Integrated Kinematic and Dynamic Optimal Synthesis of Spatial Mechanisms

A novel integrated technique permitting the simultaneous optimization of kinematic and dynamic characteristics in the synthesis of spatial mechanisms is shown. The nonlinear programming formulation determines mechanism variables which simultaneously minimize the maximum values of bearing reactions, joint torques, driving torque, shaking moment, and shaking force while constraining the maximum kinematic structural error to a prescribed bound. The method is applied to the design of a path generating RRSS spatial mechanism with prescribed input link timing. Dynamic reactions in the mechanisms synthesized using the integrated technique were substantially reduced when compared to those of a mechanism synthesized to satisfy only the specified kinematic conditions.

Representing Component Types for Design

We are interested in programs which can support the design of a variety of machines. These will require program modules which can be combined by the designer. We provide parts of a theory for such modules, and illustrate the theory with examples from a simple test program. Finally, we sketch the application of the theory to more complex problems, and outline some issues for further research.

Mechanical Product General Scheme CAD Based on Expert System Technology

This paper expounds the requirements and capabilities of the application of expert system technology to mechanical product general scheme (MPGS) in CAD: systems, discusses the structure and the characteristicsof these systems, and studies the key technology in the procedure of system design.

A PHIGS-Based Graphics Input Interface for Spatial Mechanism Design

Creation of input specifications for synthesis or analysis of spatial mechanisms can be a significant problem. A graphics preprocessor which interactively assists in the definition of spatial mechanism problems is described. New methods of depth cucing and six DOF data entry are presented. To achieve graphics device-independence, the proposed graphics standard PHIGS (Programmer’s Hierarchical Interactive Graphics System) is used. Examples of application are presented including generation of input commands for Integrated Mechanisms Program (IMP) and generation of input for spatial mechanism synthesis routines.

Design and Implementation of an Intelligent CAD System

System requirements and system design for integrating a production rule program and a computer aided design system are presented. An implementation using a commercially available graphics modeling system is described. A “suggestive mode” interface is programmed as an example with application to design for automated assembly. Initial use of the implementation indicates that encoding production rules is more difficult than with conventional text-only knowledge-based systems, but that this system is a more effective way to use artificial intelligence techniques in design.

A Hybrid-Type Expert System for the Design of Marine Power Plants Using AI Technique and Design Optimization Method

A hybrid-type expert system is developed for supporting the initial design process of marine power plants. Firstly, discussion is given generally to understand design process in the view point of applying the AI technique effectively to design. Based on the result of the discussion, a hybrid-type expert CAD system with coupling the AI technique and the numerical optimization method is developed. In the system, the design knowledge is represented in the production rules, and the data of machineries consisting the plant are described by the frame-type representation. Through the system execution, it is ascertained that the system is effective not only as the design tool assisting designers but also as the tool instructing inexperienced designers.

Comparison of Modelling Methods for Hydraulic Valve Gear

A combined lumped/distributed parameter model for the follower system of a hydraulically operated valve is compared to a lumped parameter model of the same system. Since previous results show excellent correspondence between the lumped/distributed parameter model and measured response, it is natural to attempt to simplify the model and to again perform a comparison. Results of several examples are shown.

The Complementary Roles of Symbolic Computing and Numerical Optimization in Engineering Design Software

In many engineering design problems, the designer converges upon a good design by iteratively evaluating a mathematical model of the design problem. The trial-and-error method used by the designer to converge upon a solution may be complex and difficult to capture in an expert system. It is suggested that in many cases, the design rule base could be made significantly smaller and more maintainable by using numerical optimization methods to identify the best design. The expert system is then used to define the optimization problem and interpret the solution, as well as to apply the true heuristics to the problem. An example of such an expert system is presented for the design of a valve anti-cavitation device. Because of the capabilities provided by the optimization software, the expert system has been able to outperform the expert in the test cases evaluated so far.

Analysis and Optimal Design of Spatial Mechanical Systems

This paper presents a new approach to optimal design of large multibody spatial mechanical systems. This approach uses symbolic computing to generate the necessary equations for dynamic analysis and design sensitivity analysis. Identification of system topology is carried out using graph theory. The equations of motion are formulated in terms of relative joint coordinates through the use of velocity transformation matrix. Design sensitivity analysis is carried out using the Direct Differentiation method applied to the relative joint coordinate formulation for spatial systems. Symbolic manipulation programs are used to develop subroutines which provide information for dynamic and design sensitivity analysis. These subroutines are linked to a general purpose computer program which performs dynamic analysis, design sensitivity analysis, and optimization. An example is presented to demonstrate the efficiency of the approach.

Implementation of Semi-Heuristic Reasoning for Boundedness Analysis of Design Optimization Models

A production system performing global boundedness analysis of optimal design models has been implemented in the OPS5 programming environment. The system receives as input an initial model monotonicity table and derives global facts about boundedness and constraint activity using monotonicity principles. Additional facts may be discovered by heuristic search of implicit elimination sequences that examine boundedness of reduced models with active constraints eliminated. The global facts generated automatically by this reasoning system can be used either for a global solution, or for a combined local-global active set strategy.