Center for Engineering and Scientific Computation, and College of Computer Science, Zhejiang University, Hangzhou, Zhejiang, PR China.
Position: Cheung Kong Chair Professor
Dates: Dec. 2001 - Present

The current mission is to direct the Center for Engineering and Scientific Computation, Zhejiang University.

Major research projects involved are listed here.

Taitech, Inc., Turbomachinery and Propulsion Systems Division,
NASA Glenn Research Center at Lewis Field (formerly Lewis Research Center), Cleveland, Ohio, U. S. A.
Position: Senior Research Scientist
Dates: Dec. 1998 - Mar. 2002

The research objective was to accomplish further development and applications of the DRAGON grid technology to turbomachinery and propulsion systems. This work includes investigation into 3D DRAGON gridding methodology [AE99a, AE00a, AE00b, AE00c], and conduct of flow analyses relevant to NASA Glenn programs [AE00a, AE00b, AE00c]. The research also includes the optimization of code for parallel processing on clustered workstations to take advantage of rapid advances in computer technology. [Pictures: DRAGON Grids] [Pictures: 2D Supersonic Flow in a Symmetric Convergent Channel] [Pictures: 3D Supersonic Flow in a Convergent Duct] [Pictures: Subsonic Flow through a Wavy-Wall Duct] [Pictures: Inviscid Flow around a Linear Cascade] [Pictures: Viscous Flow through an Annular Cascade]
The DRAGON grid, as a hybrid grid, is created by means of a Direct Replacement of Arbitrary Grid Overlapping by Nonstructured grid. The DRAGON grid scheme is an adaptation to the Chimera thinking. The Chimera grid is a composite structured grid, composing a set of overlapped structured grids, which are independently generated and body-fitted. The grid is of high quality and amenable for efficient solution schemes. However, the interpolation used in the overlapped region between grids introduces error, especially when a sharp-gradient region is encountered. The DRAGON grid scheme is capable of completely eliminating the interpolation and preserving the flux conservation property. It adapts and maximizes the advantages of both structured and unstructured grids, while at the same time it eliminates the weaknesses of the Chimera scheme and minimizes the memory requirement associated with the unstructured grid.

Analysis and Design Application Company (adapco), New York, U. S. A.
Position: Senior Software Scientist
Dates: Jan. 1997 - Nov. 1998

adapco is in the field of providing engineering tools and consulting for automotive, aerospace, power generation, electronics, and defense applications. It offers expertise in the areas of stress analysis, structural analysis, fluid dynamics, heat transfer analysis, structural dynamics and failure analysis.
I had taken part in the company's general practices of CFD simulation for automotive industry. With emphasis, I had been involved in the research and development of automatic mesh generation techniques for use in CFD analysis software such as STAR-CD. The specific tasks with my major participation are:
• Development and implementation of surface mesh manipulation utility.
  This utility is designed to provide certain geometric capability such as CAD repair at the level of surface meshes. The main operations include edge zipping, hole filling and various sectioning for surface meshes.
• Research and development of SAMM package.
  SAMM (semi-automatic meshing methodology) is a mesh generator using ``trimmed cell technology''. The program starts with an all hexahedral mesh and trims the cells which intersect a working surface. Cells totally inside the volume are retained as is, while cells totally outside the surface are removed from the mesh. [Pictures: SAMM meshes]

One interesting feature of the SAMM code is the capability to create a new surface using the trimming technology on a given rough surface mesh.

Department of Civil Engineering (Institute for Numerical Methods in Engineering),
University of Wales Swansea, U. K., in collaboration with the Department of Mechanical Engineering
Position : Senior Research Assistant (eq. Research Associate)
Dates: Jan. 1995 - Dec. 1996

I joined the Wolfson Laboratory for Computational Fluid Dynamics, and worked on an European Union project entitled CAESAR: Multidisciplinary User Environment for Parallel Computing Engineering. CAESAR is an acronym derived from Clusters of computationally intensive Applications for Engineering design and Simulation on scalable pARallel platforms.
The CAESAR project enables and demonstrates the exploitation of HPCN (High Performance Computing and Networking) to facilitate the optimization of design and production cycles in the manufacturing industry as product complexity increases. Over 10 organizations in four European countries as Aerospace, Automotive, Chemical Engineering and Ship Building sectors are involved.
A Parallel Simulation User Environment (PSUE) is a sub-task of the CAESAR project[CSE96a, CSE96b, CSE96c, CSE00a]. The functionality of the PSUE includes a geometry builder and geometry repair module, grid generation and statistics, numerical libraries, domain decomposition, parallel platforms, performance monitoring, database and data analysis. Within this PSUE, several multidisciplinary applications programs perform, such as Flow Simulation, Electromagnetic Simulation, Optimal Ship Design, Production Planning, and Chemical Engineering. The PSUE system is presently widely used in industries including British Aerospace, Dassault Aviation (France), DASA (Germany), Daimler Chrysler, BASF, Odense Ship Builders, NAG (UK), and ESI (France). [Pictures: PSUE]
The primary part of the PSUE is an Interactive Geometry Utility Environment (IGUE), which consists of a geometry builder (surface modeler) and grid generation interface. I am the principal developer of the IGUE[CSE02a,CSE00a, CSE00b, CSE01a, Pre96a]. In this environment, sophisticated graphical user interfaces are essential, which are based on X-Window system (Motif). The underpinning data structure of the IGUE is based on Non-Manifold topology, while the 3-D graphics rendering library employed is OpenGL. [Pictures: IGUE]
Additionally, effort had been made to enhance and further develop FEView V2.0, a visualizer I wrote initially in a previous period[CSE93c, CSE94c]. The current version is a standalone multi-platform module independent of Geomview. It is built upon Xforms (the X version of Forms) and OpenGL libraries.

Position: Senior Research Assistant (eq. Research Associate)
Dates: Jan. 1991 - Dec. 1994

In the Casting Simulation Group, I had been working on an EPSRC (Engineering and Physical Sciences Research Council, UK) project entitled CAD for Casting. The specific tasks in this project are:
• Development and implementation of 3-D mesh generators.
  I started with the creation of a 3-D structured mesh generator based on multi-block techniques, and emphasized the interactive specification of multi-block topologies[CSE93a, CSE93b].[Pictures: 3D multiblock meshes]
  After that, I concentrated on some aspects of unstructured mesh generation including point generation and Delaunay triangulation[CSE93d, CSE94a, CSE94b].
  My implementation of point generation based on domain decomposition (ie. quadtree with triangular and square shapes) has been incorporated into a 2-D Delaunay type mesh generator, so that the mesh generator has been enabled to work with adaptive analyses[CSE93d, CSE94b, CSE95b]. [Pictures: 2D adaptive meshes]
  The effort on 3-D unstructured mesh generation has led to a 3-D Delaunay mesh generator (in C), in which the Steiner point creation algorithm has been employed. The embedded sophisticated boundary recovery procedures benefit the successful utilization of this mesh generator for complex industrial components. Linear, quadratic and NURBS patches are used to defined surface geometry for the mesh generator.[Pictures: Triangular surface meshes]
  The research has resulted in the integration of a three-dimensional unstructured mesh generation toolkit[CSE94e, CSE94f, CSE94g, CSE95b].[Pictures: 3D tetrahedral meshes]
  In this research, attention has been paid to the definition of mesh quality, and visual quality control.
• Development and implementation of a visualization program for finite element applications.
  The work began with some studies of computer graphics with the inclusion of writing a hidden line elimination program, so that I got an adequate understanding of some computer graphics techniques such as hidden line removal, hidden surface removal, ray tracing, and interactive facility.
  Based on the Forms Library (a graphical user interface toolkit for SGI), I had written an interactive visualization tool (FEView V1.0, in C) for finite elements[CSE93c, CSE94c], which works as an external module to Geomview (viewing and manipulating geometric objects). [Pictures: FEView ]
• Applications of geometric modeling.
  The work was carried out to understand solid and surface modelers, and to write some interface modules for data exchange between mesh generators and geometric modelers.
  A certain amount of practices has been undertaken to get some complex shapes of real mechanical components modeled. [Pictures: Geometric modeling]
• Applications of heat conduction analyses to real casting problems.
  I carried out some 3D numerical modeling work for P.I. Casting Ltd, UK. Mechanical components in investment casting process were numerically modeled, where radiation had been taken into account.

In addition, jointly I carried out an investigation on the potential applications of intelligent preprocessing in the numerical simulation of castings[ME93a].

Position: PhD Candidate
Dates: Mar. 1989 - Dec. 1990

During this period, I was involved in four research aspects, that is, back (inverse) consolidation problems; engineering optimization; finite element methods of heat transfer; and numerical modeling of casting process and ground freezing. The corresponding results are itemized as follows:
• A methodology investigation of back analysis problems for geotechnical engineering. An application of this methodology is in soil consolidation [CE94a]. During the consolidation process, deformation is coupled with multi-phase flow in the porous medium.
• Some investigations of engineering optimization methods aiming to couple these methods with back analyses[CE92a, CE9293a, CE94b].
• Applications of numerical heat conduction modeling to casting process and ground freezing.
• Further development and implementation of an adaptive heat conduction software package, HT2UF, with particular reference to the uranium hexafluoride (UF6) filled container transportation problem [NPE9293a, NPE91a, NPE91b, NPE90a]. A fire test of the nuclear fuel container was numerically modeled according to the mechanisms of heat conduction, radiation, and phase changes. The three phases (vapor, liquid and solid) of uranium hexafluoride were taken into account in detail. The resulting package is being used in British Nuclear Fuels Plc. [Pictures: Numerical simulation of heat transfer]

Additionally, in collaboration with a colleague, I undertook consulting assignments of heat transfer analyses of casting processing for British Light Metal Foundry Association, March - December, 1990[ME91a]. [Pictures: Numerical simulation of casting process]

Department of Civil Engineering, Zhejiang University, Hangzhou, P. R. China.
Position: PhD Candidate
Dates: Sep. 1986 - Feb. 1989

Apart from the lecture courses and the site practices in geotechnical engineering, I undertook some research tasks which included:
• An investigation on soil consolidation, of which the underlying physics is multi-phase flows in porous media.
• Applications of modern physico-mathematical methods, such as grey system theory, fuzzy set theory, orthogonal design method, probabilistic approach, and expert systems, to geotechnical engineering[CE88a].
• A study of practical computer methods in geotechnology.
• An application of fuzzy set theory in solid mechanics[Me88b].

Department of Engineering Mechanics (Currently Department of Space Engineering and Mechanics), Harbin Institute of Technology, Harbin, P. R. China.
Position: Graduate Student
Dates: Sep. 1984 - Jul. 1986

Besides the lecture courses required, I mainly carried out some research work on elasto-plastic fracture of composite materials, of which one of the potential applications is the design and evaluation of fiber-reinforced composite structures to be used in astronautics and aeronautics industry. My interest in solid mechanics and structural mechanics has remained in all the research periods afterwards. Some of the more significant aspects that I was involved with through this length of time included:
• Systematic research on crack-tip fields in orthotropic elasto-plastic materials by means of analytical approaches. The types of cracks include modes I, II and III, and static, quasi-static and dynamic cases [Me8689a, Me88a, Me90b, Me86a, Me93a].
• A good understanding of manufacturing process and mechanical behaviors (such as fracture mechanism) of fiber-reinforced composite materials.
• Investigations into fracture mechanics of general media such as heterogeneous materials[ME90a, Me96a, Me00a].
• Studies on the constitutive laws of orthotropic elasto-plastic materials [Me90a], and limit loads of orthotropic plates[CE8687a].
• A study on analytical solutions of bending of rectangular plates with small deflections, subjected to complex lateral loads.

Back to Yao's homepage.