Construction of a Precast Prestressed Concrete Pavement Demonstration Project on Interstate 57 Near Sikeston, Missouri
Chapter 8. Project Evaluation and Recommendations for Future Projects
The I-57 demonstration project allowed MoDOT and local contractors to evaluate PPCP technology as a new tool for rapid pavement construction. This initial project, constructed in a rural area on a tangent section of the highway with a relatively simple geometry, allowed for the details of the construction technique to be worked out on a smaller scale project prior to incorporation into larger, time-sensitive projects. Because this was not a time-critical project with restrictions on lane closures, no attempt was made to address issues such as traffic staging and lane-by-lane construction in the development of the project layout. The use of full-width precast panels for future projects will only be possible if a full closure of at least one side of the highway is possible. Had horizontal and vertical curves or superelevations been included, a detailed survey of the project site would have been required so that precast panels could be designed to fit the roadway geometry.
The design procedures and details for the I-57 demonstration project were based primarily on experience from the demonstration projects in Texas and California.(2,3,4) A few different design details were implemented on this project, however, that should also be considered for future projects.
The design procedure essentially provides a precast pavement section with an equivalent design life to a thicker, conventional, cast-in-place (non-prestressed) pavement. This procedure analyzes stress conditions over the life of the pavement to ensure that they do not exceed those of an equivalent conventional pavement. Prestress levels are adjusted to meet this condition by increasing or decreasing the spacing of the longitudinal prestressing tendons. This procedure is believed to be very conservative in that it analyzes extreme stress conditions, which may only occur a few times each year over the life of the pavement. The prestress requirements produced by this procedure, however, are not unreasonable in terms of fabrication (pretensioning) and on-site construction (post-tensioning) requirements. While performance data are limited, the performance of existing PPCP projects indicates that the design of the PPCP projects to date has been adequate. Due to limited availability of the PSCP2 computer program, simplified design procedures or catalogs will likely need to be developed for State highway agencies to use as PPCP technology becomes more widely used.
Transverse prestress requirements were determined based on stresses from lifting and handling of the precast panels to ensure that cracking did not occur during handling. For future projects where smaller precast panels are used, and lifting and handling stresses are not as critical, in-service stresses from traffic and environmental effects, particularly slab curing, should be considered in a similar fashion as longitudinal stresses.
The design details for this demonstration project demonstrated different alternatives for the precast panels and assembly process, and overall, no major problems were encountered.
Pavement Cross Section-A single precast panel with a variable thickness was used to provide full-width pavement (including inside and outside shoulders) with a "rooftop" crown. The full-width panels provided an efficient solution in terms of the number of panels to be fabricated and installed, but may not be appropriate for projects where only one lane can be reconstructed at a time. Separate precast panels may be required for each lane on projects where only one lane is reconstructed at a time or for projects (such as unbonded overlays) where it is necessary to set the panels at the appropriate cross slope and not on a level base.
One of the disadvantages of the panels used for this project is that the cross section is essentially unique to this project, and therefore the precast panel formwork may not be useful for future projects which do not have the same cross section. For future large projects, however, the capital investment in new panel formwork will likely only be a minor percentage of the overall project cost. It may also be possible to develop standard cross sections or panel formwork that can be used for numerous projects.
Keyways-The variable-thickness precast panels precluded the use of continuous keyways across the full width of the precast panels. This resulted in some differential elevation or "faulted" joints between a few of the panels in the shoulder regions where there was only a butt joint between the panels. While this did not cause any problems with the driving surface of the pavement, future projects where similar keyways must be used may consider the use of dowels or pins to help ensure vertical alignment in regions without keyways. These pins would only be needed during panel assembly, as the longitudinal post-tensioning and epoxied joints between panels provide load transfer between panels.
Expansion Joints-This project demonstrated the viability of header-type expansion joints for PPCP. Header-type dowelled expansion joints were specified to permit diamond grinding over the expansion joints while also providing durable joints that could accommodate approximately 50 mm (2 in.) of movement. Four of the five header joints are performing well, with the poor performance of the fifth joint attributed to problems with the two halves of the joint panel bonding together and fracturing away from the actual joint opening. The joint sealant protrudes from the surface of the pavement under summer climatic conditions, but this could be corrected by ensuring that the joint width is set correctly during panel installation.
Long-term monitoring of the joint performance will provide an indication of the durability of this type of expansion joint for PPCP. Currently, header-type joints are used extensively for bridge decks and provide good performance, as they should for PPCP as well. This type of joint does require special attention to installation of the header material and seal, however, and would even permit installation of the header material at the fabrication plant if necessary. This type of expansion joint should be considered for future projects, particularly where diamond grinding is anticipated, but should still be weighed against the durability advantages of armored joints for expansion joints that are expected to open more than 25-50 mm (1-2 in.).
End Stressing-End stressing (as opposed to central stressing) was used for this project to eliminate additional central stressing panels. This required careful detailing of the anchor region in the joint panels to ensure that the stressing pockets were large enough to accommodate the stressing rams and to ensure that the prestress force would be transferred from the post-tensioning anchors back to the expansion joint. The primary disadvantage was that custom-sized pocket formers were required for each stressing pocket due to the variable thickness of the precast panels, increasing the cost of the formwork. No problems were experienced with stressing the tendons from both ends on site.
End stressing was demonstrated to be a viable alternative to central stressing. From a design standpoint, central stressing is preferable to end stressing, as the effective tendon length and associated prestress losses are lower. It also requires special attention to detailing the anchorage region, including the anchor access blockouts and reinforcement around the anchors. From a construction standpoint, end stressing eliminates the need for the central stressing panels, potentially reducing the cost of panel fabrication, but adds an additional stressing operation if the tendons are stressed from both ends. Both end stressing and central stressing should be considered for future projects, and the alternative that best fits the project constraints selected.
Overall, no major problems were encountered during the fabrication process. There were, however, some distresses observed in the precast panels at the fabrication plant. Minor changes to the precast panel details and fabrication requirements should eliminate these issues on future projects.
Tolerances-Tolerances were based on those developed from experience with previous PPCP demonstration projects. No problems were reported by MoDOT inspectors with achieving the tolerances during the panel fabrication process. The use of steel sideforms to form the keyways on this and previous projects helped to ensure that the panels would fit together properly.
Casting Bed-The "long line" casting process, where two panels were cast end to end, proved to be an efficient process for this and previous demonstration projects. While a longer casting bed would have permitted more panels to be cast each day, the cost of the custom sideforms for a longer casting bed may have been cost prohibitive, and the relatively small number of panels to be fabricated did not necessitate a longer bed. Panels with a flat bottom and variable thickness were a key design element that greatly simplified the formwork for these panels.
Finishing-The initial carpet drag finish was changed to a light broom finish due to problems applying the carpet drag that resulted in aggregate "overturning." The light broom finish provided adequate texture as a "temporary" finish prior to diamond grinding. Scaling was observed on some of the precast panels, likely due to over-finishing of the surface. Any scaling in the traffic lanes was removed by diamond grinding.
Thermal Stresses-The cracks along the long axis of the precast panels observed at the fabrication plant and on site are believed to be caused at least in part by "thermal shock" as the panels were exposed to cool ambient temperatures shortly after they were steam cured. Significant strain levels measured by the University of Missouri in the precast panels during the fabrication process indicate that high thermal stresses were likely experienced by the panels. It is important to note again, however, that the fabrication procedures used were standard established practices, and the cracking that occurred was not caused by poor fabrication practices.
To mitigate potential thermal cracking in future projects, two primary measures are recommended:
- Establishment of a "cooling" procedure for the panel fabrication process if they are steam cured under cold climatic conditions. No specific recommendation for the cooling procedure is made herein, but one should be developed with the precast supplier.
- Inclusion of temperature reinforcement in the precast panels in the transverse direction (non-pretensioned direction). This additional reinforcement will not prevent thermal cracking, but will help hold any cracks that do form tightly closed. The ACI 318-95 Building Code recommends a minimum of 0.18 percent reinforcement for "temperature steel."(14)
No major problems prevented successful construction of the PPCP section. However, there were some issues encountered during construction where improvements could be made for future projects.
The permeable asphalt-treated base proved to be a viable material for the prepared base beneath the precast panels. Stringline grade control was used for construction of the permeable asphalt-treated base, and no problems in achieving the base surface tolerance were noted by MoDOT inspectors or the contractor. Similar tolerances should be considered for future projects.
No major problems were experienced with installation of the precast panels. Some of the issues which should be addressed for future projects, however, are summarized below.
Staging/Crane Location-"Rutting" of the permeable asphalt-treated base under the weight of the crane was one of the primary issues observed during panel installation. Ideally, the crane should be located off of the base that will be supporting the precast panels, particularly if "softer" aggregate or asphalt bases are used. If it is not possible to keep the crane off of the prepared base, contact pressure under the crane should be determined prior to construction and additional measures should be taken to distribute the weight of the crane if necessary.
Panel Alignment-Deviation of the centerline of the precast panels from the true centerline of the roadway was the primary issue of concern for panel installation. An uneven gap left between two panels initiated this deviation, which was further accumulated as panel installation continued. To correct the alignment, panels were offset and shims were installed at some of the joints between panels. Although the construction report from the Texas Demonstration Project(2) recommended the use of shims (no greater than 3 mm [1/8-in.] thick) to correct deviation of alignment, strain measurements during post-tensioning of this project revealed that shims cause a nonuniform distribution of strains across the width of the pavement. Based on this experience, offsetting precast panels is recommended for correction of centerline alignment whenever possible (installing precast panels next to an existing pavement may preclude offsetting of the panels).
To prevent problems with post-tensioning duct alignment when the panels are offset, the use of larger diameter or, alternatively, flat, multistrand ducts that can accommodate offsetting are recommended. A maximum offset, based on the dimensions of the ducts used, should be specified in the contract documents prior to construction. Also, as recommended from previous demonstration projects, a mark on the surface of the precast panels directly above a designated post-tensioning duct should be used to align the precast panels as they are installed.
All of the post-tensioning tendons were successfully stressed. There were, however, some issues with installing the post-tensioning tendons and timing of the post-tensioning operation which should be addressed for future projects.
Post-Tensioning Tendon Installation-The primary causes of difficulty in feeding the post-tensioning strands through the ducts were misalignment of the ducts due to offsetting of the precast panels and obstructions in the ducts. A mechanical strand pusher was used to feed the strands through the panels, and is recommended for use on future projects as well, particularly when long tendons are used. Offsetting of the panels to correct the alignment of the pavement caused difficulty in feeding the strands, but can be mitigated on future projects by using larger diameter or flat, multistrand ducts.
Additionally, although the precast producer carefully checked and (temporarily) plugged the post-tensioning ducts at the fabrication plant, ice formed in some of the ducts and caused problems with feeding the post-tensioning strands when temperatures were well below freezing. Ensuring that the ducts are clear of water, particularly in colder climatic conditions, will help mitigate problems with ice formation in the ducts. Using compressed air to blow any water or other debris out of the ducts before plugging them at the fabrication plant or before installing them on site may also prevent problems with tendon installation.
Timing of Post-Tensioning-The original construction sequencing called for a section of panels to be installed and post-tensioned each day of construction. Unfortunately, it was not possible to get enough delivery trucks to ship an entire (e.g., 76-m [250 ft]) section of panels. Because of this, only partial sections were installed, and the epoxy applied to the panel joints served to bond these partial sections together prior to post-tensioning. This essentially created long, non-post-tensioned sections of pavement. Additionally, final post-tensioning was not completed until all four sections of precast panels had been installed, leaving long, non-post-tensioned sections of pavement for several days until post-tensioning could be completed. While not observed directly, it is believed that this may have at lease exacerbated transverse cracking in the pavement.
For future projects, it is essential that longitudinal post-tensioning be applied in a timely manner after panel installation. Ideally, final post-tensioning should be applied before final set of the epoxy used to bond the panels together. If this is not possible due to problems with panel delivery or installation, the temporary post-tensioning strands should be stressed and "locked off" at the end of a day's placement to provide a clamping force as the epoxy sets and to reduce the possibility of any transverse cracking. The temporary strands can then be de-tensioned prior to installing the remaining precast panels. As an alternative, high-strength post-tensioning bars can be specified for the longitudinal post-tensioning tendons. These bars can be cut to the length of individual panels and coupled together for "segmental" post-tensioning as the panels are installed. Bars will likely increase the cost of the post-tensioning component, but may be necessary for projects where it is not possible to construct full sections of panels in a continuous operation.
Post-Tensioning Ducts-As discussed previously, offsetting of the precast panels made feeding the post-tensioning strands difficult for some of the tendons. An alternative to help mitigate the effects of offsetting panels for future projects would be the use of either larger diameter (e.g., 50-mm [2-in.] inside diameter) or flat, multistrand (25 mm by 75 mm [1 in. by 3 in.]) ducts that could accommodate offsetting of up to 50 mm (2 in.).
Joint Seal-Although MoDOT inspectors and the post-tensioning contractor were confident that every tendon was fully grouted, it often required pumping grout into more than one vent along each tendon. Grout leakage was the result of a poor seal between segments of the ducts across the joints between precast panels. The combination of the compressible foam gaskets and epoxy along the joints between panels was beneficial, but did not guarantee leak-free joints. The epoxy and gaskets are recommended for future projects, but it may be necessary to develop a positive connection or coupler to connect duct segments between panels. This connector/coupler may add cost and complexity to the panel installation process, but would help eliminate grout leakage and the time and expense associated with it. The only known duct couplers currently available require physically fitting the coupler around the duct segments after they are in place.
The approximate final cost for the Missouri Demonstration Project was $1,057,500. This cost included panel fabrication, delivery, and installation costs. It did not include the cost of base preparation, which was required regardless of the type of pavement installed. The total translates to a unit cost of approximately $297/m² ($248/yd2) of finished pavement. Of this cost, approximately 48 percent was for the precast panels delivered to the job site. The remainder of the cost was for panel installation (including labor, crane rental, epoxy, polyethylene sheeting, etc.), post-tensioning, grouting, and traffic control costs.
The unit cost of this project was slightly higher than the cost of the California Demonstration Project ($268/m² [$224/yd2]), and significantly higher than that of the Texas Demonstration Project ($194/m² [$162/yd2]). However, many factors contribute to project cost:
- Project size.
- Project location (for contractor mobilization).
- Project distance from the precast fabrication plant.
- Local labor costs (particularly if union labor is required).
- Complexity of panel fabrication (including specialty formwork and fabrication procedures).
- Complexity of panel installation, post-tensioning, and grouting (equipment and labor requirements).
- Date of construction (variability of material, equipment, and labor costs), among others.
As such, it is difficult to accurately compare projects constructed in different States at different times, particularly when the projects are different in size and scope. However, this project provides what could be considered typical costs for a rural project constructed in the Midwestern United States. Because the project was competitively bid (even if as only a small portion of a much larger project), the costs (panel fabrication costs in particular), are likely representative of what could be expected for PPCP projects, at least during initial implementation of this technology. It should be recognized that this project was still relatively small in size, and the first of its kind for the State of Missouri. As such, higher costs are not unexpected. There will likely be economies of scale with larger projects that will reduce the unit cost, and as contractors become more familiar with the technology, fabrication and installation costs will also likely decrease.