Evaluation of Dowel Bar Retrofit Alternatives Using Accelerated Pavement Testing

This paper presents the results of accelerated pavement testing using the heavy vehicle simulator (HVS) to evaluate the performance of dowel bar retrofit (DBR) joints and cracks as measured by deflections and load transfer efficiency (LTE). Four HVS sections were constructed and tested: one with four epoxy-coated steel dowels per wheelpath, one with three epoxy-coated steel dowels per wheelpath, one with four fiber-reinforced polymer (FRP) dowels per wheelpath, and one with one joint with four epoxy-coated steel dowels per wheelpath and the other with four hollow stainless steel dowels per wheelpath. HVS and falling weight deflectometer (FWD) testing indicated that DBR significantly improved the LTE and reduced deflections compared with the undoweled pavement. Previous HVS trafficking damaged LTE on undoweled joints, but no significant loss of LTE was found after heavy HVS trafficking on DBR joints or originally doweled joints. Under FWD testing, the LTE from DBR joints (four steel dowels per wheelpath) was typically between 85 to 90% at lower temperatures before trafficking and greater than 95% for joints with originally installed dowels. The configuration using three dowels per wheelpath was found to have significantly lower LTE than that using four dowels per wheelpath. FRP and grout-filled hollow stainless steel dowels had similar performance to epoxy-coated steel dowels. The equivalent built-in temperature difference values of after-DBR sections are lower than those of before-DBR sections, likely as a result of the time of construction when the slabs were expected to be flattest. This finding indicates that tensile stresses in the concrete slabs could also be decreased by DBR.