Hidesada Kanda

Professor, Mathematical Foundation of Computer Science lab, Software Department

Dr. Eng. 1987 (University of Tokyo), M. Eng. 1969, B. Eng. 1967 (Kyoto University)

Background Data

Professor Kanda came to the University of Aizu from IBM Japan. At IBM, he researched a regional energy system which aimed to simulate energy demand and supply by the year 2000 in the Kansai region, and did major work in computational fluid dynamics to solve a problem of transition from laminar to turbulent flow. He then lectured on communication systems and UNIX systems. At the University of Aizu, he directed the Information Systems & Technology Center from 1993 to 1998. Now he belongs to the mathematical foundation of computer science lab of the software department. He lectures on numerical analysis and grid generation. His research now focuses on determining critical Reynolds numbers numerically and theoretically for circular pipe flows. This problem was experimentally found in 1883 by Reynolds and nobody has succeeded in calculating the minimum critical Reynolds number of about 2000.

Research/Educational Interests

Research Description

The introduction of the computer into engineering and science has resulted in the emergence of a completely new field termed computational physics. It is certainly not pure theoretical analysis - if anything, it is closer to experimental analysis. The beautiful display screens of computer graphics have vividly revealed some aspects of the physical world which we have never seen directly. However, does computational physics yield any new fundamental formulas of physics? There still exist a lot of unsolved physical problems. Actual discovery of physical phenomena and a new understanding of unsolved problems are both possible by numerical calculations after numerical grid generation, rigorous stability analyses and error estimates are discussed. The area of computational physics is relatively young in practice, although its roots in mathematics are old.

CALCULATION of the MINIMUM CRITICAL REYNOLDS NUMBER FOR CIRCULAR PIPE FLOWS: A conceptual model was constructed for the problem of determining in circular pipes the conditions under which the transition from laminar to turbulent flow occurs. From many previous experimental investigations, it became clear that (i) plots of the transition length versus the Reynolds number (Re) show that the transition occurs in the entrance region under the condition of a natural disturbance, and (ii) plots of the critical Re versus the ratio of bellmouth diameter to the pipe diameter show that for the case of a straight pipe the critical Re takes a minimum value of about 2000. In the entrance region, the velocity profile changes from uniform at the inlet to parabolic at the entrance length. We found that the radial component of the curl of vorticity multiplied by (2/Re), which we call the normal wall strength, works as an acceleration force and decreases inversely as Re increases. Hence, the onset of the transition should depend on whether or not the acceleration power provided by the normal wall strength exceeds a required value. In this study we calculated the acceleration power via finite difference calculations, and thus obtained the minimum critical Reynolds number of 2040 when using J0 = 101 radial grid points.

Selected Publications

e-mail: kanda@u-aizu.ac.jp