Tapered Plate Dowels for Jointed Concrete Slabs
Advisor: Professor Jeffery Roesler
Abstract
In jointed concrete slab systems, dowel bars are used to enhance the long-term load transfer efficiency at joints. Plate dowels were introduced in the 1990’s, followed by several additional tapered plate dowel (TPD) geometries. Tapered dowels were first introduced to minimize potential restraint of prismatic dowels, which may occur when misaligned at placement. Plate dowels also provide an alternative to round dowel bars through larger concrete-dowel area resulting in lower bearing stresses and differential deflections. Few published studies exist in the literature on the experimental and theoretical responses and performances of TPDs.
In this thesis, a large-scale experimental investigation using a cantilever slab-dowel test setup was conducted to compare the performance of three different TPD geometries, the elongated diamond dowel (EDD), the trapezoidal dowel (TRAP), and the hexagonal dowel (HEX). Each geometry was tested at three levels of longitudinal joint offsets, 0-,1-, and 2-inch, these tests showed a maximum difference in deflections between all dowel geometry and offset levels of 20%. The differences in deflections are relatively small given a change in dowel width at joint that ranged from 3 to 2 inches and embedded lengths from 6 to 3.65 inches. Results also showed a large change in the dowel’s punching shear capacity at failure with an 82% maximum difference.
2-D and 3-D FEM models were developed to simulate the experimental test setup to provide critical dowel responses. The 2-D simulation composed of shell elements that modeled the steel dowels with proper geometric and material inputs resting on an elastic foundation. They did not offer better deflection responses than 1-D Friberg with same support K. The 3-D FEM model represented the test and loading boundary conditions, and actual slab and dowel geometric and material inputs. Deflection profiles of 3-D models correlated well with the deflection profile of the TPD tests in the service load range, e.g., 2,000 lbf. These simulations showed maximum deflection differences between all test cases equal to 10% at the dowel-slab face.
In the 3-D FEM models, normal stress and corresponding deflection were extracted at the slab face and used to calculate modulus of dowel support, K, for 1-D Friberg’s analysis. Friberg solutions were then used to calculate critical dowel responses of the various TPD cases. The largest variation in critical responses was in the bearing stresses between TPD geometries and offsets. The EDD dowels resulted in the largest change in responses between offset levels, while the TRAP showed the least. Additionally, when compared with round dowel, bearing stress levels of TPDs were significantly lower.
A method to calculate punch shear capacity of TPDs was successfully developed using equations for breakout strength of anchors in precast concrete slabs. The accuracy of this new PSC solution was compared with 81 TPD test data from tests presented in this thesis and others from 2017 that used a similar test setup. This equation relied on design inputs including dowel embedment length, embedded dowel average width, concrete strength, and concrete cover depth. While the geometry of dowels did result in significant changes in PSCs, particularly with offset, the PSC was much more sensitive to concrete strength and slab thickness.
A predictive equation to calculate dowel-concrete support Ks for Friberg analysis using a multivariate linear regression analysis was completed. A total of 24 3-D FEM simulations were used, with variables including dowel dimensions, joint offset level, concrete cover thickness, and concrete elastic modulus. The results of the predictive equation showed an average error level of 5% with a correlation R2 equal to 99.6%. This K calculator allows for rapid estimation of support values for a given set of TPD inputs.
