Service Load testing, Numerical Simulation and Strengthening of Masonry Arch Bridges

1
Service Load testing, Numerical Simulation and
Strengthening of Masonry Arch Bridges
Carl Brookes (presenting) and Paul
Mullett Gifford and Partners 18 November 2004
2
Introduction

• The ARCHTEC method
• Pop Bottle Bridge
• Condition and strength assessments
• Supplementary Load Tests
• Objectives
• Existing condition and after strengthening
• Findings
• Measured trends
• Comparisons between test results and predictions
• Concluding Remarks

3
1. The Archtec System
1

• The basic principle of strengthening

4
Inspection, surveying and Analysis
1
5
Finite/Discrete Element (DE) technique
1

• Representation of the material characteristics
  (elastic and plastic behaviour).
• Modelling of contact-gap-friction effects along
  mortar/dry joints.
• Modelling of steel reinforcement including bond
  failure.
• A solution procedure including the calculation of
  displacement, strain and stress to efficiently
  facilitate the modelling of the constitutive
  behaviour in i), ii) and iii) above and perform
  global structural analysis.

DE Mesh
FE Mesh
Reinforcement Mesh
6
Fully Verified Structural Analysis
1

• Evaluation of predicted strength and general
  behaviour against
• Conventional methods of arch assessment.
• Published data from full-scale tests of
  unstrengthened arches carried out by others.
• Full-scale tests by TRL of bridges strengthened
  by the Archtec method.
• Overall philosophy of unifying all common bridge
  parameters in all analyses

7
Typical simulation of arch bridgeShows strength
assessment and Archtec strengthened solution
1
Shows Von Mises equivalent stresses in N/m2. Red
marks the highest levels of compressive stress in
the masonry. Stresses are not shown for fill,
surfacing and axles, hence blue (zero).
Two axle 40/44 tonne bogie with axle lift off
8
Developed by Partnering
1

• Cintec International Limited
• Anchor manufacture and project management
• Rockfield Software Limited
• Finite Element package ELFEN
• Discrete Element contact technique
• Gifford and Partners
• Engineering, assessment and design

9
Pop Bottle Bridge Arrangement Previously owned by
BRB Residuary Body now transferred to
Lincolnshire County Council
2
10
Pop Bottle Bridge ArrangementPrincipal dimensions
2
Plan of survey
11
Current Condition
2
12
General ConditionIllustrates fractured nature of
transverse cracks at some locations
2
13
Strength Assessment
2

• Original MEXE and mechanism analysis
• 13 tonnes rating calculated
• Special DE assessment by Gifford
• Rating increased to 40/44 tonnes
• Still concern about four transverse cracks
• Recommendation made to use Archtec retrofit to
  stabilise cracks

14
Unstrengthened no weight restriction in place
2
15
DE Strength AssessmentBridge survey and model
discretisation
2
16
Archtec Strengthening arrangement
2
17
Archtec Strengtheninginstallation
2
18
Supplimentary load testing objectives
3

• To demonstrate that ARCHTEC retrofitted anchors
  contribute to the structural behaviour under
  service loads and that the effects are beneficial
  and measurable.
• To validate the use of the DE analytical method
  to predict serviceability related behaviour in
  un-strengthened and strengthened arches.
• Strategy to repeat tests before and after ARCHTEC
  retrofit Tests 1 and 2

19
Why Pop Bottle Bridge?
3

• Its construction and previous use make it an
  ideal representative of British arch bridge stock
  particularly road over railways
• (Previously owned by BRB Residuary Body now
  transferred to Lincolnshire County Council)
• The disused and dismantled railway permit easy
  access for test instrumentation.
• Has significant defects that permit the
  investigation of crack control

20
Test 1 - unstrengthenedReserve of capacity
during test, must remain elastic, risk assessment
3
DE simulation included the following Full depth
transverse cracks No transverse load distribution
½ ULS capacity above which damage may result
Single 11 ½ tonne axle
Extra strength not modelled here Transverse load
distribution Spandrel walls Influence of skew
21
Test 1 2 Masonry Instrumentation
3
500mm long VW gauges
Instrumented lines
8 LVDT crack gauges
4 x 18 VW gauges
LVDTs across cracks
LVDTs vertical displacement
22
Test 2 - Anchor Instrumentation
3
23
Test 2 - Anchor Instrumentation
3
24
Bridge Loading28 cases, all based on single 11.5
tonne axles
3
25
Bridge LoadingThree types of vehicle positioning
3
2. Two vehicles, side by side in different
longitudinal
11.5 tonne axle, wheel contact area
1. Single vehicles in different longitudinal and
transverse positions
3. Two vehicles, back to back in different
longitudinal and transverse positions
26
Supplementary Load Test Findings
4

• Comparison of Unstrengthened and Strengthened
• Intrados vertical displacement (abandoned)
• Crack displacements
• Intrados macro strains
• Comparison of measured and predicted results
• Intrados macro strains
• Anchor strains

27
Results discussed here
4

• Mainly south span
• Only two gauges lost for test 2
• Trucks side by side
• 6 cases out of 28 applied
• Least transverse boundary effects

28
Measured crack displacementsSouth span
4
29
Measured crack displacementsNorth span
4
30
Apparent barrel behaviour
Classic arch type behaviour
Compression of loose material in partially open
cracks
31
Measured crack displacementsSummary of Findings
4

• Maximum absolute crack displacements for
  unstrengthened case 0.046mm
• Similar strengthened value is 0.009mm
• Archtec strengthening prevents significant
  movement of transverse cracks under live load
• Reduces load cycle derived hysteretic damage
• Likely to be beneficial to bridge service life

32
Measured intrados macro strainsSouth span, east
side, LC 9
4
Compression
33
Measured intrados macro strainsSouth span, east
side, LC 8
4
Compression
34
Measured intrados macro strains Summary of
Findings
4

• Archtec strengthening reduces intrados tensile
  strains
• Reduces the risk of loose masonry
• Likely to increase service life
• Strengthening significantly reduces crack strains
• Reduces load cycle derived hysteretic damage
• Likely to be beneficial to bridge service life

35
Predicted ResultsExample show 11.5 tonne axle at
¼ point
4
Shows intrados and macro strains Red marks
extrados results, blue intrados Thin lines show
existing condition, thick after ARCHTEC
Shows principal compressive stresses in N/m2. Red
marks the highest levels of compressive stress in
the masonry.
36
Predicted and measured intrados macro strains -
unstrengthenedSouth span, east side, LC 9
4
DE model used for predictions includes closed
full barrel thickness cracks Predicted results
calculated over 500mm length
Compression
37
Predicted and measured intrados macro strains -
strengthenedSouth span, east side, LC 9
4
Compression
38
Predicted and measured intrados macro strains -
unstrengthenedSouth span, east side, LC 8
4
Compression
39
Predicted and measured intrados macro strains -
strengthenedSouth span, east side, LC 8
4
Compression
40
Predicted and measured intrados macro strains
Summary of Findings
4

• Measured crack strains cannot be easily predicted
• Crack strain measurements can be filtered out
• The distribution of intrados macro strains can be
  predicted
• Some adjustment required to allow for 3D
  behaviour so that the magnitude of conservative
  2D results compare well
• Adjust for skew, transverse load distribution,
  influence of spandrel walls

41
Predicted and measured anchor strainsSouth span,
east side, LC 9
4
42
Predicted and measured anchor strainsSouth span,
east side, LC 8
4
43
Predicted and measured anchor strains Summary of
Findings
4

• The distribution of anchor strains can be
  predicted
• The magnitude of anchor strains can also be
  predicted and improved by further adjustment
• To allow for initially open existing transverse
  cracks
• To allow for 3D behaviour so that correlation of
  the magnitude of conservative 2D results can be
  improved (skew, transverse load distribution,
  influence of spandrel wall)

44
Concluding Remarks - 1
5

• Anchors are stressed under working loads
• Archtec strengthening reduces tensile intrados
  macro strains hence reduce likelihood of
  loosening masonry under cyclic loads
• Anchors across transverse cracks reduce cyclic
  opening and closing under repeated live loads,
  hence reduce hysteretic damage

45
Concluding Remarks - 2
5

• DE simulation results agree well with measured
  values, masonry and anchors. Results are
  conservative because of skew, transverse load
  distribution and spandrel walls
• It has been demonstrated that Archtec
  strengthening can be designed not only for the
  ultimate limit state (strength) but also for the
  serviceability limit state (deflections, strains
  and stress ranges)

46
Serviceability Checking
5

• Serviceability criteria for masonry is not
  established
• Current simple criteria is SLS loads should
  remain below ½ ULS levels
• Limits could be set to define serviceability in
  terms of the following but more research is
  required
• Live load deflections
• Crack widths or macro strains
• Maximum absolute SLS stresses
• Maximum SLS stress range/load cycles in masonry