On the stable geometry of self-formed alluvial channels: theory and practical applicationThis article is one of a selection of papers in this Special Issue in honour of Professor M. Selim Yalin (1925–2007).

On the basis of previous work by the late Professor M. Selim Yalin and the author, the process of self-formation of alluvial streams and the final (equilibrium or regime) geometry of the self-formed stream are considered in the light of thermodynamic principles, including the first and second laws, and the Gibb’s equation; the stream is treated as an isolated and irreversible system. The present analysis suggests that stream self-formation is guided by the need of the stream to progressively decrease its average flow velocity to accommodate the increase in the entropy of the system with the passage of time. The reduction in flow velocity is achieved by an appropriate alteration of stream slope, cross-sectional geometry, and effective roughness, the regime development being the process of this appropriate alteration. A method is presented for the computation of regime width, depth, and slope. The method rests on the channel formation criterion derived from thermodynamic principles and the expression of regime flow width determined on the basis of zero net cross sediment transport rate at the regime state. The regime channels computed from this method are compared with field and laboratory data from various sources.

Application of Bond Graphs to Thermofluid Processes and Systems

Over the past few years a study was focused on the development of bond graphs for thermofluid processes and systems using the true power variables of temperature and time rate of change of entropy. This paper summarizes results of the study. Discussion begins with the study of a simple case of single phase incompressible fluid flow and ends with a completely general case of multiphase, variable density flow. Variations in density require introduction of the momentum equation to the bond graph. Inclusion of entropy as a state variable necessitates the use of Gibb’s equation and its representation by means of bond graphs. This paper presents these formulations and representations, and compares dynamic results predicted by bond graphs with those of classical approaches.