Description / Table of Contents:
CONTENTS: Abstract. - Preface. - List of symbols. - Introduction. - Previous work. - Experimental design. - The radioisotope 22Na. - Description of apparatus. - Experimental procedure. - Correction of profiles. - Assumptions. - Decay correction. - Boundary correction. - Error analysis. - Results. - Salinity data. - Temperature data. - Growth velocity. - Discussion. - Brine and ice properties. - Brine salinity. - Brine density. - Brine volume. - Brine latent heat of freezing. - Brine viscosity, specific heat, and thermal conductivity. - Ice properties. - Theoretical brine expulsion model. - Continuity equations. - Thermal energy equation. - Simplified brine expulsion equations. - Brine expulsion in NaCl ice. - Results. - Discussion. - Gravity drainage in NaCl ice. - Application of results to natural sea ice. - Effective distribution coefficient. - Previous work. - Experimental procedure and results. - Conclusions. - Literature cited. - Appendix A: Profile correction data. - Appendix B: Program "correct" and sample output. - Appendix C: Tabulation of salinity data. - Appendix D: Tabulation of profile data. - Appendix E: Time-ice thickness equations (Runs 2 and 3). - Appendix F: Tabulation of distribution coefficient data.
Description / Table of Contents:
To obtain a better understanding of the desalination of natural sea ice, an experimental technique was developed to measure sequential salinity profiles of a growing sodium chloride ice sheet. Using radioactive 22Na as a tracer, it was possible to determine both the concentration and movement of the brine within the ice without destroying the sample. A detailed temperature and growth history of the ice was also maintained so that the variation of the salinity profiles could be properly interpreted. Since the experimental salinity profile represented a smoothed, rather than a true salinity distribution, a deconvolution method was devised to restore the true salinity profile. This was achieved without any significant loss of end points. In all respects, the salinity profiles are similar to those of natural sea ice. They have a characteristic C-shape, and clearly exhibit the effects of brine drainage. Not knowing the rates of brine expulsion or gravity drainage, the variation of the salinity profiles during the period of ice growth could be explained by either process. To determine the relative importance of the desalination mechanisms, a theoretical brine expulsion model was derived and compared to the experimental data. As input for the model, equations describing the variation of some properties of NaCl brine with temperature were derived. These included the brine salinity, viscosity, specific heat, thermal conductivity, and latent heat of freezing. The theoretical brine expulsion model was derived by performing mass and energy balances over a control volume of NaCl ice. A simplified form of the model, when compared to the experimental results, indicated that brine expulsion was only important during the first several hours of ice growth, and later became a minor desalination process relative to gravity drainage which continued to be the dominant mechanism for the remainder of the study period (up to 6 weeks). The rate of gravity drainage was found to be dependent on the brine volume and the temperature gradient of the ice. As either the brine volume or temperature gradient was increased, the rate of change of salinity due to gravity drainage increased. The equation commonly used to calculate the effective distribution coefficient (Weeks and Lofgren 1967) was modified and improved by taking brine drainage into account. An expression was also derived to give the distribution coefficient at very low growth velocities.
Type of Medium:
Series available for loan
vii, 85 S. : Ill.
Research report / Cold Regions Research and Engineering Laboratory, CRREL, US Army Material Command 345