Dowel Load Transfer Systems Their Evolution and Current Innovations for Sustainable Pavements presented by Mark B. Snyder, Ph.D., P.E. Staff Consultant to American Concrete Pavement Association Past President of ISCP
Presentation Outline Introduction: The Need for Mechanical Load Transfer A Brief History of Pavement Dowels in the U.S. from 1917 present The Drive to Use Alternate Dowel Materials/Shapes Determining Structural Equivalency Consideration of Shear, Bending and Bearing Stress Dowel Design: Dimensions, Placement and Materials Optimization of Dowel Location Dowel Structural Testing and Evaluation
INTRODUCTION: THE NEED FOR MECHANICAL LOAD TRANSFER
Dowels: Critical Structural Components of JCP Provide Load Transfer Reduce slab stresses Reduce slab deflections, potential for erosion of support Restraint of Curl/Warp Deformation Influence Dowel-Concrete Bearing Stress Need to last for expected pavement service life (requires corrosion resistance, other durability) 20 35 years for conventional pavement and repairs 40-100 years for long-life pavements
Load Transfer Ability of a slab to share load with neighboring slabs through shear mechanism(s) Typically quantified in terms of Load Transfer Efficiency (LTE), a deflection-based value Many factors affect LTE: Load transfer mechanisms Aggregate Interlock Dowels/Tie Bars Keyways Edge support Widened lanes, tied concrete shoulders or curb and gutter Decrease edge & corner stresses & deflections Foundation stiffness and shear resistance
Aggregate Interlock Shear between aggregate particles below the initial saw cut May be acceptable for: Few heavy loads Hard, abrasion-resistant coarse aggregate Joint opening <0.03
Effects of Dowel Load Transfer on Pavement Behavior Concrete Pavement Deflections Outside Pavement Edge 12 ft Lanes 5 D i 3 D i 3 D i 2 D i D i D i Longitudinal Centerline (acts as tied PCC Shoulder) Undoweled Transverse Joint Doweled Transverse Joint
0% Load Transfer Load Transfer Efficiency (Deflection-based) Wheel Load Direction of Traffic Approach Slab Leave Slab 100% Load Transfer Wheel Load Unloaded LT (%) = Loaded Direction of Traffic X 100 Approach Slab Leave Slab Typically specify around 70% as threshold for action
Joint Load Transfer Considerations LTE vs. Relative Deflection 1 mil ~ 0.025mm Source: Shiraz Tayabji, Fugro Consultants, Inc.
A BRIEF HISTORY OF PAVEMENT DOWELS IN THE U.S. (1917-PRESENT)
A Brief History of U.S. Dowel Design First U.S. use of dowels in PCCP: 1917-1918 Newport News, VA Army Camps Two 19mm dowels across each 3m lane joint Rapid (but non-uniform) adoption through 20s and 30s 1926 practices: two 13mm x 1.2m, four 16mm x 1.2m, eight 19mm x 0.6m By 1930s, half of all states required dowels!
A Brief History of U.S. Dowel Design Numerous studies in 20s, 30s, 40s and 50s Westergaard, Bradbury, Teller and Sutherland, Teller and Cashell, and others Led to 1956 ACI recommendations that became de facto standards into the 90s: Diameter D/8, 30 cm spacing Embedment to achieve max LT: 8*dia for 19mm or less, 6*dia for larger dowels. 45 cm length chosen to account for joint/dowel placement variability. Recent practices: Trend toward increased diameter, some shorter lengths
Current Dowel Bars (Typical) Cylindrical (round) metallic dowels Typical length = 45 cm Typical diameter Roads: 25 32mm Airports: 32 50mm Typically spaced at 30 cm across transverse joints or wheel paths Epoxy coating or other corrosion-protection typically used in harsh climates (deicing or sea salt exposure) for corrosion protection
THE DRIVE TO USE ALTERNATE MATERIALS AND SHAPES
Driving Factors for Using Alternate Materials and Shapes Improved Corrosion Resistance (Increased Service Life) Improved Performance through Reduced Bearing Stress Elimination of Joint Restraint and Alignment Problems Economy (Reduced Cost of Raw Materials, Shipping) Facilitate Construction Ease of Handling Lighter Weight Products (e.g., FRP, Pipe Dowels, Plate Dowels) Ease of Installation (e.g., plate dowel slot formers, Covex plate dowel slot cuts, etc.) Use in Thin Slabs Eliminate Magnetic Interference
PREVENTION OF CORROSION-RELATED PROBLEMS
The Corrosion Problem Corrosion - the destruction or deterioration of a metal or alloy substrate by direct chemical or electrochemical attack. Corrosion of reinforcing steel and dowels in bridges and pavements causes cracking and spalling. Corrosion costs an estimated $276B per year in the U.S. alone! Corroded dowels obtained from 19- year-old jointed concrete pavement
Effects of Corrosion on Dowels Loss of Cross-Section at Joint Poor Load Transfer Reduced Curl-Warp Restraint Loss of Joint Function (Restraint) Spalling Crack Deterioration Premature Failure
Dowel Corrosion Solutions Barrier Techniques Form Oil, Grease, Paint, Epoxy, Plastic Coating breach corrosion failure FRP Encasement Stainless Steel Cladding and Sleeves Relatively expensive Corrosion at coating breaches (including ends), accelerated due to galvanic reaction.
Dowel Epoxy Coatings Most common approach to corrosion prevention since 1970s Long-term performance has varied with environment, coating properties, construction practice and other factors Concerns with reliability over long performance periods Photo credit: Washington State DOT Photo credit: Tom Burnham, MnDOT
Typical product: AASHTO M254/ASTM 775 (green, flexible ) ASTM 934 (purple/grey, nonflexible ) has been suggested Perception of improved abrasion resistance (but green meets same spec requirement) Mancio et al. (2008) found no difference in corrosion protection What is needed: Dowel Epoxy Coatings Durability, resistance to damage in transport, handling, service Standardized coating thickness
Recommendations: Epoxy Coating Remains least expensive, potentially effective option Only effective if durable and applied with sufficient and uniform thickness Consider use of improved epoxy materials 0.25mm nominal minimum thickness meets or exceeds requirements of all surveyed states Would allow individual measures as thin as 0.2mm if average exceeds 0.25mm Probably not necessary to specify upper limit Self-limiting due to manufacturer costs Potential downside is negligible for dowels
Dowel Corrosion Solutions Corrosion-Resistant and Noncorroding Material 316/316L Stainless Steel (Solid, Tubes) Superior corrosion resistance! Expensive (solid bars and, to a less extent, groutfilled tubes) Deformation and slab cracking concerns (hollow tubes only) Microcomposite Steel and Lower-grade Stainless Steel Sufficient corrosion resistance? GFRP, FRP (Solid, Tubes) Noncorroding! Not yet widely adopted Concerns over structural behavior
Dowel Corrosion Solutions Cathodic Protection Impressed Current Useful in bridge decks, impractical for pavement dowels Galvanic (Sacrificial) 1mm zinc alloy cladding or bonded sleeve Inexpensive and self-regulating
IMPROVED PERFORMANCE THROUGH REDUCED BEARING STRESS
Dowel LT Design Considerations LT achieved through both shear and moment transfer, but moment contribution is small (esp. for joint widths of 6mm or less), so bending stress is not critical. Typical critical dowel load < 1350kg, so shear capacity of dowel is usually not critical (by inspection). What about bearing stress between dowel and supporting concrete at the joint face? Varies with load transferred, joint width, relative stiffness of dowel and concrete, etc. Maximum load transferred varies with slab thickness, foundation support, dowel layout, load placement, etc. Bearing stresses can be critical to performance! Pumping, faulting, fatigue, corner breaks, etc.
Friberg s Dowel-Concrete Bearing Stress σ b = Ky 0 = KP t (2 + βz)/4β 3 E d I d β = (Kd/4E d I d ) 0.25 I d = πd 4 /64 for round dowels I d = bh 3 /12 for rectangular dowels Assumes sufficient embedment to match behavior of Timoshenko 1925 analysis (semi-infinite embedded bar). Free web app (Friberg Single Dowel Analyzer) at: apps.acpa.org
COPES Model: Bearing Stress vs Joint Faulting
Impact of Dowel Diameter on Joint Faulting Faulting for 10 inch slab, ins 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 1" dia dowel 1.25" PCC 1.375" PCC 1.5" dia dowel 0 50 100 150 200 250 300 350 Age, months Example for 10-in slab with specific traffic and climate not a design chart!
Dowel Design Factors That Affect Bearing Stress Dowel Shape Round Elliptical Plates (Various Shapes) Dowel Stiffness Elastic Modulus Shape and Size Number and Placement of Dowels
Optimized Dowel Designs Reduce bearing stress while holding crosssectional area constant (or reducing it) Examples: Hollow Dowels (fill or use end caps) Elliptical Dowels Plate Dowels (Photos: Greenstreak, PNA Construction Technologies, Glen Eder)
Free download at: apps.acpa.org
ELIMINATION OF JOINT RESTRAINT AND ALIGNMENT PROBLEMS
Reduction of Pullout Forces Limits on Dowel Pullout Force: AASHTO M 254 Standard Specification for Corrosion-Resistant Coated Dowel Bars : maximum pullout load shall not exceed 1360 kg (3000 lb) Kansas DOT limits dowel pullout force to 1550 kg (3400 lbs) Michigan DOT limits bond stress (initial load divided by embedded surface area) to 420kPa (60 psi) Example: For a 38mm (1.5-in) diameter dowel with 22.5 cm (9 in) of embedment), maximum allowable initial load is 1160 kg (2545 lb). Some dowel manufacturer s claim that their coatings need no lubricant to meet pullout force requirements a cost savings (materials, labor)
Potential Dowel Misalignment Problems
Restraint of Movement in Area Pavements Source: PNA Construction Technologies
Source: PNA Construction Technologies
Restraint of Odd-shaped Panels and Roundabouts Isolated Circle
Plate Dowel Geometries for Contraction Joints Tolerance line Center line Tolerance line Formed void space on vertical sides of plate
Covex Dowels
Other Factors Driving the Use of Alternate Dowels Reduced Material and/or Shipping Costs Lighter weight dowels = more dowels/baskets per truck Ease of Handling (Installation) Less worker fatigue for lighter dowels Less potential for handling damage of baskets Use in Thin Slabs Eliminate Magnetic Interference Tollway gantry areas Magnetic inductance loops
DETERMINING STRUCTURAL ADEQUACY
Structural Acceptance of Alternative Dowel Systems Underlying question: what do we really need from load transfer systems? Answer: I don t think we really know what we need! ( but I think we know what has worked in the past ) Therefore, easiest path to acceptance: prove comparable behavior and performance of alternate dowel system to current standard (generally cylindrical steel dowels). Long-term field performance Accelerated testing in labs Analytical equivalence
Accelerated Load Testing
Load Transfer Efficiency (%) 100 95 90 85 80 75 70 65 60 55 50 Slab 1 - Epoxy-coated Steel Dowel Bars Slab 2 - Fiber Reinforced Polymer Dowel Bars Slab 3 - Grouted Stainless Steel Pipe Dowel Bars 0.001 0.01 0.1 1 10 100 Applied Load Cycles (in millions)
Basis for System Equivalency Deflection-based Criteria LTE Joint Stability Others? Bearing Stress Typically determined analytically High significance in many faulting models Includes influence of slab stiffness, foundation stiffness (through l)
LTE as a measure of equivalence? LTE is a measure of system behavior, not dowel equivalence. Affected by joint width, total deflection, foundation stiffness, etc. LTE has little meaning without an overall deflection reference Example #1: d UL = 0.02mm, d L = 0.04mm, LTE = 50% but is this joint bad? Example #2: d UL = 0.64mm, d L = 0.80mm, LTE = 80% but is this joint good?
Joint Stability ACI 360 definition: a joints ability to limit differential deflection of adjacent slab panel edges when a service load crosses the joint (t)he smaller the measured differential deflection number the better the joint stability.
Joint Stability Limits ACI 360.R-10): < 0.010 in. (0.25 mm) (small, hard-wheeled lift truck traffic) < 0.020 in. (0.51 mm) (larger, cushioned rubber wheels) What is appropriate for road pavements? Should the criterion vary with functional applications (e.g., streets vs highways)? Should the criterion vary with foundation design and environmental conditions (e.g., stabilized vs unbound base, and wet vs dry climate)?
Bearing Stress Basis An Analytical Approach Estimate critical dowel load Linear distribution of load Structural modeling (i.e., finite element analysis) Estimate bearing stress Friberg s analysis Structural modeling Currently best option for dowels with nonuniform cross-section
Estimating Critical Dowel Load l = (E C h 3 /12k(1 μ 2 )) 0.25 Typical critical dowel load < 1350 kg (3000 lbs) Free web app (Friberg Group Dowel Analyzer) at: apps.acpa.org
Friberg s Dowel-Concrete Bearing Stress σ b = Ky 0 = KP t (2 + βz)/4β 3 E d I d β = (Kd/4E d I d ) 0.25 I d = πd 4 /64 for round dowels I d = bh 3 /12 for rectangular dowels Assumes sufficient embedment to match behavior of Timoshenko 1925 analysis (semi-infinite embedded bar). Free web app (Friberg Single Dowel Analyzer) at: apps.acpa.org
Effects of Embedment Length 12.5cm embedment: peak bearing stress = 19.2 Mpa an 11.6% increase 22.5cm embedment: peak bearing stress = 17.3MPa
Effects of Embedment on Shear Load Capacity
ACI 325 (1956) where: f b = f c (4 d)/3 f b = allowable bearing stress, psi f c = PCC 28-day compressive strength, psi d = dowel diameter, inches Provided factor of safety of 2.5 to 3.2 against bearing stress-related cracking Withdrawn from ACI 325 in 1960s, no replacement guidance provided Still commonly cited today
Example Comparison of Alternate Dowel Behavior Dowel Type Diameter (mm) Dowel Modulus, E (GPa) Applied Shear Force (kg) Dowel Deflection at Joint Face (mm) Bearing Stress (MPa) Metallic 38 200 880 (30cm spacing) 0.023 9.69 Sch 40 Pipe 42 200 880 (30cm spacing) 0.023 9.80 FRP 38 39 880 (30cm spacing) 0.038 15.1 FRP 49 39 880 (30cm spacing) 0.023 9.60 FRP 38 39 590 (20cm spacing) 0.025 10.1
Structural Considerations for GFRP and FRP Dowels FRP, GFRP are relatively low-modulus products (<20% of steel) FRP is anisotropic - modulus varies across and along section Same diameter as steel will result in much higher bearing stresses, higher deflections, lower initial LTE values, more rapid loss of LTE under repeated loads Theory is borne out by lab tests and field experience
Many studies of FRP dowels Davis and Porter (1998) Similar joint LTE for 44mm FRP @ 20cm and 38mm steel @ 30cm Melham (1999) 38mm FRP performed comparably to 25mm steel Univ. of WV (2009) FRP performance OK with good support, close spacing, narrow joints LTE dropped from 94% to 72% for 25mm FRP at 15cm spacing after 2M load cycles when joint width increased from 6.5mm to 13mm.
OPTIMIZATION OF DOWEL LOCATION
Optimized Dowel Spacing Trend toward reducing standard dowel installations from 12 dowels per 3.7m lane to 11 Increase distance from lane edge to outside dowels to reduce incidence of paver-induced misalignment Concentrated dowels in wheel paths Common in dowel bar retrofit applications Some trends for new construction Evaluate bearing stresses for alternate spacings using DowelCAD software
Example Dowel Layout Traffic Direction Mid-depth slab Source: CP Preservation Guide 3 5 dowels/wheel path (typical) 3.7 m 0.6 m Smooth dowels 38 mm dia. 0.3 m typical 1.8 m minimum Common for repairs; some agencies use this concept for new construction (e.g., Utah DOT uses 4 dowels/wheel path)
Closure Dowels play an important role in the performance of concrete pavements They are essential for long life in pavements carrying any significant heavy truck traffic! Economics, sustainability and competition are driving a rapid increase in the development of alternate dowel materials and shapes. We must be open to improvements in pavement dowel technology, but must ensure that new dowels will meet the structural and durability requirements of our pavements.
67 Acknowledgments American Concrete Pavement Association Composite Rebar Technology (Jim Olson) Construction Materials, Inc. (MN) Dayton Superior (Glenn Eder, retired) Federal Highway Administration Fugro Consultants, Inc. (Shiraz Tayabji, now with ARA, Inc.) Jarden Zinc Products (Chris Schenk) Minnesota Department of Transportation (Tom Burnham and Maria Masten) National Concrete Consortium (Maria Masten and Tom Cackler) National Concrete Pavement Technology Center (Dale Harrington) PNA Construction Technologies (Nigel Parkes) University of Pittsburgh (Julie Vandenbossche) University of Minnesota (Lev Khazanovich)
Questions? Discussion?