TY - JOUR
T1 - Buoyancy effects of a growing, isolated dendrite
AU - Canright, D.
AU - Davis, S. H.
N1 - Funding Information:
This work was supported in part by a grant from the National Aeronautics and Space Administration, Microgravity Science and Applications Program, and in part by the Naval Postgraduate School Research Council. The authors are grateful to Dr. S.C. Huang for providing the experimental data cited, in tabular
PY - 1991/10
Y1 - 1991/10
N2 - An axisymmetric dendrite of pure material solidifies downward into an undercooled melt. Surface energy and kinetic undercooling are negligible. The Ivantsov [Dokl. Akad. Nauk SSSR 58 (1947) 567] parabolic dendrite is modified by buoyant convection. We construct an approximate solution to the growth/convection problem in powers of a buoyancy parameter G. The solution depends on Prandtl number P and Stefan number S (undercooling). When P and/or S are large enough, buoyancy enhances growth and distorts the dendrite by sharpening the tip and widening the base. These results compare well with the experiments on succinonitrile (P = 23) of Huang and Glicksman [Acta Met. 29 (1981) 701] and the local theory of Ananth and Gill [J. Crystal Growth 91 (1988) 587] up to G ≈ 1000, but overpredict convective effects for larger G. When P and S are small enough, buoyancy slows growth and flattens the tip. Physical explanations are given for the differences in buoyant effects at different P. The results suggest that near-tip effects of buoyancy should be different in metallics than in organics.
AB - An axisymmetric dendrite of pure material solidifies downward into an undercooled melt. Surface energy and kinetic undercooling are negligible. The Ivantsov [Dokl. Akad. Nauk SSSR 58 (1947) 567] parabolic dendrite is modified by buoyant convection. We construct an approximate solution to the growth/convection problem in powers of a buoyancy parameter G. The solution depends on Prandtl number P and Stefan number S (undercooling). When P and/or S are large enough, buoyancy enhances growth and distorts the dendrite by sharpening the tip and widening the base. These results compare well with the experiments on succinonitrile (P = 23) of Huang and Glicksman [Acta Met. 29 (1981) 701] and the local theory of Ananth and Gill [J. Crystal Growth 91 (1988) 587] up to G ≈ 1000, but overpredict convective effects for larger G. When P and S are small enough, buoyancy slows growth and flattens the tip. Physical explanations are given for the differences in buoyant effects at different P. The results suggest that near-tip effects of buoyancy should be different in metallics than in organics.
UR - http://www.scopus.com/inward/record.url?scp=0026238892&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0026238892&partnerID=8YFLogxK
U2 - 10.1016/0022-0248(91)90690-7
DO - 10.1016/0022-0248(91)90690-7
M3 - Article
AN - SCOPUS:0026238892
VL - 114
SP - 153
EP - 185
JO - Journal of Crystal Growth
JF - Journal of Crystal Growth
SN - 0022-0248
IS - 1-2
ER -