High Performance Concrete Pavements Project Summary
Chapter 27. OHIO 1, 2, AND 3 (US Route 50, Athens)
Under the TE-30 program, the Ohio Department of Transportation (ODOT) constructed three experimental pavement projects on US 50, approximately 8 km (5 mi) east of the city of Athens (see Figure 78). The projects incorporate a variety of experimental design features, including high-performance concrete mixtures utilizing ground granulated blast furnace slag (GGBFS) (OH 1), alternative dowel bar materials (OH 2), and alternative joint sealing materials (OH 3) (Ioannides et al. 1999; Sargand 2000; Hawkins et al. 2000). Although each project was funded separately under the TE-30 program, they are all located on the same section of roadway and share many of the same design and construction attributes, as well as the same traffic and environmental loadings; therefore, these projects are all described together in this chapter.
Figure 78. Location of OH 1, 2, and 3 projects.
The study objectives for the overall US 50 pavement project may be broken out by each specific study. For OH 1, the evaluation of GGBFS, the primary objective is to evaluate the effectiveness of GGBFS as a partial cement replacement in PCC pavements. The expectation of adding GGBFS to a concrete mix is the achievement of increased workability, increased durability, and increased long-term strength.
For OH 2, the evaluation of alternative dowel bar materials, the general purposes of the study are to evaluate dowel response under a variety of loading and environmental conditions and to compare the measured responses of different types of dowel bars (Sargand 2000). Specific objectives include the following (Sargand 2000):
- Instrument standard steel and fiberglass dowels for the monitoring of strain induced by curing, changing environmental conditions, and applied dynamic forces.
- Record strain measurements periodically over time to determine forces induced in the dowel bars during curing and during changing environmental conditions.
- Record strain measurements in the dowel bars as dynamic loads are applied with the FWD.
- Evaluate strain histories recorded for the in-service pavement.
For OH 3, the evaluation of joint sealing materials, the objectives are to (Ioannides et al. 1999):
- Assess the effectiveness of a variety of joint sealing practices employed after the initial sawing of joints, and to examine their repercussions in terms of reduced construction times and life cycle costs.
- Identify those materials and procedures that are most cost effective.
- Determine the effect of joint sealing techniques on pavement performance.
Project Design and Layout
General Design Information
The US 50 project is a 10.5-km (6.5-mi) segment of highway that was reconstructed and expanded to a new four-lane divided facility. The eastbound lanes of the project were constructed in the fall of 1997, and the westbound lanes were constructed in the fall of 1998 (Ioannides et al. 1999).
The 20-year design traffic loading for this pavement is approximately 11 million ESAL applications. The subgrade over the project site is predominantly a silty clay material (Ioannides et al. 1999).
The cross-sectional design for the projects is a 254-mm (10-in.) JRCP placed over a 102-mm (4-in.) open-graded base course. The open-graded base course in the eastbound direction is a "New Jersey" type nonstabilized base, whereas the open-graded base course in the westbound direction is a "Iowa" type nonstabilized base (Ioannides et al. 1999). A 152-mm (6-in.) crushed aggregate subbase is located beneath the open-graded bases, and is topped with a bituminous prime coat to prevent migration of fines into the open-graded layers (Ioannides et al. 1999). Table 40 provides the actual project gradations for these materials. A 102-mm (4-in.) underdrain was placed at both the outside and inside edges of the pavement to collect infiltrated moisture from the open-graded bases (Ioannides et al. 1999).
The slabs are reinforced with smooth welded wire fabric (WWF) to control random cracking (Sargand 2000). Wire style designation W8.5 x W4 - 6x12 was specified, meaning that the longitudinal wires have a cross sectional area of 54.8 mm2 (0.085 in2) and are spaced 152 mm (6 in.) apart, and the transverse wires have a cross-sectional area of 25.8 mm2 (0.04 in2) and are spaced 305 mm (12 in.) apart. This style designation translates to a longitudinal steel content of 0.14 percent.
The transverse joints are spaced at fixed 6.4-m (21-ft) intervals and contain 38-mm (1.5-in.) diameter, 457-mm (18-in.) long, epoxy-coated dowel bars on 305-mm (12-in.) centers (Sargand 2000). However, some of the joints within the alternative dowel bar project contain either fiberglass dowels or stainless steel tubes filled with concrete (Sargand 2000). Transverse joints were sealed with a preformed compression sealant except for the joints within the joint sealant project. The longitudinal centerline joint is tied with 16-mm (0.62-in.) diameter, 760-mm (30-in.) long, deformed bars spaced at 760-mm (30-in.) intervals (Ioannides et al. 1999).
Plain concrete shoulders were paved separately from the mainline pavement. These were tied to the mainline pavement using 16-mm (0.62-in.) diameter, 76-mm (30-in.) long, deformed tie bars. The outside shoulder is 3 m (10 ft) wide and the inside shoulder is 1.2 m (4 ft) wide (Ioannides 1999).
|SIEVE SIZE||TOTAL PERCENT PASSING|
|NEW JERSEY OPEN-GRADED BASE (EB)||IOWA OPEN-GRADED BASE (WB)||CRUSHED AGGREGATE SUBBASE (EB/WB)|
Project Layout Information
As described previously, the US 50 project actually includes three projects, one evaluating GGBFS, one evaluating alternative dowel bar materials, and one evaluating joint sealant materials. In addition, a control section that does not contain GGBFS is located at the western end of the project. The general layout of these projects is shown in Figure 79. More detailed information on each project is provided in the following sections.
Figure 79. Layout of experimental projects on Ohio US 50.
OH 1, Evaluation of Ground Granulated Blast Furnace Slag
The entire 10.5-km (6.5-mi) length of the US 50 project was constructed using a high-performance concrete mix consisting of a Type I cement with GGBFS replacing 25 percent of the cement (Sargand 2000). An AASHTO #8 gravel (0.13 mm [0.5 in.] top size) was used for the coarse aggregate, and a natural sand was used for the fine aggregate (Sargand 2000). A w/c of 0.44 was used in the mix design. The complete PCC mix design is shown in Table 41.
|PCC MIX DESIGN COMPONENT||QUANTITY|
|Natural Sand||1437 lb/yd3|
|AASHTO #8 Aggregate||1374 lb/yd3|
|Type I Cement||412 lb/yd3|
|Ground granulated blast furnace slag||138 lb/yd3|
|Water Reducer||11 oz/yd3|
|Air Entraining Agent||16.5 oz/yd3|
|Design Slump||3 in.|
Samples from the concrete mix used in the actual paving operation were tested in the laboratory and showed a 28-day compressive strength of 27.6 MPa (4000 lbf/in2) and a 28-day modulus of rupture of 2.76 MPa (400 lbf/in2) (Sargand 2000). The 28-day static modulus of elasticity was 25.92 GPa (3,760,000 lbf/in2) (Sargand 2000).
As previously mentioned, a control pavement section that does not contain GGBFS in the concrete mix is located at the western end of the project, between stations 92+35.4 and 104+40. Other than the mix design, the design of the control section is the same as the GGBFS section.
OH 2, Evaluation of Alternative Dowel Bars
Three types of dowel bars were used in the dowel bar project: epoxy-coated steel dowel bars, fiberglass dowel bars (manufactured by RJD Industries, Inc.), and stainless steel (type 304) tubes filled with concrete. The diameter of the steel and fiberglass dowels bars is 38 mm (1.5 in.), while the stainless steel tubes have an outer diameter of 38 mm (1.5 in.) and an inner diameter of 34 mm (1.35 in.) (Sargand 2000). All bars are 457 mm (18 in.) long.
Most of the US 50 project contains conventional epoxy-coated steel dowel bars. However, three specific test sections, each incorporating one of the load transfer devices under study, were set up near the western-most limits of the project in the eastbound direction to instrument dowel response and to compare the performance of the different load transfer devices. Each test section is made up of six consecutive joints, with the middle two joints containing instrumented dowel bars (see Figure 80). The concrete-filled stainless steel bars were not instrumented because the thin wall thickness did not permit the necessary installation operation to protect the lead wires of the gauges (Sargand 2001).
Figure 80. Layout of dowel test sections on Ohio US 50 project.
Three dowel bars within each joint are instrumented. The instrumented bars are located at distances of 152 mm (6 in.), 762 mm (30 in.), and 1980 mm (78 in.) from the outside edge of the pavement, as shown in Figure 81 (Sargand 2000)
Figure 81. Dowel instrumentation layout for Ohio US 50 project (Sargand 2000).
Each instrumented dowel bar contained a uniaxial strain gauge on the top and the bottom of the bar, and one 45-degree rosette on the side. The uniaxial gauges measure environmental and dynamic strains while the rosette gauges measure only dynamic strains (Sargand 2000).
Two thermocouple units were also installed near each instrumented joint to measure temperatures in the concrete slab. One unit housed three sensors that measure temperatures at depths of 102, 178, and 254 mm (4, 7, and 10 in.) from the surface, and the second unit consists of a single sensor measuring temperatures at a depth of 25 mm (1 in.) below the surface (Sargand 2000).
OH 3, Evaluation of Joint Sealing Materials
The joint sealant test sections are located in both the eastbound and westbound directions, and feature a total of nine different joint sealants. In addition, several pavement sections containing no sealant are included in the study.
Table 42 summarizes the location of the different sealant materials in each direction, as well as the joint channel configuration (see Figure 82) used for each material (Hawkins, Ioannides, and Minkarah 2000). The westbound sections each represent replicate sealant sections of those in the eastbound lanes, with the exception of the Watson Bowman WB-687 in the eastbound lanes, which was replicated using the Watson Bowman WB-812 in the westbound lanes (Ioannides et al. 1999). The eastbound lanes were sealed in October and November of 1997, whereas the westbound lanes were sealed in December 1998 (silicone and compression seals) and April 1999 (hot-poured sealants) (Ioannides et al. 1999).
|SEALANT MATERIAL||SEALANT TYPE||BEGIN STATION||END STATION||JOINT CONFIGURATION||SECTION LENGTH, FT||NO. OF JOINTS|
|Eastbound Direction||TechStar W-050||Preformed||154+00||160+00||5||600||29|
|Crafco 444||Hot Pour||172+00||188+00||1||1600||76|
|Watson Bowman WB-687||Preformed||194+00||200+00||5||600||27|
|Crafco 902 Silicone||Silicone||200+00||206+00||1||600||29|
|Crafco 221||Hot Pour||260+00||266+00||1||600||29|
|Westbound Direction||TechStar W-050||Preformed||133+60||139+60||5||600||29|
|Crafco 221||Hot Pour||172+00||188+00||1||1600||76|
|Crafco 444||Hot Pour||206+00||213+00||1||700||33|
|Watson Bowman WB-812||Preformed||225+00||231+00||5||600||28|
Figure 82. Joint channel configurations used in sealant study on Ohio US 50 project
(Hawkins, Ioannides, and Minkarah 2000).
State Monitoring Activities
The Ohio DOT, in conjunction with researchers from several State universities, monitored the performance of these pavements for 5 years. Annual condition surveys and profile measurements were conducted, along with special FWD testing on the instrumented joints. In addition, detailed joint sealant evaluations following SHRP procedures were performed annually on a selected samples of each sealant material.
Overall pavement performance on this project has been mixed, and may be related to the small top size of the coarse aggregate, the small percentage of reinforcing steel, and the poor support from the nonstabilized bases. Specific findings for each specific study are presented in the following sections.
OH 1, Evaluation of Ground Granulated Blast Furnace Slag
The final report by Sargand, Edwards, and Khoury (2002) provides the results for this study. Several factors related to the performance of the HPC pavement containing 25 percent GGBFS have been evaluated with the following results:
- Temperature gradients generated between the top and bottom of concrete slabs during the cure period can have a significant impact on the development of early cracks. HPC pavement sections placed in October 1997 experienced gradients of 10°C, and developed cracking within 18 hours of placement. One HPC and one standard pavement section placed in October 1998 experienced gradients of only 5°C, and did not develop cracking. The higher temperature gradient in 1997 resulted from a cold front that moved in shortly after the placement of the concrete.
- Large values of strain recorded with the vibrating wire strain gauges and maturity measurements indicated that the HP 1 and HP 2 sections could be expected to crack, as was observed in the field. HP 3 constructed 1 year later of the same concrete mix but during a period of warmer weather did not develop cracks. In this case, both strain and maturity data collected in the field indicated a low probability of cracking.
- Results from HIPERPAV also suggested that sections HP 1 and HP 2 would crack, while HP 3 would not. Predicted strength curves were calculated for the placements, in addition to those provided by the standard HIPERPAV prediction model.
- Section HP 3 had less initial warping than did section SP (standard ODOT paving concrete). Sections HP 1 and 2 developed cracking, precluding effective curling measurement of these slabs.
Based on the laboratory results and field data obtained in this study, the following conclusions were derived (Sargand, Edwards, and Khoury 2002):
- Temperature gradients generated between the surface and bottom of concrete slabs during the curing process can have a significant impact on the formation of early cracks.
- Section HP 3 had less initial warping than did section SP constructed with standard ODOT class C concrete.
- FWD data indicated that, under similar loading conditions, the HP 3 section experienced slightly less deflection at joints than the SP section.
- With limited data available, it was suggested that the moisture in the base at sealed and unsealed joints was similar. In some cases, however, moisture under sealed conditions was observed to be slightly higher, indicating that joint seals might trap moisture under the pavement.
- During FWD testing, the deflection at sealed joints was generally higher than at unsealed joints.
OH 2, Evaluation of Alternative Dowel Bars
An analysis of the strains in both the fiberglass and steel dowel bars under environmental and dynamic loading was conducted (ORITE 1998; Sargand 2000; Sargand 2001). Major findings from that analysis include (Sargand 2000; Sargand 2001):
- In addition to transferring dynamic load across PCC pavement joints, dowel bars serve as a mechanism to reduce the curling and warping of slabs due to curing, and temperature and moisture gradients in the slabs.
- Steel and fiberglass dowels both experienced higher moments from environmental factors than from dynamic loading. The dynamic bending stresses induced by a 56.9 kN (12,800 lb) load were considerably less than the environmental bending stresses induced by a 3 °C (5.4 °F) temperature gradient.
- Steel bars induced greater environmental bending moments than fiberglass bars.
- Significant stresses were induced by steel dowel bars early in the life of this pavement as it cured late in the construction season under minimal temperature and thermal gradients in the slab. Concrete pavements paved in the summer under more severe conditions may reveal even larger environmental stresses.
- Both types of dowels induced a permanent bending moment in the PCC slabs during curing, the magnitude of which is a function of bar stiffness.
- Curling and warping during the first few days after concrete placement can result in large bearing stresses being applied to the concrete around the dowels. This stress may exceed the strength of the concrete at that early age and result in some permanent loss of contact around the bars.
- Steel bars transferred greater dynamic bending moments and vertical shear stresses across transverse joints than fiberglass bars of the same size.
Given these findings, it is concluded that the effects of environmental cycling and dynamic loading both must be included in the design and evaluation of PCC pavement joints (Sargand 2001). Because of the high bearing stresses that can be generated in concrete surrounding dowel bars, this parameter should be considered in dowel bar design, especially during the first few days after placement of concrete (Sargand 2001).
It is noted that these results are based on the analysis of the instrumented steel and fiberglass dowel bars only. The stainless steel tubes were not instrumented for the reason stated earlier.
OH 3, Evaluation of Joint Sealing Materials
The results from this experiment, through the 2001 performance evaluation, have resulted in several observations (Ioannides et al. 1999; Hawkins, Ioannides, and Minkarah 2000):
- The silicone and hot-poured sealants in the eastbound lanes are in fair to poor condition, typically suffering from full-depth adhesion failure.
- The worst of the sealed sections were those with a narrow joint width of 3 mm (0.12 in). In these installations, the sealant material had overflowed and run onto the pavement surface.
- There is a significant difference in the performance of the same joint seal materials from EB (constructed in 1997) and WB (constructed in 1998). This difference is attributed to improvements in installation temperatures, experience, and equipment.
- The joints in this experiment were cleaned only by water- and air-blasting, even when the sealant manufacturers recommended sand blasting. This suggests that some of the adhesion loss may be due to an inadequate cleaning process.
- Both the Watson Bowman and the Delastic compression seals have performed by far best overall in both directions. In the WB direction, the silicones have performed best, but were poor in the EB. The performance of the hot pour materials is very different, being far better in WB in general. However, the Crafco 221 material did relatively well in one EB test section. The TechStar compression seal, however, has developed significant adhesion failure and has sunk into the joint.
- The compression seals have performed by far best overall in both directions. In the WB direction, the silicones have performed best, but were poor in the EB. The performance of the hot pour materials is very different, being far better in WB in general. However, the Crafco 221 material did relatively well in one EB test section.
- Hot pour material appears to have performed better when installed within the manufacturer's recommended temperature range. No specific temperature range is recommended for the silicone materials.
- Roughness measurements made using PSI, IRI, and Mays meter do not provide any conclusive trends relating to pavement performance.
- Assessment of joint seal efficiency has little relationship to pavement condition, at this time. It is recommended to reseal the EB sites, except for the two compression seals for continued performance monitoring.
- The Techstar W-050 material performed poorly in both directions, and is considered unsuitable for pavement applications.
- Currently, the unsealed sections seem to have more spalling, corner, and midslab cracking distress than others, although there is no conclusive pavement performance related trends as yet.
A summary of estimated joint sealant costs on this project is provided in Table 43 (Ioannides et al. 1999). These costs are based solely on the material costs themselves and do not include the costs of backer rods, adhesives, or labor.
|MATERIAL||UNIT COST||ESTIMATED COST/JOINT|
|Watson Bowman WB-812||$1.03/ft||$43.26|
|Watson Bowman WB-687||$0.72/ft||$30.24|
Points of Contact
Ohio Department of Transportation
Office of Pavement Engineering
1980 West Broad Street
Columbus, OH 43223
University of Cincinnati
Department of Civil and Environmental Engineering
741 Baldwin Hall (ML-0071)
P.O. Box 210071
Cincinnati, OH 45221-0071
Ohio Research Institute for Transportation and the Environment
Department of Civil and Environmental Engineering
Athens, OH 45701
Hawkins, B. K., A. M. Ioannides, and I. A. Minkarah. 2000. To Seal or Not to Seal: Construction of a Field Experiment to Resolve an Age-Old Dilemma. Preprint Paper No. 00-0552. 79th Annual Meeting of the Transportation Research Board, Washington, DC.
Ioannides, A. M., I. A. Minkarah, B. K. Hawkins, and J. Sander. 1999. Ohio Route 50 Joint Sealant Experiment - Construction Report (Phases 1 and 2) and Performance to Date (1997-1999). Ohio Department of Transportation, Columbus.
Ohio Research Institute for Transportation and the Environment (ORITE). 1998. Measurement of Dowel Bar Response in Rigid Pavement. ORITE-1. Ohio Department of Transportation, Columbus.
Sargand, S. M. 2000. Performance of Dowel Bars and Rigid Pavement. Draft Final Report. Ohio Department of Transportation, Columbus.
---. 2001. Performance of Dowel Bars and Rigid Pavement. Final Report. Ohio Department of Transportation, Columbus.
Sargand, S. M., W. Edwards, and I. Khoury. 2002. Application of High Performance Concrete in the Pavement System, Structural Response of High Performance Concrete Pavement. Final Report. Ohio Department of Transportation, Columbus.