UNIFORMITY OF SIMULTANEOUS DEPOSITION OF POWDER IN MULTIPLE DIES: MEASUREMENTS AND MODELING
Open Access
- Author:
- Xie, Xinsheng
- Graduate Program:
- Agricultural and Biological Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- November 23, 2005
- Committee Members:
- Virendra Puri, Committee Chair/Co-Chair
Michael Adebola Adewumi, Committee Member
Abraham S Grader, Committee Member
Harvey Bright Manbeck, Committee Member
Roy Edward Young, Committee Member - Keywords:
- Uniformity
Powder deposition
Multiple die filling
Modeling
Measurement
Pressure deposition tester
PDT-II - Abstract:
- This research aimed at testing, evaluating, analyzing, and modeling the deposition process and uniformity of powder fill in multiple dies. The second generation pressure deposition tester (PDT-II) was developed to investigate the effects of some factors (die geometry and size, die configuration and location of the die, powder characteristics, and feed shoe speed) on the deposition process and final pressure distribution. A battery powder mixture and an alumina powder were used to fill three parallel-oriented dies. Cylindrical, toroidal, and E-shaped dies were investigated. For the cylindrical and toroidal dies, feed shoe speeds of 20, 100, 200 (for the alumina powder), and 500 mm/s (for the battery powder mixture) were tested. For the E-shaped dies, 20 mm/s was used. A computed tomography (CT) scanner was employed to obtain fill density distribution for the cylindrical and toroidal dies filled with the two powders. PDT-II satisfactorily generates a real-time pressure profile of the process and final pressures at multiple locations. For the cylindrical dies filled with the battery powder mixture: 1) at 20 and 100 mm/s feed shoe speeds, the half circle close to the leeward end had higher final pressures; 2) at 500 mm/s, the final pressure distribution was more uniform; 3) the final pressure distribution was not always symmetrical about the center line. Final pressure decreased with increasing radial distance for low feed shoe speeds. The distribution was not always regular for high speed; 4) the three parallel dies did not always have similar pressure distributions; 5) the 500 mm/s feed shoe speed resulted in higher (P < 0.05) final pressures (774.5 to 1424.5 Pa) than at lower speeds (235.2 to 1136.0 Pa) at most locations; 6) at 20 and 100 mm/s feed shoe speeds, the right die tended to have higher (P < 0.05) final pressures (393.8 to 1136.0 Pa) than the center die (235.2 to 726.0 Pa). At 500 mm/s, the quantitative differences between the center (774.5 to 1246.1 Pa) and right dies (897.3 to 1424.5 Pa) were reduced (most locations P > 0.05); 7) most powder (¡Ý 90%) was deposited in the forward stroke. Results of the toroidal dies filled with the battery powder mixture showed: 1) the 0¡ã orientation area had the highest pressures (1186.7 to 2498.0 Pa). The average pressures of the remaining area were 353.7 to 648.0 Pa; 2) the pressure distribution was symmetrical; 3) the 500 mm/s feed shoe speed led to the most nonuniform and the densest filling; 4) Higher feed shoe speed did not always result in higher final pressures ; 5) the right die tended to have higher (P < 0.05) final pressures (215.0 to 2498.0 Pa) than the center die (95.4 to 2052.5 Pa). Results of the E-shaped dies indicated that: 1) the final pressures of the middle leg (308.9 to 760.7 Pa for the battery powder, and 41.9 to 130.7 Pa for the alumina) were higher (P < 0.05) than those of the left and right legs (148.9 to 530.3 Pa for the battery powder, and 26.1 to 79.6 Pa for the alumina); 2) for the battery powder, the area along the back side had the highest final pressures (1054.6 to 1303.8 Pa); 3) the pressure distribution was symmetrical about the center line; 4) neither the center die, nor the right die always had higher pressures than the other one; 5) for the alumina powder, the area of the highest final pressures (126.4 to 152.3 Pa) was in the vicinity of the junction of the middle leg and the back. For the cylindrical dies filled with the alumina powder: 1) at 20 and 100 mm/s feed shoe speeds, final pressure distribution was symmetrical. The center zone had the highest pressures (P < 0.05), and final pressures decreased with increasing radii; 2) at 200 mm/s, final pressure distribution was irregular and varied more than at lower feed shoe speeds; 3) no consistent shoe speed effect was discovered; 4) at 20 and 100 mm/s, the right die tended to have higher (P < 0.05) final pressures (49.5 to 288.8 Pa) than the center die (68.9 to 167.0 Pa). At 500 mm/s, the center (58.7 to 213.6 Pa) and the right (126.7 to 208.9) dies had similar final pressures (P > 0.05). Neither of the cylindrical and toroidal dies led to consistently higher final pressures for the battery powder mixture (P > 0.05). For the alumina powder, no consistent trend and no large differences were observed between cylindrical and toroidal dies (P > 0.05). CT scanner data of the cylindrical dies demonstrated: 1) CT number vs. bulk density did not have a strong linear correlation for the two powders (R2 ¡Ü 0.625); 2) the final bulk density distribution obtained by CT scanning was uniform within each die and among the three dies, based on the linear correlation with R2 ¡Ü 0.625 of CT number vs. bulk density; 3) CT scanner and PDT-II measured two different attributes (distribution of bulk densities and distribution of pressure values at the die bottom, respectively); 4) CT data and PDT-II data revealed that the half circle close to the leeward end had relatively high final mass (pressure) values. For cylindrical dies filled with the battery powder mixture at 20 mm/s feed shoe speed, the entire pressure profile was divided into 10 stages, which were modeled by an overall rate equation. The overall rate equation was: dPp/d¦Ó = ¦ÁPpF(¦Ó) + ¦Â, where Pp is prorated pressure, ¦Ó is normalized time, ¦Á, ¦Â are coefficients, and F(¦Ó) is a stage specific deposition rate function. The average root mean square error (RMSE) and the mean value of average relative difference (ARD) of the model for totally 17 locations in the vicinity of the center (r ¡Ü 4 mm) of the center die were 0.13 and 0.07 (or 7%), respectively. Stage 1, a very short period in the beginning of the forward stroke, deposited most of the total powder filled during the entire filling cycle. Rate equation for stage 1 was: dPp/d¦Ó = ¦ÁPp/( eb¦Ó -1), where ¦Á and b are deposition rate related coefficients. The model can serve as a qualitative tool for further investigation of the multiple die filling process.