Toward the topology design of mechanisms that exhibit snap-through behavior

https://doi.org/10.1016/j.cma.2004.02.017Get rights and content

Abstract

Topology optimization has proven to be a powerful method for the conceptual design of structures and mechanisms. In previously published work, we concentrated on the development of numerical methods that accommodate the finite deformation and incorporated these analyses into the topology optimization. We demonstrated by relatively straightforward transversely loaded clamped–clamped beam examples that topology optimization can be used to design structures that experience snap-through behavior. Here, we focus our attention on the design problem formulation where the goal is to develop a general approach for the design of mechanisms that experience more complex snap-through behavior. A multiphase design strategy is outlined, numerous significant challenges to this complex design process are discussed, and several examples are presented that demonstrate progress toward this goal.

Introduction

Stiffest structural design by topology optimization has become an automated and routine process. The extension of the methodology to more complex applications and other mechanics problems is the focus of current topology optimization research. Bendsøe and Sigmund [1] provide a comprehensive review of the current state of the art regarding structural and mechanism design by topology optimization. We apply topology optimization here to design mechanisms that exhibit snap-through behavior. For example, the structures in Fig. 1 experience large configuration changes (denoted by the dotted lines) due to elevated load levels, and in a more complicated manner, the mechanism depicted in Fig. 2 holds a workpiece (denoted by the circle) then releases/places the workpiece (at the position denoted by the square) after experiencing snap-through behavior. These problems require more computational effort, demand more attention to their formulation, and are more prone to computational difficulties than the topology optimization problems previously encountered.

Our goal is to design structures that exhibit snap-through behavior and behave like mechanisms. To illustrate the type of performance we desire, refer to the beam and spring system depicted in Fig. 3(a). The displacement response u of the system due to an applied dead load P can be highly nonlinear (refer to Fig. 3(b) for typical nonlinear load–displacement trajectories). The different paths represent the trajectories for various spring stiffnesses k, e.g. with high, medium, and low stiffnesses for paths O–A, O–B, and O–C, respectively. Structures that exhibit snap-through behavior (refer to path O–B of Fig. 3(b)) experience significant configuration changes at a given load level. Furthermore, if the structure has two stable unloaded configurations (refer to points O and C3 on path O–C of Fig. 3(b)), then it can be toggled between and held in either bistable configuration. Structures that exhibit this unique behavior are particularly attractive for actuation and sensing applications. For example, structures that exhibit snap-through behavior are well-suited for rapid switching operations, e.g. in an optical switch, and bistable structures can be used for storage applications or in circuit breakers.

The typical approach to structural and mechanism design is ad hoc and heuristic and is based heavily on designer intuition and experience. For example, structures that exhibit snap-through and usually bistable behavior are typically based on simple beam or cap designs. With the advent of high speed, low cost computing and mature computational analysis techniques, researchers have focused more attention on computer-aided synthesis techniques, e.g. see [2]. Despite their desirable performance, a systematic method to automate the conceptual design of general mechanisms that exhibit snap-through behavior has not been developed. This can be attributed to the difficulty of the design problem and not to the demand for such structures. For example, bistable mechanisms are particularly well-suited for MEMS applications, e.g. see [3], [4]. Here, we use topology optimization because it has proven to be a powerful method for the conceptual design of structures and mechanisms.

There has been some previous work by other researchers on this design problem. Jensen et al. [5] have formulated a design process based on spring and multibody dynamics, i.e. the so-called pseudo-rigid-body model, and they have designed novel and complex bistable mechanisms. However, the most difficult task of creating a suitable initial design is left to the designer, and the parameterization of its topology is fixed at the outset of the design process. Sekimoto and Noguchi [6] have used topology optimization to design toggle structures, e.g. a keyboard key, by prescribing its desired load–displacement trajectory. However, they only consider clamped–clamped beam design domains that will always exhibit snap-through behavior for a sufficiently stiff structure loaded at a sufficiently high load level, and the performance is only analyzed just beyond its limit load. Here, we use topology optimization to design such mechanisms that are not based on a fixed, initial topology and that experience full snap-through behavior.

Stiffest design formulations typically assume that the structural response can be adequately captured by linear elasticity, i.e. by a small displacement gradient assumption. However, it is imperative that the design of structures and compliant mechanisms that undergo finite deformation be obtained by embedding nonlinear structural analyses into the topology optimization [7], [8]. In previous work [9], we investigated numerical methods to analyze the complex nonlinear behavior in anticipation of the analysis requirements for the problems posed here. Also, we designed structures with clamped–clamped beam design domains that exhibit (full) snap-through behavior. This preliminary work proved that it was possible to design such structures despite the obstacles, and it suggested that the approach could be extended to more general mechanism design. Here, we more thoroughly investigate the optimization problem formulation and discuss the challenges encountered toward realizing this goal.

Numerous significant design challenges are enumerated in Section 2. In Section 3, the topology optimization problem is briefly described. More thorough descriptions of the topology optimization formulation and numerical methods for the nonlinear elastic finite element analyses are given in the appendices. Examples from previous preliminary work are briefly discussed in Section 4. In Section 5, the multiphase design approach for the design of mechanisms that exhibit snap-through behavior is discussed. We demonstrate the approach by several examples in Section 6.

Section snippets

Design challenges

There are numerous challenges to defining an appropriate optimization problem. Some of the challenges are inherent to the topology formulation and others are due to the complex nonlinear behavior that we are trying to capture. It cannot be overemphasized that the automation of the design task undertaken here is extremely difficult.

To illustrate some of the challenges that are faced, consider the design domain depicted in Fig. 4(a). Our goal is to generate a beam structure that exhibits

Topology optimization problem

The goal of the topology problem is to determine the optimal distribution of material within a design domain that minimizes a given cost function and satisfies a series of constraint functions. Details about the formulation and implementation of the topology optimization problem and its numerical methods are discussed in the appendices. Here, we describe components of the formulation that are necessary for solving our design problem.

To generate a well-posed topology optimization problem, we

Preliminary examples

Before describing the multiphase design strategy, we present some examples of previous preliminary design work to illustrate the range of complexity of the problems that we are trying to solve. The most straightforward problem is to generate a topology of a clamped–clamped beam traversely loaded at its mid-span by dead load P as described in [9]. The design domain is depicted in Fig. 8. We formulate an optimization problem in which the goal of snap-through, i.e. motion, is balanced with a goal

Design strategy

A multiphase process is necessary to design bistable mechanisms. Here, we outline the three phase design strategy.

  • In phase I, the design space is searched so that appropriate design parameters can be selected. For example, if the stiffest structural design is generated and subsequently analyzed using different applied load and spring stiffness combinations, the extreme range of motion can be determined. Alternatively, the performance of a prescribed topology based on an educated guess of the

Examples

Here, we present examples to illustrate the multiphase design strategy. In particular, we focus on phase II, i.e. to generate designs that exhibit snap-through behavior in a consistent manner. In the first example, we design a snap-fit mechanism. Next, snap-expansion and snap-inversion mechanisms are designed. Unlike the single input–output point in the transversely loaded, clamped–clamped beam design, in these examples we desire different behavior at the input and output points, e.g.

Conclusion

We have presented a multiphase strategy toward the realization of mechanism designs that exhibit snap-through behavior. At the current stage, the design process requires the designer to survey the design space and to interactively evaluate the optimization progress, and therefore, the process is not entirely automated. There are numerous design challenges enumerated in Section 2 that are inherent to these design problems, that currently cannot be systematically solved in a straightforward

Future work

There are several issues that can be further addressed in future work. Hinges appear in the optimized topologies of the compliant mechanisms, and therefore a strategy to prevent the one-node connected hinges can be implemented [15]. The location of boundary supports highly influence the success of the design for snap-through, so the support conditions may be simultaneously introduced as additional design variables [16]. Since the design space for compliant mechanisms in a continuum setting is

Acknowledgements

The authors gratefully acknowledge the financial support of the Danish Research Council through the THOR/Talent-programme: Design of MicroElectroMechanical Systems (MEMS) and the support and encouragement of the TopOpt Group members in the Department of Mechanical Engineering (Solid Mechanics) at the Technical University of Denmark.

References (21)

  • T.E. Bruns et al.

    Topology optimization of nonlinear elastic structures and compliant mechanisms

    Comput. Methods Appl. Mech. Engrg.

    (2001)
  • M.P. Bendsøe et al.

    Topology Optimization: Theory, Methods and Applications

    (2003)
  • E.J. Haug et al.

    Applied Optimal Design

    (1979)
  • M.T.A. Saif

    On a tunable bistable MEMS––theory and experiment

    J. Microelectromech. Systems

    (2000)
  • J. Qiu, J.H. Lang, A.H. Slocum, A centrally-clamped parallel-beam bistable MEMS mechanism, in: Technical Digest of the...
  • B.D. Jensen et al.

    Design of two-link, in-plane, bistable compliant micro-mechanisms

    J. Mech. Design, Trans. ASME

    (1999)
  • T. Sekimoto et al.

    Homologous topology optimization in large displacement and buckling problems

    JSME Int. J.

    (2001)
  • C.B.W. Pedersen et al.

    Topology synthesis of large-displacement compliant mechanisms

    Int. J. Numer. Methods Engrg.

    (2001)
  • T.E. Bruns et al.

    Numerical methods for the topology optimization of structures that exhibit snap-through

    Int. J. Numer. Methods Engrg.

    (2002)
  • T. Buhl et al.

    Stiffness design of geometrically nonlinear structures using topology optimization

    Struct. Multidiscipl. Optimizat.

    (2000)
There are more references available in the full text version of this article.

Cited by (56)

  • Topology optimization of bistable elastic structures — An application to logic gates

    2021, Computer Methods in Applied Mechanics and Engineering
    Citation Excerpt :

    Topology optimization has previously been used to design bistable devices. In [21] it was used to design a structure with a desired load–displacement response. In [6] a cardiovascular stent was designed to obtain a desired snap-through behavior.

  • Design of composite structures with programmable elastic responses under finite deformations

    2021, Journal of the Mechanics and Physics of Solids
    Citation Excerpt :

    At the material level, topology optimization has been adopted in many studies to design novel microstructures with engineered properties, such as negative Poisson’s ratio (Wang et al., 2014a; Wang, 2018), negative thermal expansion coefficient (Sigmund and Torquato, 1997), tunable elastic modulus (Sigmund, 1994, 1995) and stress–strain relations (Wang et al., 2014a; Wang, 2018), photonic and phononic bandgap (Bonnecaze et al., 2003; Dobson and Cox, 1999), and improved piezoelectric properties (Sigmund et al., 1998). At the structure level, topology optimization has been adopted to program unconventional force–displacement responses, including programming the snap-through (Bruns and Sigmund, 2004) and hardening/softening responses of slender structures (Lee et al., 2020) and controlling the force–displacement path of compliant mechanisms (da Silva et al., 2020). Composite materials and structures possess many attractive properties that are unavailable in single-material ones.

  • Design of compliant mechanisms using continuum topology optimization: A review

    2020, Mechanism and Machine Theory
    Citation Excerpt :

    Therefore, substantial efforts have been conducted on the size optimization of such mechanisms [304–307]. However, there have been many examples throughout the literature demonstrating the efficiency of topology optimization in obtaining bistable CMs with new configurations and better functionalities [308–311]. The seminal work of using topology optimization to design structures that exhibit snap-through behavior can be found in [312].

View all citing articles on Scopus
View full text