The Minnesota Department of Transportation (MnDOT) developed a design for a precast composite slab span system (PCSSS) to be used in accelerated bridge construction. The system consists of shallow inverted-tee precast beams placed between supports with cast-in-place (CIP) concrete placed on top, forming a composite slab span system. Suitable for spans between 20 and 60 ft., the MnDOT PCSSS is useful for replacing a large number of aging conventional slab-span bridges throughout the United States highway system. Originally developed in 2005, the PCSSS had three distinct design generations in the 12 bridges that were constructed by MnDOT between 2005 and 2011. The objective of this investigation was to evaluate the field performance of a sample of the existing bridges through detailed crack mapping and core analysis and through continued monitoring of data obtained from one of the original PCSSS bridges (Bridge No. 13004) instrumented during construction in 2005. A parametric design study was also conducted to investigate the effects of continuity design on the economy of the PCSSS.
Five of the 12 PCSSS bridges, constructed between 2005 and 2011, were selected as the sample set to conduct detailed surveys of surface cracking and examinations of extracted core specimens to evaluate effects of the design changes. Surface cracking was recorded over three different inspections between the fall of 2009 and the summer of 2011. Each inspection was done using a systematic procedure of documenting crack locations and measuring crack widths. The result was a series of crack maps for each bridge, showing the surface cracking compared to major design features. Different line types were used to distinguish relative crack widths. Core specimens were taken from each of the five inspected bridges based on anticipated reflective crack locations. The cores were partial depth through the CIP concrete, taken over either the longitudinal joint between precast panels or over the precast web corner. Each core was examined under a digital microscope for cracking with particular attention paid to the regions above the longitudinal joints and web corners. The results of the core investigation were compared to the corresponding crack maps. Overall, the field inspections indicated that the changes made between each design generation improved the performance of the PCSSS. Bridge No. 13004 in Center City, MN from the first design generation showed many short, longitudinal cracks on the deck surface with very little transverse and map cracking. The longitudinal cracks were located primarily over the precast beam web, corresponding to what appeared to be insufficient consolidation of the CIP concrete around the stirrups projecting vertically from the section to facilitate composite action, which had little clearance above the precast webs. In the second generation, more clearance was provided under the stirrups projecting from the surface. Bridge Nos. 33005 and 33008 near Mora, MN from the second generation did not show the short cracks over the webs from the first generation, but more transverse cracks and longer longitudinal cracks were observed.
Bridge No. 33008 showed significantly more longitudinal cracking than any of the other bridges. Significant longitudinal cracks were noted along several joints between the precast beams. Core specimens showed that these cracks were full-depth reflective cracks. The only other bridge to show reflective cracking from the core specimens was Bridge No. 13004, but these were not full-depth cracks. It was unclear from the design details of Bridge No. 33008 why it was in worse condition than the other bridges. This bridge also had noticeably different cambers between adjacent beams observed from the underside of the bridge, although it was unclear how this might be associated with the observed longitudinal cracking.
For the third design generation, the thickness of the precast beam flanges was decreased and the trough reinforcement spacing (consisting of trough hooks projecting horizontally from the beams across the joint, as well as a drop-in cage) was decreased from a maximum 10 in. center-to-center to 6 in. center-to-center to better control reflective cracking. The decreased spacing was accomplished by staggering the trough hooks from adjacent precast beams. Bridge Nos. 49007 and 49036 near Little Falls, MN from the third generation did not exhibit longitudinal cracking over the precast beam joints, indicating that the design changes may have had a positive impact, though not conclusively. The most significant issue observed with the third generation was shrinkage cracking, indicated by longitudinal cracks located over the precast beam webs and more extensive transverse and map cracking. Generally, bridges with a larger length to width aspect ratio (i.e., L/W) had more transverse cracking, which could be related to more longitudinal shrinkage restraint.
In addition to the field inspections, strain data from the instrumentation of Bridge No. 13004 was analyzed to evaluate performance. The bridge was instrumented in 2005 to monitor reflective cracking and continuous system behavior. Six years of strain and temperature data showed a progression of reflective cracking in several locations and significant cracking due to thermally induced restraint moments. The reflective cracking from the strain data was confirmed by observed cracks in the core specimens near the locations of the strain gauges. While the width of the reflective cracks appeared to increase over time from the strain measurements, the measurements began to plateau by the end of the six-year monitoring period. Restraint moment cracking was indicated by strain gauges attached to continuity connection reinforcement. The measured restraint moment strains were large enough for fatigue to be of potential concern, although the strains were associated with environmental effects which have a low number of cycles at once per day. Measured strains associated with both reflective cracking and restraint moments were primarily driven by seasonal and daily temperature variations, highlighting the important role of thermal effects in design.
Besides the detailed field investigations, a parametric study of PCSSS designs was conducted to determine whether there was an economic benefit of continuous system design. In particular, design implications of time-dependent and thermal gradient restraint moments and their effects on continuity were studied. Because the PCSSS is a simple-span system made continuous with a CIP deck, the effects of restraint moments must be considered in the design of continuous systems. The restraint moments are those that arise from continuity, or end restraint, over the piers due to beam creep, differential shrinkage between the CIP deck and beam, and thermal gradient. Restraint moments would not develop if PCSSS were built as a series of simple spans with no continuity provided between the spans.
Eight bridges covering the feasible range of span configurations were designed as both simple and continuous systems. Flexural design was performed for each case, resulting in optimized precast sections within practical design constraints. Primary design parameters were strand number, section depth, and precast concrete strength. These design parameters were compared between the continuous and simple-span designs for each configuration to evaluate economic benefit. Generally, continuous PCSSS designs were equally or less economical than simple-span designs. Spans less than 30 feet had a slight economic benefit with continuous design because large restraint moments did not develop. However, spans greater than 30 feet developed large restraint moments in continuous design, particularly due to thermal gradient effects. In addition, the restraint moments greatly reduced continuity, effectively negating the benefit to live-load capacity. It was recommended that the PCSSS be designed as a simply-supported system for live load. Furthermore, because most continuous system designs were less economical than the corresponding simply-supported designs, it was concluded that designing the PCSSS as simply-supported while also including a continuity connection would be unconservative without accounting for restraint moments. A simple method was developed to account for restraint moments for this case without time-intensive calculations. Further recommendations related to the analysis of negative moments over PCSSS bridge piers were also provided.
A review of current design methods and details concluded that the current PCSSS design was generally sufficient, and recommendations for future PCSSS designs were provided. Items reviewed were related to shrinkage restraint, reflective crack control, composite action, and defining tolerances for the PCSSS. In order to try to better control top surface deck cracking, recommendations included increasing the transverse reinforcement in the CIP deck to provide a gross reinforcement ratio, ρg, of 0.0063 with a spacing no greater than 9 in., based on the work of Frosch (2006). This would translate to increasing the current transverse deck reinforcement from No. 4 bars at 6 in. (ρg =0.0056) to No. 5 bars at 6 in. (ρg =0.0086) or No. 5 bars at an increased spacing of 8 in. (ρg =0.0065) to provide the needed volumetric ratio while maintaining the maximum spacing for surface crack control. The recommendations of NCHRP 10-71 for reinforcement in the trough are adopted in order to control reflective cracking In addition, it was recommended that composite action stirrups need not be used if the required shear stress transferred between the CIP concrete deck and precast beam is less than 135 psi.