The Hysteresis of Pore Water: The Non-Independence of Domains

While the Independent Domain Concept Accounts Rather Satisfactorily for the Hysteretic Behavior of Pore Water in Some Porous Bodies, There Is No Structural Reason Why This Should Be So. Some Artificial Materials Have Been Reported Whose Properties Vary Widely From the Predictions of the Independent Domain Model, and More Recently the Authors Have Obtained a Natural Sand in Which the Discrepancies Are Quite Substantial. A Theoretical Analysis Is Undertaken, Supported by Experiments, to Account for the Properties of Materials, Which May Occur Widely, for Which the Existing Theory Is Inadequate. In the Theoretical Analysis It Is Shown That the Distribution Function Underlying the Concept Has Not Adequately Accounted for the Phenomena of Wetting and Drying Suctions, Which Are More Complicated Than Presumed; In Particular, Some Interchanges of Domain Element Identity Occur That Escape Any Attempt at Detection. In the Experiments a Technique Was Used That Had Been Designed Previously by One of the Authors for Observing Sets of Wetting Scanning Curves, Including Primary Wetting Curves, Together With a Set of Primary Drying Curves for Comparison With Those Calculated From the Wetting Scanning Curves. The Methodology Is Described on a Step-by-Step Basis. Distributions at All Stages of Reversal Wetting Suction Are Obtained, and From Them Primary Drying Curves Are Calculated for Reversal Wetting Suctions Greater Than and Less Than 18.3 Cm. Distribution Diagrams Are Plotted in a Matrix of Drying Versus Wetting Suctions. The New Method of Analysis Is Said to Be Greatly Superior to Existing Methods. Several Diagrams Are Presented to Show How Additional Information Can Be Obtained From the Fine Structure of the Transfer on Nonindependent Domain Elements From One State to Another.