No abstract

Coal combustion products (CCPs), traditionally treated as wastes, can be successfully used in road construction in a variety of ways. When CCPs are used in roadway construction, however, leaching of trace elements into groundwater can be a concern. The current paper describes a model that can be used to evaluate the impacts of leaching on groundwater for typical roadways in the USA. Three analytical solutions to the advection–dispersion–reaction equation are combined to develop a method for assessing impacts to groundwater caused by leaching of trace elements from CCPs used in highway subgrade, subbase and base layers. The solutions were combined into a computer program (WiscLEACH) with a Windows® graphical user interface. The analytical method was compared to predictions made with HYDRUS-2D, a software package for simulating flow and transport in variably saturated media. Predictions made with the analytical method were also compared to field data from a highway test section where the subgrade was stabilised with coal fly ash. Parametric analyses made with the method indicate that maximum concentrations in groundwater occur near the groundwater table (GWT) and the centreline of the pavement. Peak groundwater concentrations decrease as depth to groundwater increases, the thickness of the byproducts layer decreases, the seepage velocity in the vadose zone decreases or the seepage velocity in groundwater increases. Variables having the greatest influence on maximum concentrations in groundwater are depth to the GWT, thickness of the CCP layer, hydraulic conductivity of the least conductive layer in the vadose zone, hydraulic conductivity of the aquifer material and the initial concentration in the CCP layer.

A maximum-likelihood estimate of the mean orientation M in the matrix Fisher distribution is then 1~I = AF, which expressed in Euler angles is (~0~,~b,~o2)= (171.598°,2.169°,101.824°). This mean orientation is of course different from the result that would have been obtained by simply taking the arithmetic mean of the Euler angles. Such a simple approach to orientation averaging is not recommended. By the approach of Wood (1993) and with an approximation described by Mardia & Zemroch (1977), the concentration parameter is found as D~= diag(5.311 x 104,3.156x 104,2.615 x 104). The large values for ~b, indicate, as expected, a large clustering of the X's.Consider the problem of testing whether the X's can be assumed to be uniformly distributed on SO(3). In this case, the answer is clearly no but a formal test for the hypothesis of uniformity can be obtained using the statistic R defined by (18). For the data of

Recovering the range and 3-D motion of a dynamic scene from its image sequences is a very important problem in computer vision. In this paper, the velocities and positions of features in the binocular image sequeneea are used to determine the relative translational motion between the cameras and the objects, the correspondence between features in the left and right images, and, from this correspondence, the range to the features in question. Preliminary experimental results with synthetic data are presented which indicate that the technique is quite robust in the presence of noise.

The paper addresses the problem of plate modelling within the framework of the h–p‐version of the finite‐element method. A natural hierarchy of models is constructed. The lowest member of the hierarchy is the well‐known Reissner–Mindlin model. It is shown that higher degree elements do not show any locking effects for the models under consideration.