Towards Automated Design of Toggle Switch Mechanisms

Abstract

This work deals with addressing the issues related to design of double toggle switch mechanisms with emphasis on structural, dimensional and dynamic aspects. Currently, almost all the issues related to electrical switches are dealt from electromagnetic point of view; the operating mechanism is hardly touched. It is observed that kinematic parameters influence electrical performance of switch significantly. Therefore, there is a need to develop methodologies for supporting exploration of diverse kinematic chains (KCs) for this purpose. Visual inspection is tedious and error prone even when a complete list of design criteria is available, hence, the work presented in the thesis contributes towards automated design of toggle switch mechanisms.

In this context, in house modular kinematics data structure is found useful for using it as a tool in the design of toggle switch. Modular kinematics, typically used for kinematic analysis, works on the principle of finding the configuration of a mechanism using a given set of modules by a procedure called module sequence. This module sequence is used and interpreted in a number of ways for its effective use in various design stages. Structurally, a set of seven conditions must be satisfied by a KC to exhibit double toggle. These conditions are broadly classified into three categories: criteria for KC, function assignment criteria and criteria for stoppers. These three criteria are to be checked automatically by use of module sequence in the same order as mentioned.

In the criteria for KC, one of the conditions is that, the KC should not have fractionated degrees of freedom (d.o.f.). Hence, detection of fractionation in a KC is inevitable. In literature, is was found that the algorithms for detection operate at their worst case complexity, O(n4), and some of them do not report joint fractionation. Thus, the algorithms are not only robust but also computationally expensive. Therefore, a frugal and comprehensive method O(n2) is implemented to detect fractionation using modular kinematics. Also, inherent structural pattern embedded in fractionated KCs is hardly studied in literature. It is found that the way body and joint fractionation is defined in fractionated KCs is inconsistent. So, fractionation is interpreted as symbolic partitioning of joints and links in the traditional body and joint fractionation types respectively. Based on the number of ways of partitioning, simple and multiple types of fractionation are recognized. Valid partitions are identified using the notion of fractionating and non-fractionating subchains. Relative locations of these subchains influence distribution of d.o.f. across the fractionated KC. Conventional representation of KCs as links and joints or graphs is difficult to comprehend this distribution. For this, a novel concept of fractionation graph is introduced that gives d.o.f. distribution information and the relative locations of the constituent subchains across the KC. Modular kinematics gives a constructive description of fractionated KCs. Characterization of fractionated KCs, based on presence of multiple separation links, is introduced as order of fractionation. Uniqueness for a given order of fractionation is also justified.

After the criteria for KC, a KC is tested for feasibility for function assignment criteria. This requires recognition of active and passive subchains of the KC with respect to input and output pairs. For this, module sequence is characterized for recognition of the subchains. Based on these subchains, locations of stoppers are derived. Using this information, an algorithmic approach to assign functions (functions like spring, ground link, input link, etc.) to derive distinct driving mechanisms provided isomorphic elements (links and joints) of the KC are known beforehand, is introduced.

The design parameters influencing dimensional synthesis have been identified as dimensions of links, spring anchor points and stopper locations. Sub-problems associated with each parameter are analyzed. It is found out that optimum location of stoppers for selecting operational range of motion is necessary by taking into account the considerations of timing of switch and impact velocity. Based on the analysis, an algorithmic way to design single toggle switch mechanisms is introduced.

Timing for closing or opening of a switch is one of the critical measure that determines its performance. Timing should be as low as possible without exceeding the impact velocity at the instant contacts meet each other. Timing of a switch depends on the dimensions of the links, inertial parameters, spring stiffness etc. For a given timing for a mechanism, dynamic synthesis, in this thesis, deals with finding the inertial parameters of the links using Quinn's energy distribution method, modular kinematics, and Nelder and Mead's downhill simplex method for optimization. This thesis helps the designer to use modular kinematics as a potential automated tool to select a valid design to make the solution space more meaningful in the design of toggle switch mechanisms.