losses) over time. AASHTOs LRFD Commentary section
C5.9.5.1 states that For multistage construction and/or prestressing, the prestress losses should be computed in
consideration of the elapsed time between each stage. Such computation can be handled with the time-steps
method. Further, according to AASHTO LRFD article 5.4.2.3.1, when mix-specific data are not available,
estimates of shrinkage and creep may be made according to AASHTO LRFD using the provisions of; AASHTO
LRFD Articles 5.4.2.3.2 and 5.4.2.3.3, and the CEB-FIP model code, or ACI-209.
A time-dependent analysis, that considers the material variations over time, is important to owners from
which a more accurate computation of camber and deflections can be obtained over time by performing a time-
dependent analysis using either the ACI-209, CEB-FIP, or AASHTO LRFD loss models. This statement is backed
up by AASHTO LRFDs article 5.9.5.4 on Refined Estimates of Time-Dependent Losses and parallel commentary
article C5.9.5.4.1 stating, Estimates of losses due to each time-dependent source, such as creep, shrinkage, or
relaxation, can lead to a better estimate of total losses compared with the values given in Table 5.9.5.3-1. Table
5.9.5.3-1 outlines AASHTO LRFDs approximate lump sum estimate of time-dependent losses as further outlined in
AASHTO LRFD article 5.9.5.3. Insight into the percent differences and the sensitivity between the loss model
results as applicable to spliced girder bridges will be discussed. The CONSPLICE PT software program (1) was
used for the design studies.
Should a time-dependent analysis be performed on a conventional simple span precast/prestressed concrete
girder? A simple span bridge was compared using the Standard AASHTO prestress loss equations (2) compared to
the results of a time-dependent analysis using the ACI-209 (3), CEB-FIP (4), and AASHTO LRFD models (5).
TIME-DEPENDENT MATERIALS
Concrete material properties vary with time, and various models have been developed by different code developing
organizations to predict this behavior. The concrete material properties to be considered include:

fc = Concrete Compressive Strength
fr = Modulus of Rupture
E = Modulus of Elasticity
S = Shrinkage Strain
N = Creep Coefficient
It is well known that the behavior of the concrete material properties vary over time as shown in Figure 1.
As shown, the concrete strength (fc) varies and increases over time as does the modulus of elasticity (E).
A time dependent analysis for spliced-girder bridges is warranted because;
1. The ACI-209, AASHTO LRFD, and CEB-FIP material models take into account the variation of
concrete compressive strength (fc) and modulus of elasticity (E) over time, and
2. The following bridge elements that are typically a part of the spliced-girder bridge construction
have their own, (unique and/or different) material properties:
a. Precast concrete girders
Page 2
b. Cast-in-place concrete splices
c. Cast-in-place concrete deck slab
d. Prestressing strands
e. Post-tensioning tendons
3. Different construction stages and corresponding time durations associated with each bridge
element listed above in (2.). Note, each element enters the spliced-girder construction in different
stages. Therefore an accurate model and representation of the actual construction stages specific
to spliced-girder bridges is essential in order to obtain better analytical predictions regarding the
behavior of the structure.
4. According to AASHTO LRFD article 5.5.2.3 on shrinkage and creep and parallel commentary
C5.4.2.3.1, Creep and shrinkage of concrete are variable properties that depend on a number of
factors, some of which may not be known at the time of design. Without specific physical tests or
prior experience with the materials, the use of the empirical methods referenced in these
Specifications cannot be expected to yield results with errors less than (+/-) 50 percent.

SIMPLE SPAN CONVENTIONAL PRECAST/PRESTRESSED BRIDGE
The first design study using CONSPLICE PT will be for a simple span conventional precast/prestressed bridge with
the following design information:

40 ft. single span
prestressing only (16-0.5 270 ksi strands)
precast double-T beam
HS20 Live Load
Construction stages, duration, elements, and applied loads (Table 1)

The typical design assumptions used for simple span (prestressed only) girder construction includes performing 2
analysis and designs (initial and final).
Due to owner assumptions regarding future wearing surface (FWS) and sacrificial wearing surface assumptions, it is
necessary to investigate the initial design for 2 reasons:
1. Used to compute camber and screed elevations (full deck thickness, 0 FWS, and no sacrificial wearing
surface provisions, this simulates the bridge open for traffic)
2. Verify stresses and ultimate capacity

Whereas, the final design is considered since the final design controls the number of prestressing strands because the
FWS and sacrificial wearing surface assumptions are included in the analysis and design.

What is not considered in either the initial or final analysis is the time dependent behavior of the materials.
AASHTO recognizes the need for a time-dependent analysis using either of the three loss models (AASHTO LRFD,
ACI-209, and CEB-FIP) according to AASHTO LRFD article 5.4.2.3 on shrinkage and creep. AASHTO article
5.9.5.3 addresses approximate lump sum estimates of time-dependent losses which are only applicable to
posttensioned nonsegmental members with spans up to 160 ft. AASHTO article 5.9.5.4 provide guidelines for
estimating time-dependent losses using refined methods that consider construction stages, material variations due to
creep and shrinkage, and relaxation losses.

Typically the future wearing surface (FWS) is included in the analysis up front to account for any
additional future load applied to the structure due to resurfacing maintenance over time, whereas in reality, the FWS
Page 3
load is not applied until some time into the future (i.e., after the bridge is open to traffic). What happens to the
prestress losses when the FWS is applied in the future rather than when the bridge is open to traffic?
As Figure 2 indicates, at mid-span, the prestress force decreases from approximately 400 kips (stage 6) to
350 kips (stage 7). The AASHTO equation method (3) predicts a constant prestress force of approximately 400 kips
over time, whereas the ACI-209 (4), AASHTO LRFD (6), and CEB-FIP (5) loss models all indicate that there is an
additional loss of prestress over time as shown in stages 7 and 8. These results also indicate that, for this example,
using a time-dependent analysis with either loss model will predict more losses and less prestress force over time.
SPLICED-GIRDER PROJECT DETAILS
Example 1: Twisp River Bridge
Location: State of Washington

192 ft. single span
3 individual precast segments
Temporary supports
Option for 1- or 2-stage stressing sequence for post-tensioning
Elevation view (Figure 3)
Cross section details (Figure 4)
Pre-tension and post-tension details (Figure 5)
Construction stages, duration, elements, and loads (Table 2)

The construction stages for the Twisp River Bridge are more involved than the previous example for a conventional
precast/prestressed concrete girder. The spliced-girder stages as shown account for two stages of post-tensioning
(see stages 6 and 10). There are a total of 3 tendons; PT1 and PT2 are stressed during stage 6 prior to the deck pour
and the final PT3 tendon is stressed after the deck has cured during stage 10. If should be noted that the State of
Washington showed two options for post-tensioning in the contract bid documents. The first option, as depicted in
this example, utilizes two stages of post-tensioning. The second option, which was selected by the contractor, only
utilized one stage of post-tensioning. Also, note that during stage 15, a uniform temperature and temperature
gradient was included according to AASHTO specifications (3).
Figures 6 and 7 show the prestress and post-tensioning losses/force (at mid-span) over time using the three
loss models independently. As shown, the prestress and post-tension losses/force and corresponding beam stresses
are similar.
SPLICED-GIRDER PROJECT DETAILS:
Example 2: Black Warrior Parkway Bridge (As Shown in Figure 8)
Location: State of Alabama

3-span (208-260-208 ft) continuous with drop-in segment
Variable depth 10 ft. pier segments using constant web height
Elevation view (Figure 9)
Cross section details (Figure 10)
End-span and drop-in beam details (Figure 11)
Construction stages, duration, elements, and loads (Table 3)

One important note to mention in looking at the construction stages is that a full depth deck replacement was
considered in the analysis because of the two stages of post-tensioning. Since the last tendon (PT4) was added to the
composite section (i.e., beam plus deck), engineers have to investigate the impacts of removing the deck in the
future on beam stresses since there will be 4 post-tensioning tendons acting on the non-composite (i.e., beam only)
Page 4
section. Instead of plotting prestress losses/force over time at mid-span, the bottom of beam stresses are shown in
Figure 12.
Bottom of Beam Stresses at Mid-Span (Compression Positive)
As a comparison, the allowable compressive stress of 4.2 ksi is shown in Figure 12. All computed/actual bottom of
beam stresses over time and for every construction stage are less than the allowable.
It should be noted, that the controlling stage of construction on the final design may not necessarily be the
final construction stage. This depends on the material properties, loads, and cross section (composite or non-
composite). Also the design control could be due to shear, moment, or stresses for any construction stage, therefore
a time-dependent, finite element, analysis provides an engineer this information.
CONCLUSIONS
A time-dependent analysis is warranted for spliced-girder bridges in order to more accurately account for the time-
dependent material behavior and properties of concrete, and in order to reproduce/model the construction stages that
are specific to spliced-girder bridges, and for computing more accurate camber and deflections compared to lump
sum estimates.
For the Twisp River and Black Warrior Parkway spliced-girder bridges, the ACI-209 (3), CEB-FIP (4), and
AASHTO LRFD material models (2) produced similar prestress and post-tensioning losses over time. The 40 ft.
simple span conventional precast/prestressed concrete example study showed that a time-dependent analysis predicts
more losses and less prestress force over time compared to the current AASHTO equations (5) for losses.

RECOMMENDATIONS
The authors acknowledge the need for field measured data in order to compare and/or verify the time-dependent
analytical model responses against actual spliced-girder bridges constructed throughout the United States.

REFERENCES
1. LEAP Software, Inc., CONSPLICE PT (LRFD and AASHTO Standard Design and Analysis of Spliced
Prestressed/Precast Bridge Girders and CIP Concrete Slabs), Tampa, Florida, 2001.
2. American Association of State Highway and Transportation Officials (AASHTO), AASHTO LRFD Bridge
Design Specifications, Second Edition, Washington, D.C., 1998 (including the 2000 Interims).
3. ACI Committee 209, Prediction of Creep, Shrinkage, and Temperature Effects in Concrete Structures (ACI-
209R-92), America Concrete Institute, Detroit, 1992.
4. Comite Euro-International Du Beton, CEB-FIP Model Code, 1990.
5. American Association of State Highway and Transportation Officials (AASHTO), Standard Specifications for
Highway Bridges, 16
th
Ed., Association of General Offices, Washington, D.C., 1996 (including the 1998
Interims).
Page 5
Effects of Time-Dependent Loss Models for Spliced-Girder Bridges
LIST OF TABLES AND FIGURES
FIGURE 1 Variation of Concrete Strength (fc) Over Time
FIGURE 2 Prestress Force During Each Construction Stage (Simple Span Conventional Prestressed Bridge)
FIGURE 3 Elevation View of Twisp River Spliced-Girder Bridge
FIGURE 4 Cross Section of Twisp River Spliced-Girder Bridge
FIGURE 5 Pre-Tension & Post-Tension Details of Twisp River Spliced-Girder Bridge
FIGURE 6 Prestress Force During Each Construction Stage (Twisp River Spliced-Girder Bridge)
FIGURE 7 Post-Tension Force During Each Construction Stage (Twisp River Spliced-Girder Bridge)
FIGURE 8 Black Warrior Parkway, An Example of a Spliced-Girder Bridge
FIGURE 9 Elevation View of Black Warrior Parkway Spliced-Girder Bridge
FIGURE 10 Cross Section of Black Warrior Spliced-Girder Bridge
FIGURE 11 End-Span and Drop-In Beam Details (FLBT-72)
FIGURE 12 Bottom of Beam Stresses During Each Construction Stage (Black Warrior Parkway Spliced-Girder
Bridge)

TABLE 1 Construction Stages for Conventional Prestressed Bridge
TABLE 2 Construction Stages for Twisp River Spliced-Girder Bridge
TABLE 3 Construction Stages for Black Warrior Parkway Spliced-Girder Bridge
Page 6

FIGURE 1 Variation of Concrete Strength (fc) Over Time
(ksi
(Days)
Page 7

FIGURE 2 Prestress Force During Each Construction Stage (Simple Span Conventional Prestressed Bridge)
40 ft. Single Span Double T Beam (Mid-span)
350
400
450
500
0 1 2 3 4 5 6 7 8
Sta ge
P
r
e
s
t
r
e
s
s

F
o
r
c
e

(
K
i
p
s
)
ACI-209 LRFD CEB-FIP AASHTO
Page 8
195' -6"
6"
Brg.-to- Brg. = 194'-6"
Abut.- to- Abut.
L
C
L
C
L
C
L
C
L
C
L
C
A
A
B
B
45' -7" 103' -4"
Post-Tensioning
45' -7"
C 2' -0" C.I.P.
Splice (TYP.)
L
Temporary Bent (TYP.)
Segment 1 Segment 2 Segment 3 (L= 101' -4")
(L= 45' -1")
Abutment Abutment
(L= 45' -1")

FIGURE 3 Elevation View of Twisp River Spliced-Girder Bridge
Page 9
5.275' 5.275'
45'-0"
5 SPA. @ 6.89' = 34.45'
1'-6" 1'-6"
8" Deck
W24PTMG
0.5" Haunch Thickness 0.5" Sacrificial Thickness

FIGURE 4 Cross Section of Twisp River Spliced-Girder Bridge
Page 10
4.25 in. O.D. DUCTS (TYP.)
10 -0.6"
Dia. Strand
2

i
n
.
6"
6"
3.84"
PT3
PT2
PT1

FIGURE 5 Pre-Tension & Post-Tension Details of Twisp River Spliced-Girder Bridge
Page 11

FIGURE 6 Prestress Force During Each Construction Stage (Twisp River Spliced-Girder Bridge)
195 ft. Single Span 3 Segment Bulb-T (Mid-span)
300
350
400
450
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Stage
P
r
e
s
t
r
e
s
s

F
o
r
c
e

(
K
i
p
s
)
ACI-209 LRFD CEB-FIP
Page 12

FIGURE 7 Post-Tension Force During Each Construction Stage (Twisp River Spliced-Girder Bridge)
195 ft. Single Span 3 Segment Bulb-T (Mid-span)
1500
1750
2000
2250
2500
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Stage
P
o
s
t

T
e
n
s
i
o
n

F
o
r
c
e

(
K
i
p
s
)
ACI-209 LRFD CEB-FIP
Page 13

FIGURE 8 Black Warrior Parkway, An Example of a Spliced-Girder Bridge
Page 14
260'-0" 208' -0" 208' -0"
End-span Beam
Pier Segment
Drop-in Beam
A
A
A
A
A
A
676' -0"
Temporary Bent (TYP.)
149' -6" 117' -0" 143' -0" 117' -0" 149' -6"
(Beam 1) (Beam 3)
Pier Segment
(Beam 4)
End-Span Beam
(Beam 5)
C Bent 3
L
(Beam 2)
1' -0" Cast-In-Place Splice (TYP.)
C Bent 6
L C Bent 5
L
C Bent 4
L
C Strongback
L

FIGURE 9 Elevation View of Black Warrior Parkway Spliced-Girder Bridge
Page 15
1' - 4 1/2" 1' - 4 1/2"
59' -11"
6 SPA. @ 9' -0" = 54' -0" 4'-4"
62' -8"
4'-4"
Haunch
Thickness
= 2"
8"

FIGURE 10 Cross Section of Black Warrior Spliced-Girder Bridge
Page 16
4' -1"
1'-8 3/4"
4
'
-
5
"
11"
1'-8 3/4"
4
"
2
"
2' -7"
8"
8" 11"
R
=
6
"
R
=
6
"
6
'
-
0
"
7

1
/
2
"
5

1
/
2
"
4- #9
(full length)
End-span beam length = 149' -0"
Drop-in beam length = 142' -0"

FIGURE 11 End-Span and Drop-In Beam Details (FLBT-72) of Black Warrior Spliced-Girder Bridge
Page 17
Multiple Span w/ Drop-In
0
1
2
3
4
5
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Stage
B
o
t
.

B
e
a
m

S
t
r
e
s
s

(
k
s
i
)
1990 CEB-FIP ACI-209 LRFD Allow.

FIGURE 12 Bottom of Beam Stresses During Each Construction Stage (Black Warrior Parkway Spliced-
Girder Bridge)
Page 18

Stage # Description
Duration
(days)
Total
Duration
(days)
Elements
Activated
Elements
Removed
Loads
1
Construct abutments and install
supports, Place Beams (beam age = 2)
0 0
supports/
beams
n/a beam self wt.
2 Store beams on site 60 60 n/a n/a n/a
3 Time Step (form deck) 20 80 n/a n/a n/a
4 Pour Deck + Added DL 0 80 slab n/a slab self wt.
5 Time Step 30 110 n/a n/a n/a
6 Add Live Load 0 110 n/a n/a Live Load
8
Add Future Wearing Surface + Live
Load
0 4110 n/a n/a SDL + Live Load
TABLE 1 Construction Stages for Conventional Prestressed Bridge
Page 19

Stage # Description
Duration
(days)
Total
Duration
(days)
Elements
Activated
Elements
Removed
Loads
1
Construct abutments & install
temporary bents, active beams
0 0 n/a n/a beam self wt.
2 Place Beams #1, 2, 3 in storage 60 60 n/a n/a n/a
3 Time Step (forming) 20 80 n/a n/a n/a
4
Pour Cast-in-Place Splices +
Diaphragms
0 80 splice n/a
self wt.
diaphragms
5 Time Step (splice curing) 10 90 n/a n/a n/a
6 Stress PT1 & PT2 0 90 tendons strongback n/a
7 Time Step (form deck) 20 110 n/a n/a n/a
8
Pour Deck + Supplemental Thickness
0.5
0 110
slab +
suppl.
n/a
self wt. of
slab
9 Time Step (slab curing) 14 124 n/a n/a n/a
10 Stress PT3 0 124 tendons n/a n/a
12 Remove Temporary Bents 0 135 n/a
temporary
bents
n/a
13
Add Superimposed Dead Loads
(Barrier)
0 135 n/a n/a
superimposed
dead load
15
Add Live Load + Uniform Temp +
Temp Gradient
0 165 n/a n/a
live load +
temperature
17 Sacrificial Wearing Surface 0 4165
sacrificial
thickness
n/a n/a
18 Time Step (Infinity) + Live Load 6000 10165 n/a n/a n/a
19 Deck Removal 0 10165 n/a slab n/a
20 Redeck + Live Load 0 10165 redeck n/a n/a
TABLE 2 Construction Stages for Twisp River Spliced-Girder Bridge
Page 20

Stage # Description
Duration
(days)
Total
Duration
(days)
Elements
Activated
Elements
Removed
Loads
1
Install temporary bents in end spans,
add Pier Segment beams
0 0
supports/
beams
n/a beam self wt.
2 Storage Time for Pier Segment Beams 20 20 n/a n/a n/a
3
Place End Span Beams (assumed time
for cutting strands on these beams)
0 20 beam n/a beam self wt.
4 Time Step (storage of end beams) 20 40 n/a n/a n/a
5
Place Drop-In Beams (assumed time
for cutting strands on this beam)
0 40 beam n/a beam self wt.
6 Time Step (storage of drop-in beam) 10 50 n/a n/a n/a
7
Pour Cast-In-Place Splices and
Diaphragms
0 50 splice n/a
dead load +
diaphragms
8 Time Step (splice curing) 14 64 n/a n/a n/a
9 Stress PT1 & PT2 & PT3 0 64 tendons strongback n/a
10
Time Step (form deck using stay-in-
place forms)
30 94 n/a n/a n/a
11 Pour Deck 0 94 slab n/a slab self wt.
12 Time Step (slab curing) 7 101 n/a n/a n/a
13 Stress PT4 0 101 tendons n/a n/a
14 Remove Temporary Bents 0 101 n/a
temporary
bents
n/a
15
Add Superimposed Dead Loads
(Barrier)
2 103 n/a n/a
superimposed
dead load
17 Add Live Load 0 133 n/a n/a live load
19
Add Future Wearing Surface + Live
Load
0 4133 n/a n/a
superimposed
dead load +
live load
21 Full depth deck replacement/removal 0 8133 n/a deck n/a
22 Pour Deck 0 8133 redeck n/a dead load
23 Time Step (infinity) 3000 11133 n/a n/a n/a
24 Live Load (infinity) 0 11133 n/a n/a live load
TABLE 3 Construction Stages for Black Warrior Parkway Spliced-Girder Bridge

TRB2003 001620

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

TRB2003 001620

Încărcat de

Drepturi de autor:

Formate disponibile

Effects of Time-Dependent Loss Models for Spliced-Girder Bridges

David A. Tomley, P.E., Manager, Engineering Services, LEAP Software, Inc.

S-ar putea să vă placă și