Skip to contentUnited States Department of Transportation - Federal Highway AdministrationSearch FHWAFeedback

Pavements

<<PreviousContentsNext >>

Precast Concrete Panel Systems for Full-Depth Pavement Repairs: Field Trials

Chapter 3 Michigan Field Study

Site Selection

The test sections for the Michigan field study are located along I-675 in Zilwaukee and M-25 in Port Austin. For the existing portland cement concrete (PCC) pavement, cross section structural details, traffic, and number of panels installed are summarized in Table 2. The site was selected in concert with the Michigan Department of Transportation (MDOT's) Bay and Cass City Transportation Service Center (TSC) personnel.

Table 2. Summary of Precast Panel Test Sites
PROJECTROUTE DESIGNATIONJOINT SPACINGPAVEMENT THICKNESSBASE TYPEANNUAL AVERAGE DAILY TRAFFIC (COMMERCIAL)NO. OF PANELS INSTALLED
I-675Principal arterial21.6 m (71 ft)225 mm (9 in.)Dense-graded select base10,400 (5%)8*
M-25Minor arterial30.2 m (99 ft.)900–4,000 (3–11%)12
*Nine panels were installed; however, panel 1 is a conventional full-depth repair, and panels 2–9 are precast panels.

Precast Panel Mixture Design and Fabrication Details

The precast PCC panels were fabricated by the contractor and transported to the project site. The typical PCC mixture design for this study is summarized in Table 3.

Table 3. Portland Cement Concrete Mixture Designs for the Precast Panels
MIX INGREDIENTS DESIGN
Cement312 kg/m³ (526 lbs/yd³)
Water 127 kg/m³ (212 lbs/yd³)
Fine aggregate810 kg/m³ (1,366 lbs/yd³)
Coarse aggregate 10,908 kg/m³ (1,838 lbs/yd³)
Air-entraining agents 0.59 ml/kg (0.9 fl oz/cwt)

The contractor was responsible for documenting the fresh and hardened concrete properties. Typical PCC concrete properties are summarized in Table 4. The average 28-day compressive strength based on 18 specimens was 30 MPa (4,300 lbf/in²).

Table 4. Fresh and Hardened Property Results for the Portland Cement Concrete Mixture
Test data from June 3, 2003.
TIME OF CONCRETE CASTING SPECIMEN ID AIR SLUMP CONC TEMP AIR TEMP AGE FLEX. STR.
(%)(in) (°F) (°F) (psi)
3:10 PM A6.5 2 72 67 43 hrs 533
B 43 hrs544
C
D
Test data from June 6, 2003.
TIME OF CONCRETE CASTING SPECIMEN ID AIR SLUMP CONC TEMP AIR TEMP AGE FLEX. STR.
(%) (in) (°F) (°F) (psi)
3:15 PM A 7 2.75 72 75 66 hrs 644
B 66 hrs 688
C
D
Test data from June 10, 2003.
TIME OF CONCRETE CASTING SPECIMEN ID AIR SLUMP CONC TEMP AIR TEMP AGE FLEX. STR.
(%)(in) (°F) (°F) (psi)
8:55 AM A 7 2.75 73 62 48 hrs 644
B 48 hrs 622
C
D
Test data from June 12, 2003.
TIME OF CONCRETE CASTING SPECIMEN ID AIR SLUMP CONC TEMP AIR TEMP AGE FLEX. STR.
(%) (in) (°F) (°F) (psi)
12:50 PM A 6.5 3 73 65 93 hrs 800
B 93 hrs 733
C 
D 
Note. in x 25.4 = mm; 5 (°F-32)/9 = °C; lbf/in2 x 6.89 = kPa

All panels were 1.8 m (6 ft) long, 3.7 m (12 ft) wide, and 250 mm (10 in.) thick. The precast panels were fitted with three dowel bars 38 mm (1.5 in.) in diameter, 450 mm (18 in.) long, and spaced at 300 mm (12 in.) on center along each wheelpath. Perimeter steel was included (#5 bars) to resist handling and transportation stresses. Steel mesh (10 mm [0.375 in.] in diameter placed at 150-mm [6-in.] intervals and held together with 6-mm [0.25-in.] ties) was placed at the panel mid depth to resist the potential of early-age cracking. The panels were wet-cured for 7 days using wet burlap covers. The 20 precast panels were stockpiled at the ready-mix concrete supplier's yard. Eight panels were installed at the I-675 site, and 12 were installed at the M-25 site. The typical structural details are illustrated in Figure 1.

Figure 2 summarizes the sequence of the fabrication process followed by the contractor. The standard hardware for lifting the slabs is visible in the figure. Figure 2 also illustrates the location and placement of dowel bars and temperature steel.

Figure 1. Structural details of the doweled precast panel.

Structural details of the doweled precast panel. The plan drawings show the side, end, and top views of the 6 ft by 12 ft x 10 in. panel (with load transfer). In the top view, four sets of three dowel bars each appear (two sets on either side). Each set consists of three epoxy-coated dowels 18.0 in. long, 1.5 in. in diameter, at 12.0-in. spacing. Two #5 epoxy-coated reinforcing bars are visible (four total) spanning the width of the panel are shown near its left side, at right angles to the dowel bars. Four 4-ton Dayton Swift lifts are equidistant from the edges, one in each corner of the panel. A 12-ft-side view shows the panel as 12 ft ± 0.25 in. wide with the ends of two sets of three dowels each near the two edges. The two outermost dowels are each 1 ft 3 in. (typically) from the panel's vertical edges, and the innermost dowels are 5 ft 6 in. apart. Four #5 epoxy-coated bars span the distance between the outermost dowels, two set at 3.25 in. and the other two at 6.5 in. below the top of the panel. The bars are spaced 5 ± 0.25 in. from the front of the panel and from each other. The 6-ft-side view shows the panel to be 6 ft ± 0.25 in. deep, with four #5 epoxy-coated bars on the left (front) of the panel. The panel thickness is 10 ± 0.125 in. The dowels are positioned 4 ± 0.125 in. below the top of the panel with 9 ± 0.375 in. of their length embedded within the panel. The distance between the endpoints of dowels on either side of the panel is 6 ft ± 1.25 in.

Figure 2. Fabrication of the doweled precast panels.

a) Reinforcement and dowel bar placement in the formwork.
Fabrication of the doweled precast panels. Series of photographs. Reinforcing bars, dowel bars, and lifts are placed in formworks that are partially filled with wet concrete; the remaining concrete is placed and textured; and the forms are covered for curing. Fabrication of the doweled precast panels. Series of photographs. Reinforcing bars, dowel bars, and lifts are placed in formworks that are partially filled with wet concrete; the remaining concrete is placed and textured; and the forms are covered for curing.

b) Concrete placement for precast panels. c) Texturing of fresh concrete.
Fabrication of the doweled precast panels. Series of photographs. Reinforcing bars, dowel bars, and lifts are placed in formworks that are partially filled with wet concrete; the remaining concrete is placed and textured; and the forms are covered for curing. Fabrication of the doweled precast panels. Series of photographs. Reinforcing bars, dowel bars, and lifts are placed in formworks that are partially filled with wet concrete; the remaining concrete is placed and textured; and the forms are covered for curing.

d) Curing of precast slabs.
Fabrication of the doweled precast panels. Series of photographs. Reinforcing bars, dowel bars, and lifts are placed in formworks that are partially filled with wet concrete; the remaining concrete is placed and textured; and the forms are covered for curing.

Construction Sequence

Prior to the removal of the candidate panels, a distress survey was conducted in accordance with the protocol laid out in the LTPP distress identification manual (FHWA 2003). Typical distresses observed during the survey included mid-panel transverse cracks with associated spalling and deteriorated joints with spalling and asphalt patch deterioration. Examples of the typical distresses are shown in Figure 3. During the field visit the distress information was recorded on a distress documentation form (a sample form is presented in Appendix A).

Figure 3. Typical distresses.

a) Deteriorated joint with cold patch with spalling. b) Medium-severity transverse crack.
Typical distresses. Series of photographs. Shown are a deteriorated joint with cold patch with spalling; a medium-severity transverse crack; a high-severity transverse crack; and a deteriorated joint with spalling. Typical distresses. Series of photographs. Shown are a deteriorated joint with cold patch with spalling; a medium-severity transverse crack; a high-severity transverse crack; and a deteriorated joint with spalling.

c) High-severity transverse crack. d) Deteriorated joint with spalling.
Typical distresses. Series of photographs. Shown are a deteriorated joint with cold patch with spalling; a medium-severity transverse crack; a high-severity transverse crack; and a deteriorated joint with spalling. Typical distresses. Series of photographs. Shown are a deteriorated joint with cold patch with spalling; a medium-severity transverse crack; a high-severity transverse crack; and a deteriorated joint with spalling.

The sequence of operations for offsite activities and onsite activities for the Michigan field study are listed below:

Offsite Activities

  1. Fabrication of the precast panels.
  2. Storage of the fabricated precast panels.

Onsite Activities

  1. Documentation of distresses.
  2. Identifying and marking of the repair boundary.
  3. Sawcutting panel boundaries and slab removal.
  4. Initial cleaning of the exposed base.
  5. Jackhammering of the dowel slots.
  6. Air cleaning and sandblasting of the dowel slots.
  7. Final cleaning and grade adjustment of the base.
  8. Placement of the leveling fill.*
  9. Installation of the precast panel.
  10. Adjustment of panel elevation with respect to the adjacent panels.
  11. Backfilling of the dowel slots.
  12. Sealing of joints.
*Step 8 of the construction process does not exist if the precast panel grade adjustment is done using high-density polyurethane (HDP) foam. In that case, after final cleaning and grade adjustment of the base, the precast panel is placed and portholes for injecting the HDP are drilled. The slab elevation adjustment is achieved by injecting the HDP foam.

Descriptions of the activities listed above are presented in the following section of the report. (The types of distresses addressed by the repair strategy and the panel fabrication process, illustrated in Figure 2, are presented above and will not be discussed further).

Sawcutting Panel Boundaries and Removal

The candidate repair sections were identified and marked by the MDOT personnel (Bay City and Cass City TSC). The slab boundaries were outlined by the contractor and sawcut. The limits of the pavement area to be removed were sawed in the transverse direction. Following the sawcutting operation, the lift hooks were inserted and the distressed slabs were removed using a front-end loader. During this process, the outlines for the dowel slots in the adjacent panels were also cut. This concrete was later carved out using pneumatic jackhammers. Figure 4 illustrates the slab sawing and removal process.

Figure 4. Sawcutting of slab boundaries and slab removal.

a) Sawcutting the existing slab. b) The existing slab after sawcutting.
Sawcutting of slab boundaries and slab removal. Series of photographs. Two one-lane cuts are sawn, using a saw-cutting machine, across the existing slab, delineating the section to be removed. The section is lifted out with a front-end forklift. A lifting hook is shown at one end of the section being removed. Sawcutting of slab boundaries and slab removal. Series of photographs. Two one-lane cuts are sawn, using a saw-cutting machine, across the existing slab, delineating the section to be removed. The section is lifted out with a front-end forklift. A lifting hook is shown at one end of the section being removed.

c) Removing the existing slab.
Sawcutting of slab boundaries and slab removal. Series of photographs. Two one-lane cuts are sawn, using a saw-cutting machine, across the existing slab, delineating the section to be removed. The section is lifted out with a front-end forklift. A lifting hook is shown at one end of the section being removed.

Initial Base Preparation

For all the panels the aggregate base was excavated 38–50 mm (1.5–2 in.) below the bottom of the existing slab to accommodate the thicker, 250-mm (10-in.) precast panel. At this point all concrete debris from the slab removal operation was removed. Dewatering of the base was not required at any of the project sites. Figure 5 illustrates the base preparation activity.

Figure 5. Initial cleaning and preparation of the base.

Initial cleaning and preparation of the base. Photograph. Five workers are cleaning out and smoothing the area where the slab has been removed.

Preparation of Load Transfer Slots

As shown in Figure 6, there are three dowel bars in each wheelpath. The dowel slot cutting and preparation include initial grooving to the required depth with a concrete saw; jackhammering of the concrete to carve out the dowel slot; air cleaning of dowel slot to remove debris and any loose concrete pieces; and sandblasting of the dowel slots. The dowel slots were approximately 100 mm (4 in.) wide and 133 mm (5.25 in.) deep (base of the slot cut). Figure 6 illustrates the slot cutting and preparation process.

Figure 6. Preparation of the load transfer slots.

a) Jackhammering of dowel slots. b) Debris removal from dowel slots.
Preparation of the load transfer slots. Series of photographs. Dowel slots are jackhammered in the slab adjacent to the replacement space; debris is removed from the dowel slots using a shovel; the slots are sandblasted; four sets of three dowel bar slots on the perimeter of the replacement are shown. Preparation of the load transfer slots. Series of photographs. Dowel slots are jackhammered in the slab adjacent to the replacement space; debris is removed from the dowel slots using a shovel; the slots are sandblasted; four sets of three dowel bar slots on the perimeter of the replacement are shown.

c) Sandblasting of dowel slots. d) Completed dowel slots.
Preparation of the load transfer slots. Series of photographs. Dowel slots are jackhammered in the slab adjacent to the replacement space; debris is removed from the dowel slots using a shovel; the slots are sandblasted; four sets of three dowel bar slots on the perimeter of the replacement are shown. Preparation of the load transfer slots. Series of photographs. Dowel slots are jackhammered in the slab adjacent to the replacement space; debris is removed from the dowel slots using a shovel; the slots are sandblasted; four sets of three dowel bar slots on the perimeter of the replacement are shown.

The dowel slots were placed at 300 mm (12 in.) on center. The slots are 375 mm (15 in.) from the nearest longitudinal edge (shoulder or centerline). Figure 7 shows a schematic cross section of the dowel assembly.

Figure 7. Schematic cross section of the dowel assembly.

Schematic cross section of the dowel assembly. The existing slab is 9.0 in. thick and rests on a dense graded aggregate base. A jackhammered and sandblasted dowel slot is 5.5 to 6.0 in. from the top of the pavement with a sloped end. A dowel bar–18.0 in. long and 1.5 in. thick–rests in the slot, 4.0 in. on center from the top of the pavement and half embedded in each slab. The adjacent precast slab is 10.0 in. thick and rests on a thin layer (about 1.0 in.) of flowable fill or HDP placed over the dense graded aggregate base, which is thinner under the precast slab. A space appears between the existing and replacement slabs.

Final Grading and Installation of the Precast Panel

This part of the installation process can be achieved by two different methods: slab grade adjustment using HDP foam or slab grade adjustment using flowable fill.

Elevation Adjustment Using High-Density Polyurethane Foam

Once the dowel slots were prepared and a final cleaning and grading of the base prepared the surface for receiving the precast panels, the precast panels were transported from the flat-bed truck to the excavation using a front-end loader. The HDP foam method of slab stabilization was used for 6 panels along the I-675 site and 10 panels along the M-25 site. Approximately 4–6 holes (16 mm [0.625 in.] in diameter) were drilled per panel to inject the foam. The polyurethane foam is made from two liquid chemicals that combine under heat to form a strong, lightweight, foam-like substance. The chemical reaction between the two materials causes the foam to expand and fill the voids. According to the manufacturer's specification, the HDP foam sets in approximately 15 minutes (approximately 90 percent of full compressive strength), and the precast panel is ready to carry load. For the purpose of slab stabilization, the foam density is about 64 kg/m3 (4 lbs/ft³) with a compressive strength range of 414 to 1,000 kPa (60 to 145 lbf/in²).

Once the slab elevations were verified and deemed acceptable, the dowel slots were grouted and the joints were sealed. Figure 8 illustrates the slab installation process.

Figure 8. Precast panel installation and stabilization using high-density polyurethane foam.

a) Panel installation. b) Panel installation and alignment.
Precast panel installation and stabilization using high-density polyurethane (HDP) foam. Series of photographs. Using the four attached lifts, the replacement slab is dropped into the open bed and aligned with the help of workers. Portholes are drilled for injection of the HDP foam. Precast panel installation and stabilization using high-density polyurethane (HDP) foam. Series of photographs. Using the four attached lifts, the replacement slab is dropped into the open bed and aligned with the help of workers. Portholes are drilled for injection of the HDP foam.

c) Drilling of injection portholes. d) Stabilized slab.
Precast panel installation and stabilization using high-density polyurethane (HDP) foam. Series of photographs. Using the four attached lifts, the replacement slab is dropped into the open bed and aligned with the help of workers. Portholes are drilled for injection of the HDP foam. Precast panel installation and stabilization using high-density polyurethane (HDP) foam. Series of photographs. Using the four attached lifts, the replacement slab is dropped into the open bed and aligned with the help of workers. Portholes are drilled for injection of the HDP foam.

Elevation Adjustment Using Flowable Fill

Two precast panels were stabilized at each of the test sites using flowable fill. The excavation was 38–50 mm (1.5–2 in.) below the bottom of the existing slab to accommodate the thicker 250-mm (10-in.) precast panel. The flowable fill was transported to the project site in a ready-mix concrete truck and discharged directly into the excavation. Figure 9 illustrates the flowable placement and the backfilling of the dowel slots. The fill was leveled to a depth of 250 mm (10 in.) from the surface of the existing slab. The flowable fill mixture design includes 1,020.6 kg (2,250 lb) of sand, 56.7 kg (125 lb) of cement, 136.1 kg (300 lb) of water, and 118 ml (4 fl oz) of air entraining admixture. The average 28-day compressive strength ranged from 1.0 to 1.2 MPa (150–175 lbf/in²). Once the slab elevations were verified and deemed acceptable, the dowel slots were grouted and the joints were sealed.

After the slab was installed and leveled, the dowel slots were backfilled and transverse joints were sealed.

Figure 9. Flowable fill operation and dowel slot backfilling.

a) Placing of flowable fill. b) Leveling of flowable fill.
Flowable fill operation and dowel slot backfilling. Series of photographs. Flowable fill is poured into the empty base and leveled; after the panel is placed, the dowel slots are backfilled. The panel is shown with transverse joints sealed and dowel slots filled. Flowable fill operation and dowel slot backfilling. Series of photographs. Flowable fill is poured into the empty base and leveled; after the panel is placed, the dowel slots are backfilled. The panel is shown with transverse joints sealed and dowel slots filled.

c) Back filling of dowel slots.
Flowable fill operation and dowel slot backfilling. Series of photographs. Flowable fill is poured into the empty base and leveled; after the panel is placed, the dowel slots are backfilled. The panel is shown with transverse joints sealed and dowel slots filled. Flowable fill operation and dowel slot backfilling. Series of photographs. Flowable fill is poured into the empty base and leveled; after the panel is placed, the dowel slots are backfilled. The panel is shown with transverse joints sealed and dowel slots filled.

d) Back-filled dowel slots. e) Completed panel.
Flowable fill operation and dowel slot backfilling. Series of photographs. Flowable fill is poured into the empty base and leveled; after the panel is placed, the dowel slots are backfilled. The panel is shown with transverse joints sealed and dowel slots filled. Flowable fill operation and dowel slot backfilling. Series of photographs. Flowable fill is poured into the empty base and leveled; after the panel is placed, the dowel slots are backfilled. The panel is shown with transverse joints sealed and dowel slots filled.

Construction Productivity

The construction productivity metrics include the documentation of time to complete the installation of one panel, list of possible concurrent activities, various equipment used for the installation of the panels, and panel installation crew size. The detailed panel installation activities include the following:

  • Slab (existing) demolition—A1.
    • Sawcutting of the repair boundaries.
    • Sawcutting of dowel slot outlines in the adjacent panels.
    • Removal of distresses panel.
  • Initial cleaning of the exposed base layer—A2.
    • Removal of debris.
    • Dewatering (if needed).
  • Cutting (jackhammering) of dowel slots to specification depth—A3.
  • Final cleaning and cleaning of the exposed base—A4.
  • Air cleaning and sandblasting of the dowel slots—A5.
  • Placement of the precast panel and final alignment—A6.
  • Drill holes for the high-density polyurethane foam and inject foam to stabilize and level the slab—A7.
  • Grout dowel slots and lift hook holes—A8.
  • Seal joints and open to traffic—A9.

Typical individual times, labor requirements, and equipment needed to execute the activities listed above are summarized in Table 5. Based on the proximity of the candidate panels to each other, construction activities A2–A7 can be performed somewhat concurrently, resulting in increased productivity.

Table 5. Typical Construction Time, Labor, and Equipment Needs
ACTIVITY CODE TIME, MINUTES RECOMMENDED EQUIPMENT (LABOR NEEDS)
A1 60 Concrete saw (1), front-end loader (1 operator)
A2 5 Nothing specific (2)
A3 20 Pneumatic jackhammers (2)
A4 15 Plate compactor (1)
A5 21 Sandblasting equipment (2)
A6 20 Front-end loader (1 operator, 3 additional to guide the alignment)
A7 25 Drills and high-density polyurethane injection equipment (2)
A8 26 Grout mixer (2)

Construction Activities–I-675

For the I-675 project, all nine panels were installed in 1 day under the same traffic control due to the close proximity of the panels. Figure 10 illustrates the relative location of the precast panels. Figure 11 illustrates the timeline (typical) for the installation of panel 5.

Figure 10. Relative distances of patches along the I-675 project.

Relative distances of patches along the I-675 project. Diagram. Panels 1–9 are shown, with distances between them, respectively, as follows: 189 ft. 7 in., 1,664 ft 3 in., 67 ft, 162 ft, 231 ft, 478 ft, 571 ft. 9 in. Panel 1 is a conventional full-depth repair; the others are precast panels. Panels 5 and 6 were stabilized using flowable fill.


Figure 11. Construction timeline for panel 5.

Construction timeline for panel 5. Bar chart. Performed consecutively, in minutes: slab removal ~ 5; dowel slots preparation ~ 40; base cleaning ~ 10; sandblasting ~ 10, followed by ~10 minutes inactivity, then, performed simultaneously, flowable fill placement and molding of cylinders, ~ 5, followed consecutively by leveling of fill, ~30, and slab placement, ~10. Total time is 120 minutes.

The total time required to install the panel was less than 120 minutes. The two most time-consuming activities were the preparation of the dowel slots and adjustment of the panel elevation with respect to the existing concrete pavement. The dowel slot-cutting time can be shortened by reducing the dimensions of the slots.

Patch 1 in this project had to be converted to a conventional cast-in-place, full-depth repair because the pavement at this location was superelevated and the precast panel dimensions were such that the resulting joint openings would have been unacceptable. In the future, such issues can be resolved if the contractor makes exact measurements of the panels to be replaced. This measurement data can then be used during the manufacturing of the panels. Therefore, this panel will serve as the control. The performance of the precast panels (2 through 9) will be compared with that of panel 1. Also, panels 5 and 6 were stabilized using flowable fill, whereas as the other precast panels were stabilized using the HDP foam. The impact of these two methods of slab stabilization on panel performance will be monitored and evaluated over the next 2 years as part of this study.

Construction Activities–M-25

At the M-25 project site, 12 precast panels were installed over a period of 3 days. Contributing to the longer installation time were the distances between some of the panels and the construction interruptions due to rain. Figure 12 illustrates the relative location of the precast panels.

Figure 12. Relative distances of patches along the M-25 project.

Relative distances of patches along the M-25 project. Patches are shown in northbound and southbound lanes: northbound, panels 1, 2, and 3 (stabilized using flowable fill) and 6, 8, and 9; southbound, panels 5, 4, 7, 10, 11, and 12. Distances between patches in mi beginning at 1 on the northbound lane and ending at 12 on the southbound lane: 1–5, 2.8; 5 to 4 and 2, 1 mi; 3 to 6 and 7, 1.8; 6–8, 0.10; 8 to 9 and 10, 0.75; 9 and 10 to 11, 0.02; and 11–12, 0.60 mi.


Figure 13 illustrates the contruction timeline for panels 2 and 3. Due to the close proximity of these panels, some of the construction activities overlap. Panels 2 and 3 were stabilized using flowable fill, and the remaining panels were stabilized using HDP foam.

Figure 13. Construction timeline for patches 2 and 3.

Construction timeline for patches 2 and 3. Bar chart. Time is shown for patch (P) 2, P3, and P2 and P3 simultaneously, by task: ~ 9:15–10:00 a.m., demolition of existing slab for P2; ~ 10.00–11:15 a.m., demolition for P2 and P3; ~11:15–11:25, first cleaning of exposed base layer, P2 and P3; 11:25–11:45, first cleaning of exposed base layer, P3; 11:30–11:50, cutting of dowel slots, 2 and 3; 12:00–12:15, final cleaning and cleaning of exposed base, 2 and 3; 12:15–12:25, place precast panel and align, P2; 12:25–12:35, place precast panel and align, P3; 12:35–50, drill holes and inject foam, P2 and P3; 12:25–12:40-1:20, grout slots and lift hook holes, P2 and P3.


Repair Effectiveness of the Precast Panels

The repair effectiveness of the panels was determined by conducting distress surveys and falling-weight deflectometer (FWD) tests according to the protocol illustrated in Figure 14. Test locations 1 through 8 allowed for the computation of load transfer efficiency (LTE) across the approach and leave joints. The LTE was computed using the following equation, where LTE:

LTE sub delta (%) equals delta sub u divided by delta sub L times 100

Figure 14. Falling-weight deflectometer test locations to evaluate panel effectiveness.

Falling-weight deflectometer test locations to evaluate panel effectiveness. Diagram. A slab and parts of the two adjacent slabs are shown, with four test locations in each wheelpath and a fifth location in the center of the slab. Two test locations are in each wheelpath on either side of the approach joint and two in each wheelpath on either side of the leave joint at equal distances from the fifth (center) location.


Field evaluations of the precast panels were conducted in October 2003, May 2004, October 2004, and May 2005.

Field Evaluation Results

I-675 Test Site

The FWD results are presented in Figure 15, in charts (a) through (c). The dashed lines in the charts represent the minimum LTE threshold of 70 percent and a deflection ratio threshold of 3 for doweled joints. On average, the approach and leave LTEs are in excess of 70 percent (the average LTEs range from 61 percent to 90 percent) as shown in Figure 15 (a) and (b) and in Table 6. Figure 15 (c) represents the relative deflections (peak) of the joint with respect to the mid-slab deflection (peak). The approach joint deflection ratios range from 1.2 to 2.1, whereas, the leave joint ratios range from 1.3 to 2.3, indicating an acceptable support under the panel.

Table 7 displays and summarizes the condition of the precast panels as of September 2005. The performance evaluation was conducted by MDOT. Panel 2 exhibited premature cracking along the leave joint; the remaining seven panels exhibited acceptable behavior at the time of the last performance evaluation.

For panel 2, the average before-leave joint LTE (%) ranges from 63 to 75 whereas the after-leave joint LTE (%) ranges from 68 to 91. The type of distress observed in panel 2 looks like failure at the end of the dowels due to thermal movement of the pavement. The hot pour joint sealant is pushed up in the joint, indicating the pavement closing up on the leave side of the joint. Uniform restraint was not provided in the reservoir opening between the dowels, which may have resulted in high bearing stresses at the dowel ends. Another reason for this premature joint failure could be that the dowels along the leave joint were horizontally skewed as a result of the installation. This skewing could have resulted in "locking" of the joint and impeding joint movement under environmental and traffic loads.

Figure 15. Average load transfer efficiencies and deflection ratios for the I-675 test site.

a) Leave slab loading.

Average load transfer efficiencies (LTEs) and deflection ratios for the I-675 test site. Three bar charts show average LTEs for leave slab loading, approach slab loading, and deflection ratio for 9 panels, on inner and outer wheelpaths, before and after joint. For leave slab loading, the panels reach, and many exceed, the 70 percent threshold on all four measurements except panels 2 and 8. The ranges around the averages vary from 5 to 10 percentage points, with wider ranges where the threshold was not met. For approach slab loading, most measurements exceed the threshold. Panels 2 and 8 fall short. Ranges vary from 5 to 20 percent. For deflection ratios, only panel 1 exceeds the threshold of 3.0; the others fall between 1.0 and 2.0.

b) Approach slab loading.

Average load transfer efficiencies (LTEs) and deflection ratios for the I-675 test site. Three bar charts show average LTEs for leave slab loading, approach slab loading, and deflection ratio for 9 panels, on inner and outer wheelpaths, before and after joint. For leave slab loading, the panels reach, and many exceed, the 70 percent threshold on all four measurements except panels 2 and 8. The ranges around the averages vary from 5 to 10 percentage points, with wider ranges where the threshold was not met. For approach slab loading, most measurements exceed the threshold. Panels 2 and 8 fall short. Ranges vary from 5 to 20 percent. For deflection ratios, only panel 1 exceeds the threshold of 3.0; the others fall between 1.0 and 2.0.

c) Deflection ratio.

Average load transfer efficiencies (LTEs) and deflection ratios for the I-675 test site. Three bar charts show average LTEs for leave slab loading, approach slab loading, and deflection ratio for 9 panels, on inner and outer wheelpaths, before and after joint. For leave slab loading, the panels reach, and many exceed, the 70 percent threshold on all four measurements except panels 2 and 8. The ranges around the averages vary from 5 to 10 percentage points, with wider ranges where the threshold was not met. For approach slab loading, most measurements exceed the threshold. Panels 2 and 8 fall short. Ranges vary from 5 to 20 percent. For deflection ratios, only panel 1 exceeds the threshold of 3.0; the others fall between 1.0 and 2.0.


Table 6. Summary of Approach and Leave Joint Load Transfer Efficiencies (lte), I-675
PANEL NUMBER AVERAGE APPROACH JOINT LTE (%) AVERAGE LEAVE JOINT LTE (%)
1 89.1 81.6
2 74.5 74.7
3 78.6 74.4
4 78.0 82.0
5 72.0 80.5
6 72.0 61.3
7 78.6 78.0
8 83.4 72.5
9 84.3 77.0

Table 7. Summary of Panel Performance, I-675 (Survey Date: 9/14/05)
PRECAST PANEL ID PERFORMANCE DESCRIPTION
Panel 1 There is no evidence of distress. The average LTE ranges from 81% to 89%, and the deflection ratios are less than 2.5.
Panel 1 as discussed to the left
Panel 2 Panel cracking observed in the vicinity of the dowel bars. The average joint LTE is 74.5%, and the deflection ratios are less than 2.
Panel 2 as discussed to the left
Panel 3 Panel in good condition. Some spalling was observed along the joint. The average joint LTE ranges between 75% and 78%, and the deflection ratios are less than 2.
Panel 3 as discussed to the left
Panel 4 The panel is in good condition. The average joint LTE ranges between 78% to 82%, and the deflection ratios are less than 2.
Panel 4 as discussed to the left
Panel 5 The panel is in good condition. The average joint LTE ranges between 73% to 80%, and the deflection ratios are less than 2. The panel was stabilized using flowable fill.
Panel 5 as discussed to the left
Panel 6 Panel is in acceptable condition. Some spalling of the dowel slot back-fill was observed. The average joint LTE ranges between 61% and 70%, and the deflection ratios are less than 2. The panel was stabilized using flowable fill.
Panel 6 as discussed to the left
Panel 7 The panel is in good condition. A high spot was observed at the leading edge of the repair. The average joint LTE is 78%, and the deflection ratios are less than 2.
Panel 7 as discussed to the left
Panel 8 The panel is in good condition. The average joint LTE ranges from 73% to 80%, and the deflection ratios are less than 2.
Panel 8 as discussed to the left
Panel 9 The panel is in good condition. A high spot was observed at one end of the repair. The average joint LTE ranges from 77% to 84%, and the deflection ratios are less than 2.
Panel 9 as discussed to the left

M-25 Test Site

The FWD results are presented in Figure 16 (a) through (c). The dashed lines represent the minimum LTE threshold of 70 percent and a deflection ratio threshold of 3 for doweled joints. On average, the approach and leave LTEs are in excess of 70 percent (the average LTEs range from 72 percent to 90 percent) as shown in Figure 16 (a) and (b) and Table 8. Figure 16 (c) represents the relative deflections (peak) of the joint with respect to the mid-slab deflection (peak). The approach joint deflection ratios range from 1.2 to 2.3, whereas the leave joint ratios range from 1.0 to 3.0, indicating an acceptable support under the panel.

Figure 16. Average load transfer efficiencies and deflection ratios for the M-25 test site.

a) Leave slab loading.

Average load transfer efficiencies and deflection ratios for the M-25 test site. Three bar charts show average LTEs for leave slab loading, approach slab loading, and deflection ratio for 12 panels, for inner and outer wheelpaths, before and after joints. For leave slab loading, all panels exceed the 70 percent threshold on all four measurements except panels 4, 9, and 12. Ranges around the average vary from 10 to 40 percentage points. For approach slab loading, most measurements exceed the threshold. Panel 4 falls short on the outer wheelpath, before-joint measure. Ranges vary from 5 to 30 percent. For deflection ratios, no averages exceed the threshold of 3.0, falling between 1.5 and 3.0.

b) Approach slab loading.

Average load transfer efficiencies and deflection ratios for the M-25 test site. Three bar charts show average LTEs for leave slab loading, approach slab loading, and deflection ratio for 12 panels, for inner and outer wheelpaths, before and after joints. For leave slab loading, all panels exceed the 70 percent threshold on all four measurements except panels 4, 9, and 12. Ranges around the average vary from 10 to 40 percentage points. For approach slab loading, most measurements exceed the threshold. Panel 4 falls short on the outer wheelpath, before-joint measure. Ranges vary from 5 to 30 percent. For deflection ratios, no averages exceed the threshold of 3.0, falling between 1.5 and 3.0.

c) Deflection ratio.

Average load transfer efficiencies and deflection ratios for the M-25 test site. Three bar charts show average LTEs for leave slab loading, approach slab loading, and deflection ratio for 12 panels, for inner and outer wheelpaths, before and after joints. For leave slab loading, all panels exceed the 70 percent threshold on all four measurements except panels 4, 9, and 12. Ranges around the average vary from 10 to 40 percentage points. For approach slab loading, most measurements exceed the threshold. Panel 4 falls short on the outer wheelpath, before-joint measure. Ranges vary from 5 to 30 percent. For deflection ratios, no averages exceed the threshold of 3.0, falling between 1.5 and 3.0.


Table 8. Summary of Approach and Leave Joint Load Transfer Efficiencies (LTE)
PANEL NUMBER AVERAGE APPROACH JOINT LTE (%) AVERAGE LEAVE JOINT LTE (%)
1 85.0 82.0
2 90.2 88.3
3 88.0 82.2
4 78.5 73.8
5 87.8 82.4
6 87.9 87.9
7 87.5 85.8
8 86.9 87.8
9 89.1 72.6
10 84.6 89.7
11 89.4 90.3
12 88.0 62.0

Table 9 summarizes (by way of photographs) the condition of the precast panels in September 2005. The performance evaluation was conducted by MDOT. Of the 12 panels, 10 exhibited acceptable to good behavior at the time of the last performance evaluation.

Figure 17 summarizes the structural response of the leave and approach joints for panels 4 and 9. It is evident that for both panels there was a significant drop in leave joint efficiencies from October 2003 and May 2004. The plot also shows that there is also a significant loss in relative support along the distressed joint during the same time frame. During the summer of 2004 the thumb area of Michigan experienced a series of 32 °C (90 °F) days that may have resulted in "abnormal" expansion of the pavement slabs. Such expansion could have resulted in a joint blowout. A possible contributor to joint blowout is horizontal misalignment of dowel bars. If such misalignment occurred during installation of the panels, the joint may not have been flexible enough to accommodate slab expansion caused by the high ambient temperatures. Figure 18 illustrates how dowel misalignment could reduce joint flexibility.


Figure 17. Structural responce of panels 4 and 9.

a) Leave joint.

Structural response of panels 4 and 9. Graphs depict leave joint and approach joint responses for panels 4, 9, and "others" as measured by falling-weight deflectometer testing on three occasions: October 2003, May 2004, and October 2004 (x axis). The leave joint data on load transfer efficiency (LTE, 0 to 100 in increments of 12.5 percent; y axis) for panel 4 on the three occasions, respectively, were approximately 85, 87, and 49; for panel 9, 88, 90, and 50; for "others," 86, 83, and 78. The leave joint data deflection ratios (0.0–4.0 in increments of 0.5, y axis) for panel 4 on the three occasions, respectively, were approximately 1.4, 1.7, and 2.5; for panel 9, 1.9, 2.3, and 3.5; for "others," 1.8, 1.7, and 2.0. The approach joint data on LTE for panel 4 were approximately 85, 85, and 66; for panel 9, 88, 88, and 87; for "others," 87, 86, and 73. The approach joint data deflection ratios for panel 4 were approximately 1.5, 1.6, and 1.7; for panel 9, 2.2, 2.4, and 1.9; for "others," 1.7, 1.9, and 1.9.

b) Approach joint.

Structural response of panels 4 and 9. Graphs depict leave joint and approach joint responses for panels 4, 9, and "others" as measured by falling-weight deflectometer testing on three occasions: October 2003, May 2004, and October 2004 (x axis). The leave joint data on load transfer efficiency (LTE, 0 to 100 in increments of 12.5 percent; y axis) for panel 4 on the three occasions, respectively, were approximately 85, 87, and 49; for panel 9, 88, 90, and 50; for "others," 86, 83, and 78. The leave joint data deflection ratios (0.0–4.0 in increments of 0.5, y axis) for panel 4 on the three occasions, respectively, were approximately 1.4, 1.7, and 2.5; for panel 9, 1.9, 2.3, and 3.5; for "others," 1.8, 1.7, and 2.0. The approach joint data on LTE for panel 4 were approximately 85, 85, and 66; for panel 9, 88, 88, and 87; for "others," 87, 86, and 73. The approach joint data deflection ratios for panel 4 were approximately 1.5, 1.6, and 1.7; for panel 9, 2.2, 2.4, and 1.9; for "others," 1.7, 1.9, and 1.9.


Figure 18. Line sketch illustrating the possible dowel skew and uneven joint opening.

Line sketch illustrating the possible dowel skew and uneven joint opening. A precast panel with two sets of dowel bars on either side is shown. The dowels are skewed horizontally, shifting the panel and creating uneven joint openings on both sides.


The finite element model EverFE (Davids 2003) was used to approximate the effects of horizontal skew and high ambient temperatures on the tensile stresses on the precast concrete panel and vertical shear stresses in the dowel bars. Three variables were included in this limited analysis:

  • Dowel horizontal skew angles: 0, 3, 6, and 10 degrees.
  • Temperature differential (maximum ambient–construction): 2°C (35°F) and 16°C (60°F).
  • Axial friction force: 0, 45, 445, 4,448, 44,482 kN (0, 10,000, 100,000, 1,000,000, 10,000,000 lbf).

Figure 19 summarizes the findings of the analysis.


Figure 19. Results of the finite element analysis.

a) Effect of temperature differential on concrete tensile stresses.

Results of the finite element analysis. Two graphs. The first shows the effect of temperature differential on concrete tensile stresses, with maximum principal stress in psi (0 to 140) (y axis) plotted over axial friction in kips (0 to 12,000) (x axis). For 35 °F, 4 curves (0-degree, 3-degree, 6-degree, and 10-degree) rise from about 50 psi to 80 psi over 0 to 1,000 kips, then level and remain around 80 psi to the ending point of 10,000 kips. For 60 °F, 4 curves (0-degree, 3-degree, 6-degree, and 10-degree) rise from 60–65 to about 118 psi over 0 to 1,000 kips, then level and remain at 120 psi between 1,000 and 10,000 kips; the 10-degree curve continues to rise slightly, ending at about 125 psi. The second graph shows the effect of horizontal dowel skewness on dowel vertical shear force. Maximum shear, 0 to 1.4 kips (y axis) in increments of 0.2, is plotted over axial friction, 0 to 12,000 in increments of 2,000 (x axis).

b) Effect of horizontal dowel skewness on dowel vertical shear force.

Results of the finite element analysis. Two graphs. The first shows the effect of temperature differential on concrete tensile stresses, with maximum principal stress in psi (0 to 140) (y axis) plotted over axial friction in kips (0 to 12,000) (x axis). For 35 °F, 4 curves (0-degree, 3-degree, 6-degree, and 10-degree) rise from about 50 psi to 80 psi over 0 to 1,000 kips, then level and remain around 80 psi to the ending point of 10,000 kips. For 60 °F, 4 curves (0-degree, 3-degree, 6-degree, and 10-degree) rise from 60–65 to about 118 psi over 0 to 1,000 kips, then level and remain at 120 psi between 1,000 and 10,000 kips; the 10-degree curve continues to rise slightly, ending at about 125 psi. The second graph shows the effect of horizontal dowel skewness on dowel vertical shear force. Maximum shear, 0 to 1.4 kips (y axis) in increments of 0.2, is plotted over axial friction, 0 to 12,000 in increments of 2,000 (x axis).

Table 9. Summary of Panel Performance, M-25 (Survey Date: 9/14/05)
PRECAST PANEL ID PERFORMANCE DESCRIPTION
Panel 1 The panel is in good condition. The average joint LTE ranges from 82% to 85% and the deflection ratios are less than 2.
Panel 1 as discussed to the left
Panel 2 The panel is in good condition. The average joint LTE ranges from 88% to 92% and the deflection ratios are less than 2. The panel was stabilized using flowable fill.
Panel 2 as discussed to the left
Panel 3 The panel is in good condition. The average joint LTE ranges from 82% to 88% and the deflection ratios are less than 2. The panel was stabilized using flowable fill.
Panel 3 as discussed to the left
Panel 4 The repair is severely deteriorated. The cracking along the dowels was first observed during the summer of 2004. The cracking was observed along the approach and leave joint of the panel. The average joint LTE ranges from 73% to 78% and the deflection ratios are less than 2.
Panel 4 as discussed to the left
Panel 5 The panel is in good condition. The average joint LTE ranges from 82% to 88% and the deflection ratios are less than 2.
Panel 5 as discussed to the left
Panel 6 The panel is in good condition. The average joint LTE is 88% and the deflection ratios are less than 2.5.
Panel 6 as discussed to the left
Panel 7 The panel is broken in one corner most probably due to snowplow damage, as there was a high spot along the approach joint. The average joint LTE ranges from 86% to 88% and the deflection ratios are less than 2.
Panel 7 as discussed to the left
Panel 8 The panel is in good condition. The average joint LTE is 87% and the deflection ratios are less than 2.
Panel 8 as discussed to the left
Panel 9 The repair is deteriorated. The cracking along the dowels was first observed during the summer of 2004. The cracking is along the leave joint of the panel. The average joint LTE ranges from 72% (leave joint of the panel) to 89% and the deflection ratios are approaching 3 along the leave joint of the panel.
Panel 9 as discussed to the left
Panel 10 The panel is in good condition. The average joint LTE ranges from 85% to 90% and the deflection ratios are less than 2.5.
Panel 10 as discussed to the left
Panel 11 The panel is in good condition. The average joint LTE is 90% and the deflection ratios are less than 2.
Panel 11 as discussed to the left
Panel 12 The panel is in acceptable condition. Trailing edge (leave joint of the panel) broken in the old concrete. The average joint LTE ranges from 62% (leave joint) to 88% and the deflection ratios are less than 2.5. However, the deflection ratio at the broken edge is approaching 3.
Panel 12 as discussed to the left

Recommendations for Precast Panel Installation

Based on the experience of the Michigan field trials, the following practices are recommended for future precast panel installations for full-depth repair of jointed concrete pavements:

  1. Provide an expansion cap at one end of the dowel bar to accommodate slab movement due to environmental loading and to prevent closing of the joint.
  2. Provide expansion material along the joint to accommodate joint movement due to thermal expansion and contraction.
  3. Diamond grinding of the panel, especially the joints, is recommended to remove high spots. These high spots have a potential to "break off" as a result of snowplowing operations.
  4. The width of the dowel slot should be kept as small as possible to accommodate the dowel bar. This will reduce construction time and also reduce the potential for dowel skewing in the horizontal plane.
  5. Care needs to be given to sawing the existing concrete for the outline of the patch. The saw cuts should be perpendicular to the centerline to avoid skewing the patch when it is placed (which leads to problems with the dowel bars locking in the slots).
  6. To reduce construction time, multitasking during the installation process should be encouraged whenever possible. Construction and installation time can be positively impacted if the repair locations are close to each other, allowing for the installation of multiple panels under one traffic control setup.
<<PreviousContentsNext >>

FHWA
United States Department of Transportation - Federal Highway Administration