Jenmu Wang

The Study of Wind Force Coefficient Predictions for Rectangular High-Rise Buildings

In wind-resistant design of structures, the calculation of wind coefficients is usually based on data from wind tunnel tests. The process is very time-consuming and expensive. In order to predict wind coefficients of rectangular buildings, polynomial and nonlinear regression were studied. Also, artificial neural networks (ANNs) were used as well to train, simulate and forecast wind coefficients using terrain, side ratio (D/B) and aspect ratio (H/B) as inputs. The neural networks used include BP (Back Propagation), RBF (Radial Basis Function) and GR (General Regression) neural networks. According to the investigation presented in this paper, RBF neural network is the most effective mean to predict wind coefficients.

Web System for CFD Pedestrian Wind Simulation and Assessment

This paper reports the development of a web-based wind environment simulation and pedestrian wind assessment system using computational fluid dynamics (CFD). Graphical user interfaces were built using Internet browsers to link to the server side CFD programs, which hides the complexity of CFD programs and makes the tool accessible to engineers at the early architectural design stage for quick evaluation of wind field around buildings. The paper describes the system function and architecture, case testing and comparison, and possible future improvement and expansion.

Numerical Simulation and Wind Tunnel Studies of the Wind Load on Wind Turbine Structures

This paper investigated the structural responses of the wind turbine due to wind loads by performing the wind tunnel test and the Computational Fluid Dynamics, (CFD). The base shear force and the base moment of the wind turbine measured by the wind tunnel test were compared with the numerical simulation results. Both the numerical dynamic mesh and sliding mesh models were selected for the numerical simulations. The results showed that the dynamic mesh model was better than the sliding model by comparing to the wind tunnel test result. In the case of the k-epsilon RNG turbulence model, the prediction of the bending moment affecting by acrossswind was more than 50%, and the prediction of the force affecting by acrosswind was less than 3%. The both simulation results of the prototype and the full scale wind turbine were obtained by CFD model. The comparisons of the result showed that the error of Fx was about 15% and My was about 13.5%.

A Virtual Reality Framework for RC Building Design and Construction Cooperation

With the advance in computer technology, virtual reality (VR), which allows users to explore and interact within three-dimensional (3D) virtual environment, becomes affordable and begins to play a vital role in various engineering practices. By establishing 3D VR models, engineers are able to sense, examine, simulate and evaluate their design works and identify inconsistency between design and construction to ensure work quality. Building a 3D VR model is time-consuming and labor-intensive. Hence, it is important to develop an effective approach to relax the aforementioned limitation for rapid prototyping VR models. In this paper, we propose a framework that correlates engineering design process and World Wide Web technology to generate Virtual Reality Modeling Language (VRML) models. An information system for the design and construction of reinforced concrete (RC) building structures has been developed to demonstrate the versatility and robustness of the proposed framework.

Learning design plans in a knowledge-based blackboard system

The control information along a decision route that leads to the creation of a design solution is often referred to as design plan. Knowledge-based design systems frequently use precompiled design plans to control and schedule design activities. It is, however, difficult for design systems to learn new plans. The article introduces the concept of memory-oriented learning and presents an approach to learn design plans from normal design sessions in a blackboard design model.