epw1 curved steel girders - university at...

Curved Steel GirdersCurved Steel GirdersCurved Steel GirdersCurved Steel Girders

Ed Wasserman, P.E.Civil Engineering DirectorCivil Engineering Director

Tennessee DOT

1

Report to AASHTO TECHNICAL COMMITTEEFOR STEEL DESIGN (T‐14)

Ramp B Over IRamp B Over I‐‐4040Fi ld I ti tiFi ld I ti tiField InvestigationField Investigation

Field Instrumentation and Summary ofField Instrumentation and Summary of Preliminary ResultsA t 13 2010August 13, 2010

(Sponsored by the State of Tennessee)(Sponsored by the State of Tennessee)

2

Ramp B over I40Ramp B over I40Ramp B over I40Ramp B over I40

3

Bridge SummaryBridge SummaryBridge SummaryBridge Summary• Curved I‐girder bridge• Eight spans, total of 1516 ft, Rmin = 750 ft• Five 68 inch deep girders• Five 68 inch deep girders• 43 ft wide deck

5


NCHRP Project 12NCHRP Project 12‐‐7979Guidelines for Analytical Methods &

Erection Engineering of Curved & Skewed Steel Deck‐Girder Bridges

Don White Georgia TechDon White, Georgia TechAugust 2010

6

Project ObjectivesProject ObjectivesProject ObjectivesProject Objectives

1.1. GuidelinesGuidelines1.1. GuidelinesGuidelines– When are simplified 1D or 2D analysis methods

sufficient?– When are 3D methods necessary?

for assessing constructability & f di ti th t t d t& for predicting the constructed geometry

2.2. RecommendationsRecommendationsl f l– Level of analysis

– Plan detailSubmittals– Submittals

7

Sample of Bridges Studied to DateSample of Bridges Studied to Date

8

Bridges StudiedBridges Studied to Dateto DateBridges StudiedBridges Studied to Dateto DateSuite 1, I‐Girder Bridges

FHWA I Girder Test Bridge SR8002 R A 1 Ki f P i PAFHWA I‐Girder Test Bridge(EISCR1)

SR8002 Ramp A‐1, King of Prussia, PA(EISCS3) NHI Example Bridge, 2007

(XICSS5)Ramp B over Robertson Ave & I‐40, p ,

Davidson County, TN(EICCR22a)

9

Bridges StudiedBridges Studied to Dateto DateBridges StudiedBridges Studied to Dateto DateSuite 1, Tub‐Girder Bridges

NCHRP 12‐52 Tub‐girder curved example(XTCCR8 )

NHI Design Example Straight Tub Girder (XTCSN3)

10

Bridges StudiedBridges Studied to Dateto DateBridges StudiedBridges Studied to Dateto DateSuite 3, I‐Girder Bridges

MN/DOT bridge No 27998, Mi li MN

I‐235 EB over E. University Ave., Polk Co., IA(EICSS2)

Minneapolis, MN(EICCS10)

R 0581 Section A01, Cumberland Co., PA

(EISSS5)( )

Ford City Bridge Ford City PAFord City Bridge, Ford City, PA(EICCR11)

11

Bridges StudiedBridges Studied to Dateto DateBridges StudiedBridges Studied to Dateto DateParametric I‐Girder Bridges

(NISCR7), Suite P1 (NISCR2), Suite P1( ),

(NISSS16), Suite P3(NISCR5), Suite P2(NISSS13), Suite P3 12

Data Synthesis & FindingsData Synthesis & FindingsData Synthesis & FindingsData Synthesis & Findings• Effects of neglecting girder warping torsion

o Vertical deflection predictionso Girder layover predictions

f f f h h d• Significance of using cross‐frames without top chords• Influence of type of cross‐frame detailing on straight I girder bridgesI‐girder bridges

• Cases where one should be wary of global 2nd‐order amplificationp

13

Vertical Displacements Vertical Displacements –– Effect of Neglecting Effect of Neglecting pp g gg gWarping Torsion, Curved IWarping Torsion, Curved I‐‐Girder BridgesGirder Bridges

NHI 2010 Example Bridge (XICCS7)NHI 2010 Example Bridge (XICCS7)

Lb/R = 0.031; Lb/Las= 0.137

0.00

2.00

)

GIRDER 1

‐4.00

‐2.00

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

lacemen

t (in

.)

‐8.00

‐6.00

Vertical Disp

G1‐10.00

Normalized Length

u3‐FEA‐NL u3‐1D u3‐MDX u3‐LARSA (grid) 14

Girder Layovers Girder Layovers –– Effect of Neglecting Effect of Neglecting yy g gg gWarping Torsion, Curved IWarping Torsion, Curved I‐‐Girder BridgesGirder Bridges

NHI 2007 Example Bridge (XICCS7)NHI 2007 Example Bridge (XICCS7)

Lb/R = 0.031; Lb/Las= 0.137

15.00

20.00

ent (in

.)

GIRDER 1

0.00

5.00

10.00

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

adial D

isplaceme

‐15.00

‐10.00

‐5.00

Relativ

e Ra

Normalized Length

FEA‐Nonlinear LARSA(grid)

16

Layover PredictionsLayover PredictionsLayover PredictionsLayover Predictions

For “well connected” curvedFor “well connected” curved bridges… The girder layover is predicted accurately at the cross‐frame locationsframe locations

h k d b dStraight skewed bridges areless sensitive to limitations in the modeled torsional stiffness

17

Significance of usingSignificance of usingSignificance of using Significance of using CrossCross‐‐Frames without Top Chords Frames without Top Chords

in Iin I‐‐Girder BridgesGirder Bridges

18

II‐‐Girder Bridges havingGirder Bridges havingII Girder Bridges having Girder Bridges having CrossCross‐‐Frames without Top ChordsFrames without Top Chords

Bridge on SR 1003 (Chicken Road) over US 74, Robeson Co., NC (EISSS3)

L = 133 ft, w = 30.1 ft, left = right = 46.2o, left right4 girders

Total Dead Load Deformations scaled x10 19

Cases where one should be wary Cases where one should be wary of global 2of global 2ndnd‐‐order effectsorder effectsof global 2of global 2 ‐‐order effectsorder effects

20

OnOn‐‐going Researchgoing ResearchOnOn going Researchgoing Research

• Detailed studies of existing bridges• Detailed studies of existing bridges• Walk through of findings with design

i k b idengineers on key bridges • Field measurements & observationsdd l d f b d• Additional studies of parametric bridges

• Thoroughly examine & quantify the conditions where different methods of analysis lose accuracy or fail to predict the y y pbridge responses

21


Ramp B Over IRamp B Over I‐‐4040Fi ld I ti tiFi ld I ti tiField InvestigationField Investigation

Field Instrumentation and Summary ofField Instrumentation and Summary of Preliminary ResultsA t 13 2010August 13, 2010

(Sponsored by the State of Tennessee)(Sponsored by the State of Tennessee)

22

Objectives of this ResearchObjectives of this ResearchObjectives of this ResearchObjectives of this Research

• Instrument selected girders and monitor gstresses throughout the bridge

• Track deformations at two key spansTrack deformations at two key spans• Track thermal effects

f• Compare results with those from analytical work on NCHRP 12‐79 to assess the ability f d ff h d hof different techniques to predict the

behavior during construction

23

Research TeamResearch TeamResearch TeamResearch Team

• PIs: • URAs: – Roberto Leon– Donald White

– Julie Dykas– Kevin Mayer

– Jochen Teizer

• GRAs:

– Fabio Molina

G s– Towhid Bhuyan– Andres Sanchez

• Applied Geomechanics:

– Cagri Ozgur– Juan Jimenez

– Naia Suszek– Jeff Keller

24

CollaboratorsCollaboratorsCollaboratorsCollaborators• Bell and Associates Construction:

– Jeremy Mitchell Project• PDM Steel (Palatka, FL)

– Chris Price– Jeremy Mitchell, Project Manager

– Dennis Howell, Project Superintendent

Chris Price– Jeremy Duckett, Project

Manager– Ben Bristol, Plant Manager

– Joe Howell• Powell Steel Erection:

– David Horton, Project Manager

• HDR (via NCHRP Project 12‐79)– Domenic Coletti– Brandon Chavel

Manager– Mike Trent, Project

Superintendent– Harold West, Project Foreman

• High Steel– Bob Cisneros

• TN DOT:, j• Parsons Brinckerhoff:

– Bryan Estock– Mathew Smith

– Tim Huff, Bridge Designer– Ed Wasserman, Civil

Engineering Director

– Mike Sage

25

RampRamp BB Field InstrumentationField InstrumentationRampRamp B B Field InstrumentationField Instrumentation

26

Instrument SelectionInstrument SelectionInstrument SelectionInstrument Selection• Long‐term stability:

– Low or no drift– Low probability of zero shifts

• Robustness:– Survive erection– Insensitive to wide temperature and humidity changes

l f ld ll• Simple field installation• Automated data acquisition

27

Vibrating wire gaugesVibrating wire gaugesVibrating wire gaugesVibrating wire gauges• Simple and rugged device consisting of a

tensioned wire anchored to two blockstensioned wire anchored to two blocks welded to girder

• Changes in length (i.e., strain) are measured as changes in vibration frequency of the wire

• Includes a thermistor to correct for• Includes a thermistor to correct for temperature variations

• Not a current or voltage device, so it can be connected and disconnected at will

• Limited range (± 1500 ųε)St ti t l• Static measurements only

28

ClinometersClinometersClinometersClinometers• Simple and rugged device consisting of

an electrolytic fluid and several contactan electrolytic fluid and several contact points

• Changes in rotation are measured as h i i t f th fl idchanges in resistance of the fluid

• Includes a thermistor to correct for temperature variationsp

• The transducer is excited and read by stable, low‐noise electronics

• Wide range (± 25 degrees)• Referenced to gravity• Static measurements only• Static measurements only

29

Typical InstallationTypical Installationypyp

• Six vibrating wire gauges, four at flange tips and two at third point on web

• Two clinometers, one at center of web and one at flange tip

30

Total StationTotal StationTotal StationTotal Station

• Combines:• angle measurement, • distance measurement,

t ti t t iti• automatic target recognition

• Accuracy:Accuracy:• 0.6 mm + 1 ppm to prism • 2 mm + 2 ppm to any surface

• Automated storage of data

31

Laser ScannerLaser ScannerLaser ScannerLaser Scanner

•Generates a point cloud a leap forward in terrain and•Generates a point cloud – a leap forward in terrain and environment understanding

• Both distance and intensity data are measured• Detailed 3D processing

32

Instrument LocationsInstrument LocationsInstrument LocationsInstrument LocationsDisplacements Scaled 50xDisplacements Scaled 50xCompletion of Bridge Deck PlacementCompletion of Bridge Deck PlacementCompletion of Bridge Deck PlacementCompletion of Bridge Deck PlacementBridge Deck Not ShownBridge Deck Not Shown

Girders 2 & 5 Girders 1 & 5

Girders 1 & 5 instr. @ midspan &

instr. @ midspan & near piers 1 & 2

Girders 1 & 5 instr. @ midspan & near piers 6 & 7

near pier 7

Girders 1, 2 & 5 instr. @ midspan &midspan & near pier 1 Girders 2 &

5 instr. near pier 2

Girder 1 instr. @ midspan &

Girder 1 instr. @midspan &

Girder 1 instr. @ midspan &

near piers 3 & 4 @ midspan & near piers 4 & 5 near piers 5 & 6

33

Instrument LocationsInstrument LocationsInstrument LocationsInstrument Locations

Cross‐section view showing location ofCross‐section view showing location of different sensors

34

Typical InstallationTypical InstallationTypical InstallationTypical Installation

Gage alignment jigg g j g

Welded blocks35

Data AcquisitionData AcquisitionData AcquisitionData Acquisition• Six boxes, each with up to 32 gauges and 10 clinometers

• Data downloaded by radio communication• Powered by solar panels• Powered by solar panels

DACQ System Diagram

36

Typical InstallationTypical InstallationTypical InstallationTypical Installation

Connecting boxes after erection

Installation of data acquisition boxes 37

Crane Placement, Start of Stage 1Crane Placement, Start of Stage 1Crane Placement, Start of Stage 1Crane Placement, Start of Stage 1

38

Stage 1, Placing 2Stage 1, Placing 2ndnd Girder over Pier 1Girder over Pier 1Stage 1, Placing 2Stage 1, Placing 2 Girder over Pier 1Girder over Pier 1

39

Stage 1, Placing 3Stage 1, Placing 3rdrd Girder over Pier 1Girder over Pier 1Stage 1, Placing 3Stage 1, Placing 3 Girder over Pier 1Girder over Pier 1

40

Photo During Laser ScanningPhoto During Laser ScanningPhoto During Laser ScanningPhoto During Laser Scanning

41

LaserLaser‐‐Scan Image, End of Stage 1Scan Image, End of Stage 1LaserLaser Scan Image, End of Stage 1Scan Image, End of Stage 1

42

Overview of Erection Procedure & Overview of Erection Procedure & Construction Stages Targeted for StudyConstruction Stages Targeted for Study

(Plan Sketches and(Plan Sketches and(Plan Sketches and (Plan Sketches and Corresponding Photos)Corresponding Photos)

43

Summary of Preliminary ResultsSummary of Preliminary ResultsSummary of Preliminary ResultsSummary of Preliminary Results

Lifting of Lifting of Girder G170A14Girder G170A14Girder G170A14Girder G170A14

60

Girder layover during lifting

4.00

6.00

)

GIRDER 170A14 ‐ Holding Crane

y g g

6 00

‐4.00

‐2.00

0.00

2.00

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

al Disp

lacemen

t (in

.yover(in.)

6.5 in. (predicted)

‐10.00

‐8.00

‐6.00

Latera

Normalized Length

u1‐bf‐nonlinear u1‐tf‐nonlinear u1‐rel‐nonlinear u1‐Measured

8.4 in. (measured)

Lay

62

Stresses: (11:36AM‐12:00PM, 3 min readings)

1.00

GIRDER 1 ‐ BOTTOM FLANGE

‐1.00

‐0.50

0.00

0.50

0.8 0.8 0.9 0.9 0.9 0.9 0.9 0.9

fb (ksi)

‐2.50

‐2.00

‐1.50

Normalized LengthFEA Predictions Field Measurements Field Measurements Field Measurements

2.50

3.00

GIRDER 1 ‐ TOP FLANGE

0.50

1.00

1.50

2.00

fb (ksi)

‐0.50

0.00

0.8

Normalized Length

FEA Predictions Field Measurements Field Measurements 63

Lifting of Girder 170Lifting of Girder 170Lifting of Girder 170Lifting of Girder 170

6

Rotation of Girder 170

Splices and cross-frames being placed Steady state

2

4

6

Beam in the air

‐2

0

tion (degrees)

Transverse Rotation

Approx. 5.5o of layover recovered as beam is being placed

‐6

‐4Rotat

Longitudinal Rotation

‐10

‐8

3/7/10 10:57 3/7/10 11:57 3/7/10 12:57 3/7/10 13:57

i

Approx. 6.7o of layover during lifting

Time

64

Girder 170 after InstallationGirder 170 after InstallationGirder 170 after InstallationGirder 170 after Installation

401400

Girder 170, Line 20

30

35

40

1000

1200

1400

20

25

600

800

1000

Microstrain

perature (o

C)

5

10

15

200

400

600M

Temp

0

5

0

200

3/7/10 10:00 3/7/10 16:00 3/7/10 22:00 3/8/10 4:00 3/8/10 10:00 3/8/10 16:00 3/8/10 22:00 3/9/10 4:00 3/9/10 10:00

Time

Top Flange North Bottom 1/4 Web North Bottom Flange North

Bottom Flange South Top Flange South Average Temperature65

Summary of Preliminary ResultsSummary of Preliminary ResultsSummary of Preliminary ResultsSummary of Preliminary Results

Comparison of Analysis MethodsComparison of Analysis Methodsp yp y

66

ComparisonComparison of Simplified vs. 3D FEA Resultsof Simplified vs. 3D FEA ResultsComparisonComparison of Simplified vs. 3D FEA Resultsof Simplified vs. 3D FEA Results


5.00

10.00


0.00

5.00

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0fb(ksi)

‐10.00

‐5.00(ksi)

‐15.00

Normalized Length

fb FEA NL fb 1D fb MDX fb LARSAfb‐FEA‐NL fb‐1D fb‐MDX fb‐LARSA

67


GIRDER 2 TOP FLANGE

2.50

3.00


0.50

1.00

1.50

2.00

f

‐1.00

‐0.50

0.00

0.50

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

fl(ksi)

‐2.00

‐1.50

Normalized Length

FEA‐NL 1D‐Max MDX‐Max LARSA‐Max FEA‐Max

68


GIRDER 1

1.00

2.00

.)

GIRDER 1

‐1.00

0.00

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

acem

ent (in

‐3.00

‐2.00

ertical Disp

la

‐4.00

Ve

Normalized Length

u3‐FEA‐NL u3‐1D u3‐MDX u3‐LARSAu3 FEA NL u3 1D u3 MDX u3 LARSA

69


GIRDER 1

1 00

1.50

.)

GIRDER 1

0.50

1.00

acem

ent (in

.

‐0.50

0.00

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Radial Disp

la

‐1.00

Relativ

e R

Normalized Length

FEA‐Nonlinear LARSA

70

SummarySummarySummarySummary•The comparisons show that the simplified 1D and p p2D methods are sufficient to match the expected response of the bridge, as predicted by the 3D FEA

•Major‐axis bending stresses are accurately captured, even at Span 7, where the girder is subject to negative moment along its entire length

• Flange lateral bending stress predictions based on the V‐Load method equation have the same trend as the 3D prediction… The V‐load method vertical di l t l t i S 8displacements are a less accurate in Span 8

71

Summary (cont )Summary (cont )Summary (cont.)Summary (cont.)• The vertical displacement predictions are reasonably accurate for all the methods The limitations of the 2D gridaccurate for all the methods… The limitations of the 2D grid models in representing the torsional stiffness of the I‐girders does not affect the response

• Radial displacements (layover) are accurately predicted by the 2D grid model at the cross‐frame locations… At these positions the response of the group predominates over thepositions the response of the group predominates over the response of each individual girder… The 2D model is able to capture this behavior.

• Within the unbraced length (i.e., between cross‐frames), the layover prediction is inaccurate… The reason is the absence f i tiff t ib ti i th tof warping stiffness contributions in the computer

formulation of the 2D grid elements72

OnOn‐‐Going ResearchGoing ResearchOnOn Going ResearchGoing Research

• Live load testing of completed structureg p• Additional analyses of construction stages and of live load testing

• Final report (December 2010)

73

Live LoadLive LoadLive Load Live Load

•Heavy short wheelbase vehicles (fully loadedHeavy, short wheelbase vehicles (fully loaded tri‐axle dump trucks)

• 10 trucks• 10 trucks• Each truck, approx. 72 kips, truck plus aggregate

• 20 positions on the bridge

Load Position B8Load Position B8Load Position B8Load Position B8

G1, Loading B8, Vertical DisplacementG1, Loading B8, Vertical DisplacementG1, Loading B8, Vertical DisplacementG1, Loading B8, Vertical Displacement2.0

1 0

0.0

1.0

t (in

.)

‐3.0

‐2.0

‐1.0

placem

ent

‐5.0

‐4.0

rtical Disp

‐6.0

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Ver

NormalizedBridgeLengthNormalized Bridge Length

With Parapets Without Parapets Field Measurements

G5, Loading B8, Vertical DisplacementG5, Loading B8, Vertical DisplacementG5, Loading B8, Vertical DisplacementG5, Loading B8, Vertical Displacement2.0

0.0

1.0

t (in

.)

3 0

‐2.0

‐1.0

lacemen

t

‐5.0

‐4.0

‐3.0

tical Disp

‐6.0

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Vert

NormalizedBridgeLengthNormalized Bridge Length

With Parapets Without Parapets Field Measurements

G1, Loading B8, Bottom Flange StressG1, Loading B8, Bottom Flange StressG1, Loading B8, Bottom Flange StressG1, Loading B8, Bottom Flange Stress15

5

10

0fb (ksi)

‐10

‐5

‐15

0.6 0.7 0.8 0.9 1.0

NormalizedBridgeLengthNormalized Bridge LengthWith Parapets Without parapets Field Measurements

G1, Loading B8, Bottom Flange StressG1, Loading B8, Bottom Flange StressG1, Loading B8, Bottom Flange StressG1, Loading B8, Bottom Flange Stress3

0

1

2

‐2

‐1

0

f (ksi)

‐4

‐3

( )

‐6

‐5

0.6 0.7 0.8 0.9 1.0

Normalized Bridge LengthWith Parapets Without parapets Field Measurements

SummarySummarySummarySummary•Unique set of data being developed –distribution of thermal stresses for example•Robust and comprehensive measurements

bl f l fcapable of giving a complete picture of system behavior•Successful measurements during different•Successful measurements during different stages of construction•Good initial correlation between analytical and yexperimental values

83

Partial/Preliminary FindingsPartial/Preliminary FindingsPartial/Preliminary FindingsPartial/Preliminary Findings•Second Order Amplification of the Overall Torsional Displacements and Flange Lateral Bending Stresses can be large when the

l l l dstructural system is relatively narrow, compared to span length•Overall overturning instability and high 2nd•Overall overturning instability and high 2ndOrder Amplification of the girder responses are independent phenomena.p p•Concrete slab placement on narrow stages in phased construction can be problematic.

84

Partial/Preliminary Findings (Cont.)Partial/Preliminary Findings (Cont.)Partial/Preliminary Findings (Cont.)Partial/Preliminary Findings (Cont.)

•Basic beam or frame elements commonly available in analysis and design packages can give rater poor predictions of displacements in cases

l h f ll binvolving the following attributes, or some combination thereof:More highly curved alignmentsMore highly curved alignmentsGirders connected with fewer cross‐framesThere are fewer number of girders in the gbridge cross‐section (final or intermediate stages)Girders have more narrow flanges and/or thinner plates 85

Partial/Preliminary Findings (Cont.)Partial/Preliminary Findings (Cont.)Partial/Preliminary Findings (Cont.)Partial/Preliminary Findings (Cont.)•Common analysis models in software packages do not include torsional stiffness quantification associated withinclude torsional stiffness quantification associated with the warping response (flange lateral bending) of I‐girder bridges, leading to poor displacement predictionsg , g p p p

•Generally, flange lateral bending stresses in positive b di i i d i d b blbending regions in many curved girders can be reasonably estimated with the basic equations in the AASHTO LRFD commentary In the negative bending regions thesecommentary. In the negative bending regions, these stresses are significantly less accurate. This is because the cited equations presume uniform bending, whereas the bending moment gradient in negative bending is quite steep. 86

Partial/Preliminary Findings (Cont.)Partial/Preliminary Findings (Cont.)Partial/Preliminary Findings (Cont.)Partial/Preliminary Findings (Cont.)

•The effect of sequential deck concrete placement on overall bridge response appears to be rather minor compared to the responses obtained from an idealized

d k l l l f hone‐step deck placement, particularly if the positive moment areas are placed first and negative moment areas are placed lastareas are placed last.

87

epw1 curved steel girders - university at...

Documents