Austroads Bridge Conference 2004 Hobart May 2004

Bridge Deck Behaviour Revisited Doug Jenkins Interactive Design Services

Overview:

Changes in computing technology Illustration dynamic analysis animation Features of alternative analysis methods Hamblys comments on FE analysis Summary of findings in paper Examples of advanced analysis techniques Conclusions Copy of presenation: www.interactiveds.com.au

Changes in Computing Technology

1976 Text based punched card input 200 nodes; grillage

Linear static analysis Printed text output

2004 Interactive graphical input 100,000 nodes; 3D brick elements Non-linear and dynamic analysis Interactive, animated graphical output

ftp://download.intel.com/research/silicon/moorespaper.pdf

Speed of Computer Systems Prof. E.L. Wilson

http://www.csiberkeley.com/support_technical_papers.html

Computer Performance
1963 - 2003
10 0000

Relative Performance

10 000 10 00 10 0 10 1 0. 1 19 60

19 70

19 80 Year Main Frame

19 90 D esktop

20 00 Trend

20 10

Animation

Hambly comments

Powerful and versatile analytical ... with a sufficiently large computer,

Often requested by clients, or proposed to a client,

Cumbersome to use and is usually expensive. Choice of element type can be extremely critical Full time occupation which cannot be carried out ... by the senior engineer responsible for the design. Unlikely to have time to understand or verify ... data. Difficult to place his confidence in the results

Features of Analyses
Downstand Grillage Transverse variation in the level of the neutral axis. Transverse and longitudinal in-plane forces Distortion of beam members Torsional and distortional warping effects Local bending effects Model skew decks exactly Plates with downstand beams 3D Brick models

Grillage 1a

Grillage 1b

Grillage 1c

Grillage 2

Grillage 2 - detail

Plate Slab with Downstand Beams

Brick Elements

Brick Elements - detail

Deflections at mid-span
-2 -3 -4 -5
Deflection, mm

10

12

14

-6 -7 -8 -9 -10 -11 -12 Position, m Grillage 1a Grillage 1b Grillage 1c Grillage 2 Plate slab Brick

Top Face Stress

0 0.00 -500

2.00

4.00

6.00

8.00

10.00

12.00

14.00

Stress, kPa

-1000

-1500

-2000

-2500 Position, m Grillage 1a Grillage 1b Grillage 1c Grillage 2 Plate slab Brick

Bottom Face Stress

4500 4000 3500 3000
Stress, kPa

2500 2000 1500 1000 500 0 0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

Position, m Grillage 1a Grillage 1b Grillage 1c Grillage 2 Plate slab Brick

Beam and Plate Analysis Deflected Shape

Beam and Plate Analysis Long. Stress

Brick Elements Long. Stress

Brick Elements Cutting Plane

Brick Elements Cutting Plane, Selected Element

Link Slab

Deflections at mid-span
-2 -3 -4 -5
Deflection, mm

10

12

14

-6 -7 -8 -9 -10 -11 -12 Position, m Grillage 1c Brick Grillage + Link Brick + Link Plate No hinges Plate + Link

Animation

Dynamic Analysis
Deflection at mid Span, Outer Beam
0.002

0.000 0.00

0.50

1.00

1.50

2.00

2.50

-0.002

Deflection, m

-0.004

-0.006

-0.008

-0.010

-0.012

Time, S Plate dynamic Brick Dynamic Plate static Brick Static

Animation

Summary of findings in paper Advantages

Transverse distribution of live loads - significantly reduce maximum design stresses in longitudinal members. Distribution of wheel loads - more accurate estimate top slab bending moments, without the need for introducing separate local analyses. Analysis of secondary effects such as differential temperature and shrinkage Ends of skew decks and link slabs modelled more exactly, including three dimensional effects.

Summary of findings in paper No Longer a Problem

Finite element models may now be produced and analysed using standard computing equipment in a shorter time than a grillage analysis would have taken in the recent past. The accuracy of complex models may be checked against grillage analysis, or individual elements may be checked against simple analysis methods. Three dimensional contour plots of stresses or plots of the deformed shape of structures are easily produced, allowing engineers not directly involved in the analysis to review the results, and check the validity of the model.

Summary of findings in paper Disadvantages

More difficult to extract member actions, particularly for large elements such as bridge beams. Design engineers must be trained in the use of complex software to use it efficiently. Verification process may be more difficult, particularly if detailed analysis has resulted in lower design actions than a simpler analysis.

Review Hambly comments

Powerful and versatile analytical ... with a sufficiently large computer,

Often requested by clients, or proposed to a client,

Cumbersome to use and is usually expensive. Choice of element type can be extremely critical Full time occupation which cannot be carried out ... by the senior engineer responsible for the design. Unlikely to have time to understand or verify ... data. Difficult to place his confidence in the results

Recommendations

Standard analysis procedure:- Plate slab model with longitudinal beam members Use pre and post-processor software, specifically designed for bridge decks. Use brick models to further refine the design, or to investigate the behaviour of nonstandard features. Consider the use of non-linear analysis and slab membrane action - potential for significant refinement of deck slab design.