Satish C. Negi

Neural networks for predicting nitrate-nitrogen in drainage water

Two artificial neural network (ANN) models, a trainable fast back-propagation (FBP) network and a self-organizing radial basis function (RBF) network, were developed for simulation of subsurface drain outflow and nitrate-nitrogen concentration in tile effluent. Experimental data collected at the Greenbelt Research Farm of Agriculture Canada over a 40-month period were used to train and validate the two models. The available field data were divided into training and testing scenarios, with the training file consisting of eight inputs and two outputs. A sensitivity analysis was performed by varying the network parameters to minimize the prediction error and determine the optimum network configuration. The best architecture for the FBP model comprised of 20 neurons in the hidden layer and a learning rate of 0.02, while the RBF network with a tolerance of 20 and a receptive field of 15 yielded 547 neurons in the hidden layer. Overall, the performance of the RBF neural network was superior to the FBP model in predicting the concentration of nitrate-nitrogen in drain outflow due to the application of manure and/or fertilizer. This information, in turn, can be used for proper fertilizer management; thereby, reducing not only the loss of valuable nitrogen fertilizer but also the potential for pollution of subsurface water by nitrate.

Modeling Sediment and Phosphorous Movement through Vegetative Filter Strips using Artificial Neural Networks and GRAPH

Present study explores the potential of Artificial Neural Networks (ANNs) which function somewhat like a black box, and GRAPH (GRAss PHosphorous), a deterministic hydrological model, to predict the movement of sediments and phosphorous through vegetative filter strips (VFSs). Data collected from a series of field experiments conducted at Elora and at Guelph Turf grass Institute, Guelph, Canada were used for evaluating different models. The synthetic runoff, a mixture of water, sediment, and cattle manure in known proportion, was allowed to pass through filter strips of different lengths and with different vegetation covers. Runoff flow rates at inlet and outlet of the strips were recorded. Collected samples were analyzed for total suspended solids, total and soluble phosphorous, and particle size distribution. Furthermore, experiments were conducted for different inflow rates. The available field data were used to train a back propagation (BO) neural network and a radial basis (RBF) neural network for predicting the performance of vegetative filter strips. The statistical comparisons between observed and predicted values of sediment and phosphorous amounts indicated that among the various models tried during the study, the best results were obtained from the back propagation neural network model. Although results indicated limited success with GRAPH, the model’s capabilities to provide insight into various physical processes governing the transport of sediment and attached pollutants cannot be overlooked. In this paper, details of the field experiment, approach adopted to develop various models, and the significance of the obtained results are presented and discussed.

Computer-Aided Prediction of Silo-Wall Pressures

THE design equations for predicting the distributions of lateral pressures and friction forces in farm tower silos are described. Three cases are presented which take into account the mode of unloading operation and the wetness of the ensiled crop. A computer-aided design package is developed to expedite the numerical computations and the plotting of the pressure-depth curves. The package is comprised of three files: a program for performing the design calculations and graphing the pressure distributions on the monitor screen, a plotter program for hardcopy plots, and the reference manual. These programs are intended to be as self-explanatory and user-friendly as possible.

Measurements of the at-rest pressure ratio of silage and grains

A transducer was developed for measurements within the mass of the grain of the ratio of horizontal to vertical pressure (K) in grain bins. It recorded three normal stresses and the tilt of the device with respect to two reference axes in the horizontal plane; thereby, providing two independent values of the parameter K. This ratio was also determined for silage by measuring the frictional resistance caused by drawing two blades, one oriented horizontally and the other vertically, through the confined mass. For the granular materials, the pressure ratios were measured to be 0·39 for barley and 0·48 for wheat. These values appear to be directly related to the bulk density of the material, initially 690 and 780 kg/m 3 for barley and wheat respectively. For whole-plant corn silage the K ratio was influenced by the orientation of fibres relative to blade surfaces. The pressure ratios of the granular materials were essentially independent of the vertical stress level. Good agreement was found between the experimental results and code recommendations.

Lateral Pressures in Farm Tower Silos

ABSTRACT THE paper discusses two discrete pressure distributions which can occur in farm tower silos depending upon the wetness of the contained silage material. Formulas and graphs based on available data in the literature are presented for both cases. A new method is put forward to account for the effective fiber pressures developed by a saturated silage mass. An example is included illustrating the step-by-step approach to calculation of lateral wall pressures.

Sulphide, Sulphate and Sulphuric Acid a Corrosion of Concrete in Laboratory Tests

Portland cement (PC) concrete is generally a highly durable structural material. Nevertheless, certain chemical actions and aggressive environments in a livestock building can cause deterioration and total collapses of structures have occurred long before they have reached their design life.

Tower silo design: Principles and computer implementation☆

Abstract Farm tower silo design in Canada is changing from the use of strictly empirical formulae to analytical methods. This means that several types of structures and a variety of silage materials at a range of moisture contents can be accomodated. This paper attempts to summarize the wall pressures caused by whole plant materials in bottom and top unloading silos. Both dry and wet silages are considered. Guidelines are provided also for determining pressures from high moisture grains. The determination of internal hoop tensile forces is also discussed. A design aid is provided for bottom unloading silos where the internal hoop tensile forces are difficult to calculate. A computer program was developed using the various algorithms to aid in the rapid determination of a pressure diagram and of the resulting hoop forces, given the silo geometry and the specification of the material intended to be stored.

Sulfide and Sulfate Attack on Reinforced Concrete of Livestock Buildings

Although reinforced concrete is one of the most durable construction materials, in farm buildings it is subjected to high levels of hydrogen sulfide and sulfate that result in corrosion. This leads to premature deterioration of slatted floors and other concrete components of liquid manure storages. The rehabilitation of reinforced concrete structures due to corrosion is expensive and impractical. Prevention is the better approach. Hazardous gases are released from stored manure. Of these, hydrogen sulfide is the most corrosive. Sulfatereducing bacteria generate sulfuric acid as the end of their metabolism 48 concrete cylinders, 100mm diameter and 100mm high, were made with Portland cement, and various combinations of slag, fly ash and silica fume. Each has a reinforcing steel bar embedded in it. 24 cylinders are half immersed in sodium sulfate (20,000ppm SO4 2-) and also subjected to hydrogen sulfide gas (1,000ppm H2S). The second set of 24 is subjected only to hydrogen sulfide gas. In each set, there are 8 different treatments. Initial results indicate that after 11 cycles of testing over about 22 months that the PC concrete with 0.5 W/C ratio is the least corrosion resistant. All treatments containing silica fume, fly ash or slag, except the fly ash/silica fume combination, performed better thus far than PC concrete with 0.4 W/C ratio. Also, there is an indication that concrete with sulfate resistant cement is more resistant than type I Portland cement. Indeed, sulfate resistant cement concrete was one of the best performers in both sets of tests.

Sulfide and Sulfate Corrosion of Concrete in Livestock Buildings

Concrete in farm buildings is subjected to severe hydrogen sulphide and sulphate concentrations that result in the corrosion of reinforced concrete. This leads to premature deterioration of walls and floors, especially slatted floors, to the point of requiring replacement. Corrosive gases are released from stored manure. Of these, hydrogen sulphide is the most corrosive agent that leads to the rapid deterioration of concrete in barns. As well, manure has high concentrations of sulphates. In the present study concrete cylindrical specimens with a reinforcing steel bar in the center were exposed to hydrogen sulphide and sulphates in solution. One half of the specimens is partially immersed in sodium sulphate (20,000 ppm SO4 2-) and also subjected to hydrogen sulphide gas (1,000 ppm H2S). The second set is subjected to hydrogen sulphide gas only. Each set consists of 8 different treatments including Portland cement (PC) concrete with 0.4 and 0.5 W/C ratios, PC concrete with 8% silica fume replacement, 25% fly ash and 35% slag of the total amount of cementitious material, and specimens made of PC concrete with combinations of silica fume and fly ash (6%/25%), and silica fume and slag (6%/25%). Finally one treatment is made with sulphate resistant cement.

On the Storage Pressures and Properties of Alfalfa in a Bottom-Unloading Silo

ABSTRACT Haylage pressures in a bottom unloading monolithic concrete silo were measured for two consecutive years with 15 transducers placed symmetrically around the circumference at five levels. The experimental data were used to estimate the ratio of horizontal to vertical pressure and the coefficient of friction between silage and silo wall. The results indicate that the lower and upper limits for pressure ratio of 0.5 and 0.6 may be taken for alfalfa haylage in concrete silos. Values for friction coefficient of 0.4, 0.5, and 0.6 can be used corresponding to moisture contents of 60%, 50%, and 40% (wet basis) for this material and silo type. The predictions based on the proposed Canadian standard for bottom-unloading silos agree closely with the measured wall pressures at all levels above the unloader. This code also provides a conservative estimate of the overpressures in the unloader region.