Method and means for sand reblending

A method of reblending a mixture of aggregates from a given location comprising:

• 1) providing a classification tank having an inlet and a plurality of stations Si, where i=1 to N, arranged at operatively progressively greater distances from the inlet to receive a raw feed input slurry comprising fluid and a mixture of aggregates of a plurality of mesh sizes M;
• 2) providing at each station Si a bed level sensor and a plurality of discharge valves made active only when the bed level sensor indicates that the mixture of aggregates has filled the station Si to a desired level, the plurality of discharge valves including at least a primary product discharge valve and a waste valve;
• 3) providing a computer connected to a programmable logic controller that is in turn connected to the bed level sensors and the discharge valves for automated control of the classification tank, the programmable logic controller setting open times on a percentage basis for each of the discharge valves within each of the stations in the tank based on input from the computer and recording and relaying to the computer the actual valve open times for each of the discharge valves;
• 4) inputting into the computer an initial data set for startup including an initial M×N Station Analysis matrix and initial discharge flow rates;
• 5) delivering the input slurry to the classification tank for a calibration run;
• 6) calibrating the tank by setting the primary product discharge valve at each station open 100% of a calibration cycle time, running the tank for the calibration cycle time, sampling and sieve testing the aggregate from the given location to determine raw feed input, sampling and sieve testing the aggregate discharged at each station Si by the tank, and developing a Calibrated M×N Station Analysis Matrix of actual tank performance in terms of individual percent retained for each mesh size Xj at each station Si, then
• 7) inputting data from the previous step into the computer and, by using the assumption that the raw feed input must be equal to the output discharged from the tank, calculating a 1×N matrix of discharge flow rate multipliers stated in terms of total valve open time at each station over the total valve open times at all of the stations so as to satisfy said assumption;
• 8) using the flow rate multipliers from the previous step, calculating in the computer a mathematical matrix model MathModelXY including an initial Product A array, the initial Product A array=Calibrated M×N Station Analysis Matrix×Matrix of Flow Multipliers;
• 9) inputting into the computer production specifications for a primary product, the production specifications for the primary product including a desired total quantity of the primary product to be discharged, a series of integers 1 to M designating a corresponding series of sieves Xj, where j=1 to M, having progressively smaller mesh sizes, and for each of the sieves or mesh sizes Xj respectively a desired specification range for the percent of the discharged aggregate capable of passing through said sieve;
• 10) simulating operating the tank in a production mode for a given cycle time;
• 11) comparing the raw feed input to the specifications for the primary product and calculating with the computer a variance from the desired range of percent passing for each mesh size Xj;
• 12) if all the variances are zero, sending a signal to the programmable logic controller to set the primary discharge valve at each station Si of the tank open 100% of the cycle time and skipping to step 15;
• 13) if any of the variances is nonzero,
• a) determining which of the variances is largest and which mesh or sieve size Xj is associated with the largest variance;
• b) establishing in the computer a simulation goal of reducing said largest variance;
• c) beginning with station 1, simulating setting the primary product discharge valve to 0% at one of the stations Si at a time and calculating a simulated Product A array with said one of the stations excluded;
• e) ceasing the previous step at a station Sn which first reduces the largest variance;
• f) using the Calibrated Station Analysis Matrix from the most recent calibration step, searching in ascending numerical order from station Sn to SN and identifying a station Xe with the highest percent retained for the mesh size Xj+l;
• g) simulating setting the primary product discharge valve to be open a lower percentage of the cycle time at station Xe to form a 1×N Usage Matrix;
• h) calculating a revised Product A array=Usage Matrix×Calibrated Station Analysis Matrix;
• i) calculating a resultant gradation including percent passing values for each mesh size;
• j) comparing the resultant gradation simulated in the previous step to the specifications for the primary product, Product A;
• k) calculating the variances from the desired range of percent passing for each mesh size Xj;
• l) repeating steps a)-k) and simulating reducing open times or excluding additional stations Xe from anywhere in the tank as dictated by steps a)-j) to bring the simulated resultant gradation of the primary product, Product A, into the specification range for all of the mesh sizes;
• 14) if Product A can be produced in the specification range for all of the mesh sizes, sending a signal from the computer through the programmable logic controller to the primary discharge valve at each station Si to establish primary discharge valve settings for a production cycle based upon the above simulations and calculations;
• 15) running the tank in a production mode for the production cycle;
• 16) during the production cycle, monitoring and recording the actual open times of the primary discharge valves with the programmable logic controller and relaying said actual open times to the computer, and then
• 17) recalculating the flow discharge rates, the raw feed input, and the variances based upon the actual open times in order to make any valve setting adjustments necessary before a next production cycle begins.