Percolation Segregation in Multi-Size and Multi-Component Particulate Materials: Measurement, Sampling, and Modeling

Open Access
- Author:
- Jha, Anjani Kumar
- Graduate Program:
- Agricultural and Biological Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- October 19, 2007
- Committee Members:
- Virendra Puri, Committee Chair/Co-Chair
Eileen Wheeler, Committee Member
Douglas Brian Beegle, Committee Member - Keywords:
- Modeling
Sampling
Coarse and Fines Sizes
Primary Segregation Shear Cell
Segregation
Continuous Mixtures
Particulate Materials - Abstract:
- Particulate materials are routinely handled in large quantities by industries such as, agriculture, electronic, ceramic, chemical, cosmetic, fertilizer, food, nutraceutical, pharmaceutical, power, and powder metallurgy. These industries encounter segregation due to the difference in physical and mechanical properties of particulates. The general goal of this research was to study percolation segregation in multi-size and multi-component particulate mixtures, especially measurement, sampling, and modeling. A second generation primary segregation shear cell (PSSC-II), an industrial vibrator, a true cubical triaxial tester, and two samplers (triers) were used as primary test apparatuses for quantifying segregation and flowability; furthermore, to understand and propose strategies to mitigate segregation in particulates. Toward this end, percolation segregation in binary, ternary, and quaternary size mixtures for two particulate types: urea (spherical) and potash (angular) were studied. Three coarse size ranges 3,350-4,000 µm (mean size = 3,675 µm), 2,800-3,350 µm (3,075 µm), and 2,360-2,800 µm (2,580 µm) and three fines size ranges 2,000-2,360 µm (2,180 µm), 1,700-2,000 µm (1,850 µm), and 1,400-1,700 µm (1,550 µm) for angular-shaped and spherical-shaped were selected for tests. Since the fines size 1,550 µm of urea was not available in sufficient quantity; therefore, it was not included in tests. Percolation segregation in fertilizer bags was tested also at two vibration frequencies of 5 Hz and 7Hz. The segregation and flowability of binary mixtures of urea under three equilibrium relative humidities (40%, 50%, and 60%) were also tested. Furthermore, solid fertilizer sampling was performed to compare samples obtained from triers of opening widths 12.7 mm and 19.1 mm and to determine size segregation in blend fertilizers. Based on experimental results, the normalized segregation rate (NSR) of binary mixtures was dependent on size ratio, mixing ratio, material, strain rate, and strain. Segregated fines mass of potash and urea particles was significantly different for the same size ratio (p<0.05). The (NSR) and segregation rate of fines for binary mixtures were higher for larger size ratios, as expected (2.4:1.0>2.0:1.0>1.7:1.0). Segregation rate was the highest and lowest for mixing ratios 33:67 and 67:33, respectively, when coarse mean size was 3,675 µm. The NSR decreased when the strain rate was decreased from 1.0 Hz>0.5 Hz>0.25 Hz for the binary size ratios 1.7:1.0, 2.0:1.0, and 2.4:1.0 (p<0.05). The NSR was dependent on multi-size mixtures (binary>ternary>quaternary). At strain rate of 0.5 Hz for the size ratio 2.0:1.7:1.0 in ternary mixture, the NSR for potash (0.83 kg/kg-h) was higher than the NSR for urea (0.21 kg/kg-h) (p<0.05). The NSR increased with the increase in strain from 2% to 10%. At strain of 6% and strain rate of 0.5 Hz, for the size ratio 2.0:1.7:1.0 in ternary mixture, the NSR for potash (0.83 kg/kg-h) was higher than the NSR for urea (0.21 kg/kg-h) in ternary mixtures (p<0.05). For size ratios 2.0:1.0 and 1.7:1.0, only 2.8% and 7.0% of decrease in NSRs were recorded for increase in relative humidity by 10 points (from 40% to 50%), respectively, whereas 36.0% and 45.0% decrease in NSRs were recorded for increase in relative humidity by 20 points (from 40% to 60%), respectively (p<0.5). Additionally, flowability was quantified using a Cubical Triaxial Tester (CTT) for size ratios 2.0:1.0 and 1.7:1.0, angle of internal friction increased from 31.3° to 35.9° to 39.0° and 27.4° to 32.0° to 36.0° when relative humidity increased from 40% to 50% to 60%, respectively (p<0.05), whereas cohesion remained close to zero (p>0.05). An innovative time-sequence procedure for sampling of bags was devised and implemented. The size guide numbers (SGNs) of 10-10-10 blend samples using 19.1 mm width trier were larger than those obtained using the 12.7 mm width trier, there were no substantial differences between the SGNs and uniformity index (UIs) of the two different width triers, i.e., except for 10-10-10 (sample from second quarter) from blend plant 1, all SGNs and UIs were within 7 and 2, respectively. Eleven out of the twelve samples from bagged fertilizers using 12.7 mm vs. 19.1 mm had the same outcomes, i.e., only one sample from blend plant 3, of 10-10-10 (sample from third quarter) using 12.7 mm vs. 19.1 mm had a conflicting outcome – the sample obtained using 19.1 mm width trier (SGN=259, UI=47) passed, whereas, the sample with 12.7 mm trier (SGN=256, UI=47) failed the AOAC chemical analysis test. Two triers of opening widths 12.7 mm and 19.1 mm were used to determine percolation segregation under vibration conditions. The percent mass retained on sieve number 8 (opening width = 2.36 mm) was the highest (28%). Sieve numbers 7 (2.80 mm) and 10 (2.00 mm) retained 21% and 17% and followed the sieve number 8 (2.36 mm). The particle size distributions were not significantly affected at the frequency of 5 Hz (p>0.05). The SGN and UI increased with time at the frequency of 7 Hz. The SGN and UI of samples collected by 12.7 mm and 19.1 mm triers were not significantly different (p>0.05). Results of binary size ratios of potash at three strains of 2%, 6%, and 10% each at strain rates of 0.25 and 0.5 Hz and for urea at strains of 2%, 6%, and 10% each at strain rates of 0.25 and 1.0 Hz were used to develop and validate a convective and diffusive model. Based on validation results the segregated fines mass values for size ratio 2.0:1.0 were: (1) within the 95% CI for potash at 6% strain and 0.5 Hz strain rate, (2) similar but not within the 95% CI for potash at 10% strain and 0.5 Hz strain rate, and (3) close but not within the 95% CI for urea at 6% strain and 0.5 Hz strain rate. A dimensional analysis model included the binary, ternary, quaternary mixtures of urea and potash at strains of 2%, 6%, and 10% and strain rates of 0.25, 0.5, and 1.0 Hz. Strain rate of 1.0 Hz was included at strain of 6% for binary mixtures only. Developed dimensional analysis model was validated for binary and ternary size ratios of urea and potash at strain rate of 0.5 Hz. Based on the experimental data, it seems binary mixtures of potash results were sufficient to represent the percolation segregation in binary, ternary, and quaternary mixtures of urea and potash with reasonable accuracy (CoV of ±15%). The validation results showed that the CoV of the modeled values to the experimental values were 18%, 15%, and 11% for binary mixtures of urea and potash at strains of 2%, 6%, and 10% and strain rate of 0.25 Hz. In conclusion, the binary, ternary, and quaternary mixtures provided the needed framework to predict the segregation behavior in continuous mixtures under different motion conditions. Based on initial results, only a limited number of tests need to be performed using multi-size and multi-component mixtures to determine the extent of segregation in continuous mixtures, the number of sizes and components should be determined based on operating conditions. Percolation segregation in blended fertilizer could be minimized by changing physical, mechanical and environmental parameters of particulate materials. The percolation segregation was modeled for time-dependent and time-independent conditions to minimize the number of tests to improve materials and process efficiency.