Uncertain design optimization of automobile structures: A survey

Xiang Xu; Chuanqiang Huang; Chongchong Li; Gang Zhao; Xiaojie Li; Chao Ma; Xiang Xu; Chuanqiang Huang; Chongchong Li; Gang Zhao; Xiaojie Li; Chao Ma

doi:10.3934/era.2023062

Electronic Research Archive

2023, Volume 31, Issue 3: 1212-1239. doi: 10.3934/era.2023062

Previous Article Next Article

Review Special Issues

Uncertain design optimization of automobile structures: A survey

1.
School of Engineering, Anhui Agriculture University, Hefei 230036, China
2.
School of Intelligent Manufacturing, Jiangsu Vocational Institute of Architectural Technology, Xuzhou 221116, China

Academic Editor: Jianguang Fang

Received: 03 December 2022 Revised: 13 December 2022 Accepted: 14 December 2022 Published: 03 January 2023

In real life, there are a lot of uncertainties in engineering structure design, and the potential uncertainties will have an important impact on the structural performance responses. Therefore, it is of great significance to consider the uncertainty in the initial stage of structural design to improve product performance. The consensus can be reached that the mechanical structure obtained by the reliability and robustness design optimization method considering uncertainty not only has low failure risk but also has highly stable performance. As a large mechanical system, the uncertainty design optimization of key vehicle structural performances is particularly important. This survey mainly discusses the current situation of the uncertain design optimization framework of automobile structures, and successively summarizes the uncertain design optimization of key automobile structures, uncertainty analysis methods, and multi-objective iterative optimization models. The uncertainty analysis method in the design optimization framework needs to consider the existing limited knowledge and limited test data. The importance of the interval model as a non-probabilistic model in the uncertainty analysis and optimization process is discussed. However, it should be noted that the interval model ignores the actual uncertainty distribution rule, which makes the design scheme still have some limitations. With the further improvement of design requirements, the efficiency, accuracy, and calculation cost of the entire design optimization framework of automobile structures need to be further improved iteratively. This survey will provide useful theoretical guidance for engineers and researchers in the automotive engineering field at the early stage of product development.

Keywords:

Citation: Xiang Xu, Chuanqiang Huang, Chongchong Li, Gang Zhao, Xiaojie Li, Chao Ma. Uncertain design optimization of automobile structures: A survey[J]. Electronic Research Archive, 2023, 31(3): 1212-1239. doi: 10.3934/era.2023062

Related Papers:

[1]	Elias Maciel, Inocencio Ortiz, Christian E. Schaerer . A Herglotz-based integrator for nonholonomic mechanical systems. Journal of Geometric Mechanics, 2023, 15(1): 287-318. doi: 10.3934/jgm.2023012
[2]	Xavier Rivas, Daniel Torres . Lagrangian–Hamiltonian formalism for cocontact systems. Journal of Geometric Mechanics, 2023, 15(1): 1-26. doi: 10.3934/jgm.2023001
[3]	Jordi Gaset, Arnau Mas . A variational derivation of the field equations of an action-dependent Einstein-Hilbert Lagrangian. Journal of Geometric Mechanics, 2023, 15(1): 357-374. doi: 10.3934/jgm.2023014
[4]	Alexander Pigazzini, Cenap Özel, Saeid Jafari, Richard Pincak, Andrew DeBenedictis . A family of special case sequential warped-product manifolds. Journal of Geometric Mechanics, 2023, 15(1): 116-127. doi: 10.3934/jgm.2023006
[5]	Marcin Zając . The dressing field method in gauge theories - geometric approach. Journal of Geometric Mechanics, 2023, 15(1): 128-146. doi: 10.3934/jgm.2023007
[6]	Robert I McLachlan, Christian Offen . Backward error analysis for conjugate symplectic methods. Journal of Geometric Mechanics, 2023, 15(1): 98-115. doi: 10.3934/jgm.2023005
[7]	Álvaro Rodríguez Abella, Melvin Leok . Discrete Dirac reduction of implicit Lagrangian systems with abelian symmetry groups. Journal of Geometric Mechanics, 2023, 15(1): 319-356. doi: 10.3934/jgm.2023013
[8]	Markus Schlarb . A multi-parameter family of metrics on stiefel manifolds and applications. Journal of Geometric Mechanics, 2023, 15(1): 147-187. doi: 10.3934/jgm.2023008
[9]	Valentin Duruisseaux, Melvin Leok . Time-adaptive Lagrangian variational integrators for accelerated optimization. Journal of Geometric Mechanics, 2023, 15(1): 224-255. doi: 10.3934/jgm.2023010
[10]	Francesco Bonechi, Jian Qiu, Marco Tarlini . Generalised Kähler structure on $\mathbb{C}P^2$ and elliptic functions. Journal of Geometric Mechanics, 2023, 15(1): 188-223. doi: 10.3934/jgm.2023009

Abstract

1. Introduction

An invariant volume is a powerful tool for understanding the asymptotic nature of a dynamical system. In particular, it is a well-known fact that unconstrained mechanical systems are volume-preserving. The case of nonholonomic systems is more nuanced as these systems generally fail to preserve the symplectic form (which follows from the fact that nonholonomic systems are not variational) and hence, the induced volume form. This makes the study of invariant volumes in nonholonomic systems nontrivial and the existence of an invariant volume is an exceptional scenario which should not be expected in general ^[1,2]. A famous example of this is the Chaplygin sleigh; this system, although energy-preserving, experiences volume "dissipation" which results in asymptotic stability (cf. ^[3] for a general discussion on stability of nonholonomic systems or ^[4] for an interpretation via impact systems).

The existence of an invariant volume for a nonholonomic system offers two key insights. The first is the usual case in dynamical systems where an invariant volume allows for the use of the Birkhoff Ergodic Theorem (cf. e.g. 4.1.2 in ^[5]), for recurrence (with the caveat of the volume being finite although some results still hold for infinite measures ^[6]), and as an obstruction to asymptotic stability. The other is unique to nonholonomic systems; even though nonholonomic systems are not Hamiltonian, "nonholonomic systems which do preserve volume are in a quantifiable sense closer to Hamiltonian systems than their volume changing counterparts, " ^[7] (see also ^{[8,9,10,11,12]}). Therefore, being able to find an invariant volume for a nonholonomic system allows for understanding its asymptotic behavior and can provide a way to "Hamiltonize" a nonholonomic system (although most systems with an invariant volume still cannot be Hamiltonized, cf. ^[13]).

There exists an abundance of research into finding invariant volumes for nonholonomic systems where the constraints are linear and symmetries are present: Chaplygin systems are studied in, e.g. ^{[11,14,15,16,17,18,19]}, Euler-Poincaré-Suslov systems are studied in, e.g. ^[20,21], systems with internal degrees of freedom are studied in, e.g. ^[20,22,23], and ^[24] studies the case of symmetric kinetic systems where the dimension assumption does not hold. Related work on asympotic dynamics may be found in ^[25]. There are also results demonstrating that no invariant volumes exist, e.g. ^[26,27].

The contribution of this work is to consider conditions for the existence of invariant volume in nonholonomic systems such that

1. the constraints need not be linear/affine in the velocities,

2. the analysis can be carried out on the ambient manifold rather than resorting to local coordinates, and

3. absolutely no symmetry assumptions are used.

Such an approach seems to be new.

Our main result is an existence condition for an invariant measure for a hyper-regular Lagrangian system with an admissible set of nonlinear nonholonomic constraints obeying Chetaev's rule. In the linear case the existence of a basic invariant density reduces to checking whether or not a certain 1-form can be made to be exact, ^[28], and this special case is similar to results in ^[15,24].

Preliminaries on nonholonomic systems are presented in Section 2. Section 3 presents the construction of the "extended nonholonomic vector fields" on the entire cotangent space $T^*Q$ which restricts to the nonholonomic vector field on $M$ . Although there does not exist a canonical choice for the extended vector field, they will agree once restricted. An advantage of this approach is that one is able to design extended vector fields which ensure that the constraint manifold is an exponentially stable invariant manifold. The divergence calculation for a nonholonomic system is performed in Section 4 via extended nonholonomic vector fields. The main results, Theorem 5.2 and Theorem 5.6, are proved in Section 5 (cf. Theorems 5.2 and 5.6). Additionally, it is shown that the invariant volume does not depend on the choice of extended nonholonomic vector field chosen and is unique up to a constant of motion. Section 6 presents an interpretation of the linear constraint case to the torsion of the nonholonomic connection, which seems to be a novel observation. This paper concludes with examples in Section 7.

This paper is a continuation of the work done in ^[29] and, as such, many of the results below can be found there.

2. Preliminaries and notation

2.1. Unconstrained mechanics

We will first briefly cover the case of unconstrained mechanical systems before discussing nonholonomic systems. A smooth (finite-dimensional, $\mathrm{dim}(Q) = n$ ) manifold $Q$ is called the configuration space, the tangent bundle $TQ$ is called the state space, and the cotangent bundle $T^*Q$ is called the phase space. The bundle projection maps will be denoted by $\tau_Q:TQ\to Q$ and $\pi_Q:T^*Q\to Q$ .

2.1.1. Lagrangian systems

Lagrangian systems are described by a smooth real-valued function on the state space called the Lagrangian, $L:TQ\to Q$ . The dynamics generated from the Lagrangian function arise from Hamilton's principle and are given by the Euler-Lagrange equations,

$\begin{equation*} \frac{d}{dt}\frac{\partial L}{\partial \dot{q}} - \frac{\partial L}{\partial q} = 0. \end{equation*}$

2.1.2. Hamiltonian systems

In constrast to Lagrangian systems, Hamiltonian systems evolve on the phase space. For a given Lagrangian, denote $\mathbb{F}L:TQ\to T^*Q$ as its fiber derivative.

Definition 2.1. A Lagrangian, $L:TQ\to \mathbb{R}$ , is hyperregular if the fiber derivative is a global diffeomorphism. Furthermore, a Lagrangian is natural if it has the form

$\begin{equation*} L(v) = \frac{1}{2}g(v,v) - \tau_Q^*V(v), \end{equation*}$

where $g$ is a Riemannian metric on $Q$ and $V\in C^\infty(Q)$ .

Throughout this work, all Lagrangians will be assumed to be hyperregular. Moreover, when affine constraints are considered, the Lagrangian will be assumed to be natural.

For a given Lagrangian, we define the Hamiltonian $H:T^*Q\to\mathbb{R}$ via the Legendre transform:

$\begin{equation*} H(p) = \langle p,v\rangle - L(v), \quad p = \mathbb{F}L(v). \end{equation*}$

Let $\omega = dq^i\wedge dp_i$ be the canonical symplectic form on $T^*Q$ . The resulting Hamiltonian vector field arising from the Hamiltonian $H$ and the symplectic form $\omega$ is denoted by $X_H$ and is given by

$\begin{equation*} i_{X_H}\omega = dH, \end{equation*}$

where $i_X\omega = \omega(X, \cdot)$ is the contraction. An important feature of Hamiltonian vector fields is that they preserve the symplectic form, and thus are volume-preserving.

Theorem 2.2 (Liouville's Theorem). Hamiltonian dynamics preserve the symplectic form and, additionally, preserve the volume form $\omega^n$ .

Liouville's theorem does not apply to general nonholonomic systems as they are generally not symplectic. Determining when an analogous version of this theorem applies to constrained systems is the main goal of this work.

2.2. Constrained dynamics

For unconstrained Lagrangian systems, motion is allowed to occur in the entire state space, $TQ$ . For constrained systems, motion is restricted to lie on a constraint manifold $N\subset TQ$ and we tacitly assume that $\tau_Q(N) = Q$ . For the purposes of this work, we will not differentiate between nonholonomic and holonomic constraints as the machinery for the former will suffice for the latter.

Let $N\subset TQ$ be the constraint submanifold which is, locally, described via the zero level-set of a regular function $G:TQ\to\mathbb{R}^k$ , $k < n = \dim(Q)$ . Let us denote each component of this function by $\Psi^\alpha$ , $\alpha = 1, \ldots, k$ , i.e.

$\begin{equation*} N = \bigcap\limits_{\alpha = 1}^k \, (\Psi^\alpha)^{-1}(0). \end{equation*}$

To transfer to the Hamiltonian point of view, we assume that the Lagrangian, $L:TQ\to\mathbb{R}$ , is hyper-regular. The constraints on the cotangent bundle become

$\begin{equation*} \Phi^\alpha = (\mathbb{F}L^{-1})^*\Psi^\alpha, \quad M = \mathbb{F}L(N) = \bigcap\limits_{\alpha = 1}^k \, \left( \Phi^\alpha\right)^{-1}(0). \end{equation*}$

These constraints are admissible if the following matrix is non-singular ^[18].

$\begin{equation*} m^{\alpha\beta} = \mathcal{C}^*d\Phi^\beta\left( X_{\Phi^\alpha}\right), \quad (m_{\alpha\beta}) = (m^{\alpha\beta})^{-1}, \end{equation*}$

where $X_{\Phi^\alpha}$ is the Hamiltonian vector field generated by $\Phi^\alpha$ and $\mathcal{C}:T(T^*Q)\to T(T^*Q)$ is the $\mathbb{F}L$ -related almost-tangent structure, explained below.

The constrained Euler-Lagrange equations become modified via Chetaev's rule (which is not necessarily the physically correct procedure, cf. ^[30] for a discussion), which will provide equivalent results to those in ^[31] where the "almost-tangent" structure of the tangent bundle is utilized. For Lagrangian systems, Chetaev's rule states that if we have the nonlinear constraints $\Psi^\alpha$ , then the constraint forces have the following form

$\begin{equation} \frac{d}{dt}\left( \frac{\partial L}{\partial \dot{q}}\right) - \frac{\partial L}{\partial q} = \lambda_\alpha\cdot \mathcal{S}^*d\Psi^\alpha, \end{equation}$

(2.1)

where $\mathcal{S}:T(TQ)\to T(TQ)$ is a (1, 1)-tensor called the almost-tangent structure and, in local coordinates, is given by

$\begin{equation*} S = \frac{\partial}{\partial\dot{q}^i}\otimes dq^i. \end{equation*}$

If, rather than being general nonlinear, the constraints are affine in the velocities,

$\begin{equation*} \Psi^\alpha(v) = \eta^\alpha(v) + \tau_Q^*\xi^\alpha(v), \end{equation*}$

where $\eta^\alpha\in\Omega^1(Q)$ are 1-forms and $\xi^\alpha\in C^\infty(Q)$ are functions, then (2.1) becomes

$\begin{equation*} \frac{d}{dt}\left( \frac{\partial L}{\partial \dot{q}}\right) - \frac{\partial L}{\partial q} = \lambda_\alpha\cdot \tau_Q^*\eta^\alpha. \end{equation*}$

However, this work will focus on the Hamiltonian formalism. The constraint manifold, $M\subset T^*Q$ , is locally described by the joint level-set of the functions $\Phi^\alpha:T^*Q\to\mathbb{R}$ where $\Phi(p) = \Psi\circ\mathbb{F}L^{-1}(p)$ . Moving (2.1) to the cotangent side, the constrained Hamiltonian equations of motion become

$\begin{equation} i_{X_H^M}\omega|_M = dH|_M + \lambda_\alpha\cdot \mathcal{C}^*d\Phi^\alpha|_M, \end{equation}$

(2.2)

where $\mathcal{C}:T(T^*Q)\to T(T^*Q)$ is the $\mathbb{F}L$ -related almost-tangent structure, i.e. if $L$ is natural, then

$\begin{equation*} \mathcal{C}^*\left( \alpha_idq^i + \beta^jdp_j\right) = g_{ij}\beta^jdq^i. \end{equation*}$

2.3. Notation

Throughout this work, $L:TQ\to\mathbb{R}$ will be a hyperregular Lagrangian with corresponding Hamiltonian $H:T^*Q\to\mathbb{R}$ . The constraint manifold will be called $N\subset TQ$ or $M = \mathbb{F}L(N)\subset T^*Q$ . These submanifolds will be locally described by the joint zero level-set of a collection of smooth functions,

$\begin{equation*} N = \bigcap\limits_{\alpha = 1}^k \, (\Psi^\alpha)^{-1}(0),\quad M = \bigcap\limits_{\alpha = 1}^k\, (\Phi^\alpha)^{-1}(0). \end{equation*}$

If the constraints are affine in the velocities/momentum, the Lagrangian will be assumed to be natural with Riemannian metric $g$ . In this case, the constraining functions will have the form

$\begin{equation*} \Psi^\alpha(v) = \eta^\alpha(v) + \tau_Q^*\xi^\alpha(v), \quad \Phi^\alpha(p) = P(W^\alpha)(p) + \pi_Q^*\xi^\alpha(p), \end{equation*}$

where $\eta^\alpha$ are 1-forms, $\xi^\alpha$ are functions, $W^\alpha = \mathbb{F}L^{-1}(\eta^\alpha)$ are vector fields, and $P(W^\alpha)$ is the vector field's momentum

$\begin{equation*} P(W^\alpha)(p) = \langle p, W^\alpha(\pi_Q(p))\rangle. \end{equation*}$

In this case of affine constraints, $\mathcal{D}\subset TQ$ will be the distribution

$\begin{equation*} \mathcal{D}_q = \bigcap\limits_{\alpha = 1}^k\, \ker \eta^\alpha_q, \end{equation*}$

and $\mathcal{D}^* = \mathbb{F}L(\mathcal{D})\subset T^*Q$ .

Finally, the nonholonomic vector fields with Hamiltonian $H$ and constraint manifold $M$ will be denoted by $X_H^M$ .

3. Extended nonholonomic vector fields

Given a constraint manifold, $M\subset T^*Q$ , we can determine the nonholonomic vector field, $X_H^M\in\mathfrak{X}(M)$ via (2.2). Commonly local, noncanonical, coordinates are chosen for $M$ (cf. §5.8 in ^[20] and ^[32]). However, we will instead work with a tubular neighborhood (for ease, this will be written as the entire manifold $T^*Q$ ) and define an extended vector field $X_H^{ext}\in\mathfrak{X}(T^*Q)$ such that $X_H^{ext}|_{M} = X_H^M$ . This section outlines an intrinsic (albeit non-unique) way to determine such a vector field.

Definition 3.1. For a given constraint submanifold $M\subset T^*Q$ , a realization of $M$ is an ordered collection of functions $\mathscr{C} \; : = \{\Phi^\alpha:T^*Q\to\mathbb{R}\}$ such that zero is a regular value of $G = \Phi^1\times\ldots\times \Phi^k$ and

$\begin{equation*} M = \bigcap\limits_{\alpha}\, (\Phi^\alpha)^{-1}(0). \end{equation*}$

If the functions $\Phi^\alpha$ are affine in momenta, i.e. $\Phi^\alpha = P(X^\alpha)+\pi_Q^*f^\alpha$ , then the realization is called affine.

Remark 3.2. In the case where the Lagrangian is natural (which provides a Riemannian metric on $Q$ ) and the constraint submanifold is affine, we can choose the realization to be affine:

$\mathscr{C} = \{P(W^1)+\pi_Q^*\xi^1,\ldots,P(W^k)+\pi_Q^*\xi^k\},$

where $W^i = \mathbb{F}L^{-1}\eta^i = (\eta^i)^\sharp$ .

By replacing $M$ with a realization $\mathscr{C}$ , we can extend the nonholonomic vector field to a vector field on $T^*Q$ such that $\Phi^\alpha$ are first integrals. Recall that the form of the nonholonomic vector field is $i_{X_H^M}\omega = dH + \lambda_\alpha\mathcal{C}^*d\Phi^\alpha$ . We construct the extended nonholonomic vector field, $\Xi_H^{\mathscr{C}}$ , by requiring that:

(NH.1) $i_{\Xi_H^\mathscr{C}}\omega = dH + \lambda_\alpha\mathcal{C}^*d\Phi^\alpha$ for smooth functions $\lambda_\alpha:T^*Q\to\mathbb{R}$ , and

(NH.2) $\mathcal{L}_{\Xi_H^\mathscr{C}}\Phi^\alpha = 0$ for all $\Phi^\alpha\in\mathscr{C}$ .

Under reasonable compatibility assumptions on $\mathscr{C}$ (cf. §3.4.1 in ^[18], see Definition 3.4 below), such a vector field exists and is unique. However, given two different realizations, $\mathscr{C}$ and $\mathscr{C}'$ , of the same constraint manifold $M$ , it is not generally true that $\Xi_{H}^\mathscr{C} = \Xi_H^{\mathscr{C}'}$ , however $\Xi_{H}^\mathscr{C}|_{M} = \Xi_H^{\mathscr{C}'}|_{M}$ .

Remark 3.3. The constraint manifold is given by the joint zero level-sets of the $\Phi^\alpha$ while the realization provides additional irrelevant information off of the constraint manifold. This is why $\Xi_H^\mathscr{C}\ne \Xi_H^{\mathscr{C}'}$ but they agree once restricted as will be proved in Proposition 3.6 below.

Definition 3.4. For a realization $\mathscr{C} = \{\Phi^1, \ldots, \Phi^k\}$ , the constraint mass matrix, $(m^{\alpha\beta})$ , is given by "orthogonally" pairing the constraints,

$\begin{equation*} m^{\alpha\beta} = \mathcal{C}^*d\Phi^\alpha\left( X_{\Phi^\beta}\right). \end{equation*}$

The realization is admissible if this matrix is non-singular on a tubular neighborhood of $M$ . The inverse matrix will be denoted by $(m_{\alpha\beta}) = (m^{\alpha\beta})^{-1}$ .

When the constraints are affine (along with a natural Lagrangian), the constraint mass matrix becomes

$\begin{equation*} m^{\alpha\beta} = g(W^\alpha,W^\beta) = \eta^\alpha(W^\beta), \end{equation*}$

and admissibility of the constraints is equivalent to the constraints being linearly independent.

We can now write down a formula for $\Xi_H^\mathscr{C}$ . Using (NH.1) and (NH.2), we get that (where $\{\cdot, \cdot\}$ is the standard Poisson bracket)

$\begin{equation*} \begin{split} \mathcal{L}_{\Xi_H^\mathscr{C}}\Phi^\beta & = i_{X_{\Phi^\beta}}\omega (\Xi_H^\mathscr{C}) \\ & = -i_{\Xi_H^\mathscr{C}}\omega(X_{\Phi^\beta}) \\ & = -dH\left( X_{\Phi^\beta}\right) - \lambda_\alpha \mathcal{C}^*d\Phi^\alpha(X_{\Phi^\beta}) \\ & = \left\{ \Phi^\beta,H \right\} - \lambda_\alpha \mathcal{C}^*d\Phi^\alpha(X_{\Phi^\beta}) = 0. \end{split} \end{equation*}$

This implies that $\left\{ \Phi^\beta, H \right\} = m^{\alpha\beta}\lambda_\alpha$ . Due to the constraint mass matrix being nondegenerate, the multipliers are uniquely determined and the extended nonholonomic vector field is determined by

$\begin{equation} { i_{\Xi_H^{\mathscr{C}}}\omega = dH - m_{\alpha\beta}\left\{H,\Phi^\alpha\right\} \mathcal{C}^*d\Phi^\beta } \end{equation}$

(3.1)

Definition 3.5. The 1-form on $T^*Q$ given by

$\begin{equation*} {\nu_H^\mathscr{C}} : = dH - m_{\alpha\beta}\left\{H,\Phi^\alpha\right\} \mathcal{C}^*d\Phi^\beta, \end{equation*}$

is called the nonholonomic 1-form with respect to the realization $\mathscr{C}$ .

Proposition 3.6. Given two different realization, $\mathscr{C}$ and $\tilde{\mathscr{C}}$ , the extended nonholonomic vector fields given by (3.1) agree on $M$ .

Proof. Let $\mathscr{C} = \{\Phi^\alpha\}$ and $\tilde{\mathscr{C}} = \{\tilde{\Phi}^\alpha\}$ be two different realizations of the same manifold $M$ and let $x\in M$ . Their differentials span the annihilator,

$\begin{equation*} \mathrm{span}\left\{ d\Phi^\alpha_x\right\} = \mathrm{span}\left\{d\tilde{\Phi}^\alpha_x\right\} = \mathrm{Ann}(T_xM). \end{equation*}$

As the map $\mathcal{C}:T(T^*Q)\to T(T^*Q)$ is linear, we have

$\begin{equation*} \mathrm{span}\left\{ \mathcal{C}^*d\Phi^\alpha_x \right\} = \mathrm{span}\left\{ \mathcal{C}^*d\tilde{\Phi}^\alpha_x \right\}. \end{equation*}$

As the multipliers are unique, we must have

$\begin{equation*} m_{\alpha\beta}\left\{H,\Phi^\alpha\right\}\mathcal{C}^*d\Phi^\beta|_M = \tilde{m}_{\alpha\beta}\left\{H, \tilde{\Phi}^\alpha\right\}\mathcal{C}^*d\tilde{\Phi}^\beta|_M, \end{equation*}$

and therefore the resulting extended nonholonomic vector fields must agree on $M$ .

Remark 3.7. A consequence of Proposition 3.6 is that this procedure of extending the nonholonomic vector field is still valid when $M$ is not globally defined as the level-set of functions. Suppose that $U, V\subset M$ are two neighborhoods characterized by the zero level-sets of functions $\Phi, \phi:T^*Q\to\mathbb{R}$ . Then $\Phi$ and $\phi$ both define different realizations on $U\cap V$ . However, their corresponding vector fields agree on the intersection. This observation will make the constructions required for Theorem 5.2 well-defined even if $M$ cannot be globally defined as a level set.

3.1. Stabilizing extended vector fields

The extended nonholonomic vector field, $X_H^M$ , is a vector field on the ambient space $T^*Q$ and has $M$ as an invariant manifold. As such, an integral curve of $X_H^M$ is only a valid trajectory if its initial condition lies in $M$ . In practice using numerical techniques, round off errors can result in trajectories drifting away from $M$ . To counter this effect, the condition (NH.2) can be replaced to include something in the spirit of a "control Lyapunov function" ^[33]. This modification will only be explored in this section while the rest of this paper will be concerned with the unmodified approach explained above.

Let the modified feedback nonholonomic criteria be:

(NH.1) $i_{\Xi_H^\mathscr{C}}\omega = dH + \lambda_\alpha \mathcal{C}^*d\Phi^\alpha$ , for smooth functions $\lambda_\alpha:T^*Q\to \mathbb{R}$ , and

(NH.2') $\mathcal{L}_{\Xi_H^\mathscr{C}}\Phi^\alpha = -\kappa^\alpha \Phi^\alpha$ for some positive constants $\kappa^\alpha$ ,

where there is no summation in (NH.2').

This approach has the advantage that if $x_0\not\in M$ , its trajectory exponentially approaches $M$ , i.e. suppose that $\varphi_t$ is the flow of the extended feedback nonholonomic vector field, then

$\begin{equation*} \Phi^\alpha\left( \varphi_t(x_0)\right) = e^{-\kappa^\alpha}\Phi^\alpha\left(x_0\right) \to 0. \end{equation*}$

Mimicking the derivation of (3.1), we arrive at

$\begin{equation} i_{\Xi_H^\mathscr{C}}\omega = dH - m_{\alpha\beta}\left( \left\{ H, \Phi^\alpha \right\} - \kappa^\alpha \Phi^\alpha\right)\cdot \mathcal{C}^*d\Phi^\beta. \end{equation}$

(3.2)

Example 1 (Simple Pendulum). Consider the simple pendulum viewed as a constrained system in $\mathbb{R}^2$ . The Hamiltonian and constraint are

$\begin{equation*} H = \frac{1}{2m}\left( p_x^2+p_y^2\right) + mgy, \quad \Phi = xp_x+yp_y = 0, \end{equation*}$

where $m$ is the mass of the pendulum and $g$ is the acceleration due to gravity. The equations of motion from (3.2) are

$\begin{equation} \begin{split} \dot{x} & = \frac{1}{m}p_x,\quad \dot{y} = \frac{1}{m}p_y, \\ \dot{p}_x & = \frac{x}{x^2+y^2}\left[ mgy - \frac{1}{m}(p_x^2+p_y^2) - \kappa\left( xp_x+yp_y\right) \right], \\ \dot{p}_y & = \frac{y}{x^2+y^2}\left[ mgy - \frac{1}{m}(p_x^2+p_y^2) - \kappa\left( xp_x+yp_y\right) \right] - mg. \end{split} \end{equation}$

(3.3)

A valid initial condition, $z_0 = (x(0), y(0), p_x(0), p_y(0))$ , for (3.3) needs to satisfy the constraint $\Phi(z_0) = 0$ . However, due to running numerical errors, the constraint will not be preserved as the state is integrated. By choosing $\kappa > 0$ , a correcting feedback is introduced to help preserve $\Phi = 0$ as shown in Figure 1.

Figure 1. The introduction of a positive

$\kappa$ helps to stabilize the numerical trajectory. Numerical integration was performed with Matlab's

$\texttt{ode45}$ .

DownLoad: Full-Size Img PowerPoint

3.2. Ideal constraints

We end this section with a brief consideration of ideal constraints - constraints that perform no work on the system. As discussed in ^[31], a constraint $\Psi$ on $TQ$ is ideal if and only if

$\begin{equation*} \left.\dot{q}^i\frac{\partial \Psi}{\partial \dot{q}^i}\right|_M = 0. \end{equation*}$

On the cotangent side, this condition becomes

$\begin{equation*} \left.\mathcal{C}^*d\Phi\left( X_H\right)\right|_M = 0. \end{equation*}$

In particular, if all the constraints are ideal, then energy is conserved. Indeed,

$\begin{equation*} \dot{H} = m_{\alpha\beta}\{H,\Phi^\alpha\}\mathcal{C}^*d\Phi^\beta(X_H). \end{equation*}$

Constraints will not be assumed to be ideal. Indeed, examples 7.1.2, 7.1.3, and 7.2.1 will all be subject to non-ideal constraints. In particular, these examples will be volume-preserving but not energy-preserving (although a modified energy integral exists for the first two examples ^[34]).

4. Nonholonomic volume

For a given nonholonomic vector field $X_H^M\in\mathfrak{X}(M)$ , a volume-form $\mu\in \Omega^{2n-k}(M)$ is invariant if and only if

$\begin{equation*} \mathrm{div}_\mu(X_H^M) = 0. \end{equation*}$

This section offers a way to choose a volume-form $\mu$ based off a (non-canonical choice) of realization $\mathscr{C}$ as well as showing how to compute the divergence of the nonholonomic vector fields with respect to this volume.

4.1. Nonholonomic volume form

The symplectic manifold $T^*Q$ has a canonical volume form $\omega^n$ . However, the nonholonomic flow takes place on a submanifold $M\subset T^*Q$ which is $2n-k$ dimensional. Therefore, $\omega^n$ is not a volume form on $M$ . Here, we construct a volume form on $M$ which is unique up to the choice of realization. The derivation of this will be similar to the construction of the volume form on an energy surface in §3.4 of ^[35]. For the realization $\mathscr{C} = \{\Phi^1, \ldots, \Phi^k\}$ , define the $k$ -form

$\begin{equation*} \sigma_\mathscr{C} : = d\Phi^1\wedge\ldots\wedge d\Phi^k. \end{equation*}$

Definition 4.1. If we denote the inclusion map by $\iota:M\hookrightarrow T^*Q$ , then a nonholonomic volume, $\mu_\mathscr{C}$ , is given by

$\begin{equation*} \mu_\mathscr{C} = \iota^*\varepsilon,\quad \sigma_\mathscr{C}\wedge\varepsilon = \omega^n. \end{equation*}$

Proposition 4.2. Given an ordered collection of constraints, $\mathscr{C}$ , the induced volume form $\mu_\mathscr{C}$ is unique.

Proof. Suppose that $\varepsilon$ and $\varepsilon'$ are two forms satisfying $\sigma_\mathscr{C}\wedge\varepsilon = \omega^n$ . Then

$\begin{equation*} \varepsilon - \varepsilon' = \alpha,\quad \sigma_{\mathscr{C}}\wedge\alpha = 0. \end{equation*}$

Now let $\iota:M\hookrightarrow T^*Q$ be the inclusion. Then from the above, we see that

$\begin{equation*} \iota^*\varepsilon = \iota^*\varepsilon' + \iota^*\alpha. \end{equation*}$

The result will follow so long as $\iota^*\alpha = 0$ . Suppose that $\iota^*\alpha\ne0$ and choose vectors $v^1, \ldots, v^{2n-k}\in T_xM\subset T_xT^*Q$ such that $\alpha(v^1, \ldots, v^{2n-k})\ne 0$ . Complete this collection of vectors to a basis of $T_xT^*Q$ : $v^1, \ldots, v^{2n-k}, v^{2n-k+1}, \ldots, v^{2n}$ such that $\sigma_{\mathscr{C}}(v^{2n-k+1}, \ldots, v^{2n})\ne 0$ . Then we have

$\begin{equation*} \sigma_{\mathscr{C}}\wedge \alpha \left(v^1,\ldots,v^{2n}\right) = (-1)^{(2n-k)k}\alpha(v^1,\ldots,v^{2n-k})\cdot\sigma_{\mathscr{C}}(v^{2n-k+1},\ldots,v^{2n}) \ne 0, \end{equation*}$

which is a contradiction.

Remark 4.3. Notice that for an ordered collection of constraints the volume form is unique. However, changing the order of the constraints changes the sign of the induced volume form and re-scaling constraints re-scales the volume form. In this sense, $\mathscr{C}$ uniquely determines $\mu_\mathscr{C}$ , but $M$ only determines $\mu_\mathscr{C}$ up to a multiple. This is to be expected as all volume-forms are related by a multiplicative factor.

While examining the failure of Liouville's theorem (Theorem 2.2) for nonholonomic systems, we will see when $\mu_\mathscr{C}$ is preserved under the flow of $X_H^M$ . More generally, we will consider the existence of a smooth density $f\in C^\infty(M)$ when $f\mu_\mathscr{C}$ is preserved.

4.2. Divergence

Let $\omega = dq^i\wedge dp_i$ be the standard symplectic form on $T^*Q$ . This in turn induces a volume form $\omega^n$ . A measure of how much a flow fails to preserve a volume form is described by its divergence. We proceed with computing the divergence of a nonholonomic vector field, $\mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M)$ . When this is nonzero, we will be interested in finding a density, $f$ , such that $\mathrm{div}_{f\mu_{\mathscr{C}}}(X_H^M) = 0$ . This problem will be addressed in §5.

Before we begin with the divergence calculation, we first present a helpful lemma which allows us to relate the divergence of the extended nonholonomic vector field with the corresponding restricted vector field.

Lemma 4.4. If $\mathscr{C}$ is a realization of the constraint manifold $M\subset T^*Q$ , then

$\begin{equation*} \left.\mathrm{div}_{\omega^n}\left(\Xi_H^{\mathscr{C}}\right)\right|_{M} = \mathrm{div}_{\mu_\mathscr{C}} \left(X_H^M\right). \end{equation*}$

Proof. Leibniz's rule for the Lie derivative provides

$\begin{equation*} \begin{split} \mathcal{L}_{\Xi_H^{\mathscr{C}}}\omega^n & = \mathcal{L}_{\Xi_H^{\mathscr{C}}}\left( \sigma_\mathscr{C}\wedge\varepsilon\right) \\ & = \left(\mathcal{L}_{\Xi_H^{\mathscr{C}}}\sigma_\mathscr{C}\right)\wedge\varepsilon + \sigma_\mathscr{C}\wedge\left(\mathcal{L}_{\Xi_H^{\mathscr{C}}}\varepsilon\right). \end{split} \end{equation*}$

However, $\mathcal{L}_{\Xi_H^{\mathscr{C}}}\sigma_\mathscr{C} = 0$ because the constraints are preserved under the flow. Applying this, we see that

$\begin{equation*} \mathcal{L}_{\Xi_H^{\mathscr{C}}}\omega^n = \sigma_\mathscr{C}\wedge \left(\mathcal{L}_{\Xi_H^{\mathscr{C}}}\varepsilon\right), \end{equation*}$

which gives

$\begin{equation*} \left(\mathrm{div}_{\omega^n}\left(\Xi_H^\mathscr{C}\right)\right) \sigma_{\mathscr{C}}\wedge\varepsilon = \sigma_{\mathscr{C}}\wedge\left( \mathcal{L}_{\Xi_H^\mathscr{C}}\varepsilon\right). \end{equation*}$

Due to the fact that the Lie derivative commutes with restriction, the result follows.

This lemma allows for us to calculate the divergence of the extended nonholonomic vector field and to restrict to the constraint distribution afterwards.

Using Cartan's magic formula along with Lemma 4.4, we see

$\begin{equation*} \begin{split} \mathcal{L}_{\Xi_H^\mathscr{C}}\left( \omega^n\right) & = i_{\Xi_H^\mathscr{C}}d\omega^n + di_{\Xi_H^\mathscr{C}}\omega^n \\ & = n\cdot d\left( i_{\Xi_H^\mathscr{C}}\omega\wedge\omega^{n-1}\right) \\ & = n\cdot d\left( i_{\Xi_H^\mathscr{C}}\omega\right)\wedge\omega^{n-1} - n\cdot\left( i_{\Xi_H^\mathscr{C}}\omega\right)\wedge d\omega^{n-1} \\ & = n\cdot d\nu_H^\mathscr{C}\wedge\omega^{n-1}. \end{split} \end{equation*}$

The divergence of the system is controlled by the failure of $\nu_H^\mathscr{C}$ to be closed. Its derivative is given by

$\begin{equation*} d\nu_H^\mathscr{C} = ddH - d\left(m_{\alpha\beta}\left\{ H, \Phi^\alpha\right\} \cdot \mathcal{C}^*d\Phi^\beta\right). \end{equation*}$

Let $\phi_\alpha = m_{\alpha\beta}\Phi^\beta$ . As $\Phi^\beta = 0$ on $M$ , the derivative becomes (where restriction to $M$ is implied)

$\begin{equation*} d\nu_H^\mathscr{C} = -d\{H,\phi_\beta\}\wedge \mathcal{C}^*d\Phi^\beta + \{H,\phi_\beta\}\cdot d\mathcal{C}^*d\Phi^\beta. \end{equation*}$

The form $d\nu_H^\mathscr{C}$ is a general 2-form. However, the only terms that survive once it is wedged with $\omega^{n-1}$ are the diagonal terms, $dq^i\wedge dp_i$ . Using local coordinates, $a_i^\beta dq^i = \mathcal{C}^*d\Phi^\beta$ , we have

$\begin{equation*} \left( d\nu_H^\mathscr{C}\right)_{\mathrm{diag}} = \left(\frac{\partial\{H,\phi_\beta\}}{\partial p_i}a_i^\beta - \{ H,\phi_\beta\} \frac{\partial a_i^\beta}{\partial p_i} \right) dq^i\wedge dp_i. \end{equation*}$

Therefore, the divergence is given by

$\begin{equation*} \mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M) = n\cdot \left(\frac{\partial\{H,\phi_\beta\}}{\partial p_i}a_i^\beta - \{ H,\phi_\beta\} \frac{\partial a_i^\beta}{\partial p_i} \right). \end{equation*}$

The first component of this expression can be written intrinsically. Notice that

$\begin{equation*} d\pi_Q\cdot X_f = \frac{\partial f}{\partial p_i}\frac{\partial}{\partial q^i}, \end{equation*}$

where $\pi_Q:T^*Q\to Q$ is the cotangent bundle projection. Therefore, the divergence can now be written as

$\begin{equation} \mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M) = -n\cdot \left(\mathcal{C}^*d\Phi^\beta\left( [X_H,X_{\phi_\beta}]\right) + \{ H,\phi_\beta\} \frac{\partial a_i^\beta}{\partial p_i} \right). \end{equation}$

(4.1)

4.3. Intrinsic forms of the divergence

Suppose the Lagrangian is natural with Riemannian metric $g = (g_{ij})$ . Then

$\begin{equation*} \mathcal{C}^*d\Phi^\beta = g_{ij}\frac{\partial \Phi^\beta}{\partial p_j} dx^i, \quad a_i^\beta = g_{ij}\frac{\partial \Phi^\beta}{\partial p_j}, \quad \frac{\partial a_i^\beta}{\partial p_i} = g_{ij}\frac{\partial^2 \Phi^\beta}{\partial p_i\partial p_j}. \end{equation*}$

Let us call this final double sum $\mathcal{M}^\beta$ . Applying this to (4.1), we have

$\begin{equation*} \mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M) = -n\cdot \mathcal{C}^*d\Phi^\beta\left( [X_H,X_{\phi_\beta}] \right) - n\cdot \left\{ H,\phi_\beta\right\}\mathcal{M}^\beta. \end{equation*}$

Next, assume that (in addition to the Lagrangian being natural), the constraints are affine which makes $\mathcal{M}^\beta = 0$ . Also the matrix, $m^{\alpha\beta}$ , does not depend on $p$ and therefore, the divergence can be written as

$\begin{equation} \mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M) = -n\cdot m_{\alpha\beta}\cdot \pi_Q^*\eta^\beta\left( [X_H, X_{\Phi^\alpha}] \right), \end{equation}$

(4.2)

as $\mathcal{C}^*d\Phi^\beta = \pi_Q^*\eta^\beta$ for affine constraints.

For affine systems with natural Lagrangians, the divergence can be expressed intrinsically via (4.2). To understand the divergence in the general nonlinear case (with potentially non-natural Lagrangians), we need an intrinsic way to interpret the term $\mathcal{M}^\beta$ . This can be accomplished via the following procedure:

The divergence for a nonholonomic system subject to nonlinear constraints is thus

$\begin{equation} \mathrm{div}_{\mu_\mathscr{C}}(X_H^M) = \dim(Q)\cdot \left( \mathcal{C}^*d\Phi^\beta\left( \left[ X_H, X_{\phi_\beta}\right] \right) + \left\{ H, \phi_\beta\right\} \Delta_\mathcal{C}\left(\Phi^\beta\right) \right), \end{equation}$

(4.3)

which is the key to proving Theorem 5.2 below.

5. Invariant volumes and the cohomology equation

In general, the divergence of a nonholonomic system does not vanish as (4.1) shows. When does there exist a different volume form on $M$ that is invariant under the flow? i.e. does there exist a density $\rho > 0$ such that $\mathrm{div}_{\rho\mu_{\mathscr{C}}}(X_H^M) = 0$ ? Finding such a $\rho$ requires solving a certain type of partial differential equation which is known as the smooth dynamical cohomology equation. Solving this PDE is generally quite difficult. However, when the constraints are affine and the density is assumed to be basic, the problem becomes considerably more tractable

5.1. The cohomology equation

What conditions need to be met for $\rho$ such that $\rho\mu_\mathscr{C}$ is an invariant volume form? Using the formula for the divergence as well as the fact that the Lie derivative is a derivation yields:

$\begin{equation*} \mathrm{div}_{\rho\mu_\mathscr{C}}\left(X_H^M\right) = \mathrm{div}_{\mu_\mathscr{C}}\left(X_H^M\right) + \frac{1}{\rho}\mathcal{L}_{X_H^M}(\rho). \end{equation*}$

Therefore the density, $\rho$ , yields an invariant measure if and only if

$\begin{equation} \frac{1}{\rho}\mathcal{L}_{X_H^M}(\rho) = -\mathrm{div}_{\mu_\mathscr{C}}\left( X_H^M\right). \end{equation}$

(5.1)

Notice that the left hand side of (5.1) can be integrated to

$\begin{equation*} \frac{1}{\rho}\mathcal{L}_{X_H^M}(\rho) = d\left(\ln \rho \right) \left( X_H^M\right). \end{equation*}$

Calling $\lambda = \ln \rho$ , we have the following lemma.

Lemma 5.1. For a nonholonomic vector field, $X_H^M$ , there exists a smooth invariant volume, $\rho\mu_\mathscr{C}$ , if there exists an exact 1-form $\alpha = d\lambda$ such that

$\begin{equation} \begin{split} \alpha\left( X_H^M\right) & = -\mathrm{div}_{\mu_\mathscr{C}}\left(X_H^M\right). \end{split} \end{equation}$

(5.2)

Then the density is (up to a multiplicative constant) $\rho = e^\lambda$ .

Therefore the existence of invariant volumes boils down to finding global solutions to the PDE (5.2). Using the divergence calculation, (4.3), we can state the main result.

Theorem 5.2. The nonholonomic Hamiltonian vector field $X_H^M$ possesses a smooth invariant volume, $\rho\mu_{\mathscr{C}}$ , if there exists an exact 1-form $\alpha\in\Omega^1(T^*Q)$ such that

$\begin{equation*} -\alpha\left( X_H^M\right) = \dim(Q)\cdot \left(\mathcal{C}^*d\Phi^\beta\left( [X_H,X_{\phi_\beta}]\right) + \{ H,\phi_\beta\} \Delta_\mathcal{C}\left(\Phi^\beta\right) \right). \end{equation*}$

The remainder of this section deals with uniqueness of solutions and a necessary condition for solutions to exist.

Remark 5.3. PDEs of the form $dg(X) = f$ for a given smooth function $f$ , vector field $X$ and with $g$ as the unknown are called cohomology equations ^[36,37]. Thus the equation (5.2) is a cohomology equation.

5.1.1. Uniqueness

The problem of existence is quite difficult in general and we postpone that discussion until the next subsection where we assume that the solution has the form $\rho = \pi_Q^*\tilde{\rho}$ . In the meantime, assuming that there exists a function $\lambda\in C^\infty(M)$ that solves (5.2), do there exist other solutions? Suppose that $\lambda_1$ and $\lambda_2$ both solve (5.2). Then their difference must be a first integral of the system: $\mathcal{L}_{X_H^M}(\lambda_1-\lambda_2) = 0$ . Solutions of (5.2) are then unique up to constants of motion. i.e. if $\lambda$ solves (5.2), then every invariant density has the form (again, up to a multiplicative constant)

$\begin{equation*} \rho = \exp\left( \lambda + \mathrm{constant\; of\; motion} \right). \end{equation*}$

Therefore invariant measures can be thought of as an affine space with dimension being equal to the number of first integrals of the nonholonomic system.

5.2. Special case: basic densities and affine constraints

In general, solving the cohomology equation (5.2) is quite difficult. It turns out, however, that it is relatively easy to determine necessary and sufficient conditions on the solvability when the density is assumed to be basic and the constraints are affine.

For the remainder of this section, unless otherwise stated, the constraints are assumed to be affine. In particular, if $N\subset TQ$ , then the constraints have the form

$\begin{equation*} \Psi^\alpha(v) = \eta^\alpha(v) + \tau_Q^*\xi^\alpha(v), \end{equation*}$

for 1-forms $\eta^\alpha$ and function $\xi^\alpha$ . On the cotangent side, the constraints become

$\begin{equation} \Phi^\alpha(p) = P(W^\alpha)(p) + \pi_Q^*\xi^\alpha(p), \end{equation}$

(5.3)

where $W^\alpha = \mathbb{F}L^{-1}\eta^\alpha$ .

Definition 5.4. A density $\rho:T^*Q\to\mathbb{R}$ is said to be basic if $\rho = \pi_Q^*\tilde{\rho}$ for some $\tilde{\rho}:Q\to\mathbb{R}$ .

Under this assumption, (4.2) can be presented in a surprisingly nice way. In this case, the divergence can be described by an equivalence class of "affine-forms." The density form and density function, defined below, is a representative element from this class.

Definition 5.5. Let $\mathscr{C}$ be an affine realization of $M\subset T^*Q$ of the form (5.3). Then, define the density form and the density function to be

$\begin{equation*} \begin{split} \vartheta_{\mathscr{C}} & = m_{\alpha\beta}\cdot\mathcal{L}_{W^\alpha}\eta^\beta, \\ \zeta_\mathscr{C} & = m_{\alpha\beta}\cdot\mathcal{L}_{W^\alpha}\xi^\beta, \end{split} \end{equation*}$

respectively.

Studying this pair, $(\vartheta_{\mathscr{C}}, \zeta_\mathscr{C})$ , provides necessary and sufficient conditions for the existence of basic densities.

Theorem 5.6. For a natural Hamiltonian system subject to affine constraints, there exists an invariant volume of the form $(\pi_Q^*\tilde{\rho})\mu_{\mathscr{C}}$ if and only if there exists functions $\varphi_\gamma$ such that

$\begin{equation} \begin{split} n\cdot \vartheta_{\mathscr{C}} - \varphi_\gamma\eta^\gamma & = -d\ln\tilde{\rho}, \\ n\cdot \zeta_\mathscr{C} - \varphi_\gamma\xi^\gamma & = 0. \end{split} \end{equation}$

(5.4)

Proof. To prove this, we will show that $-n\cdot\pi_Q^*\vartheta(X_H^M) - n\cdot \pi_Q^*\zeta_\mathscr{C} = \mathrm{div}_\mathscr{C}(X_H^M)$ . Recall that the differential of a 1-form is given by $d\alpha(X, Y) = X\alpha(Y)-Y\alpha(X) - \alpha([X, Y])$ and that $\pi_Q^*\eta^\beta(X_H)|_M = -\xi^\beta$ . Returning to (4.2), we have

$\begin{equation*} \begin{split} \mathrm{div}_{\mu_\mathscr{C}}(X_H^M) & = -n\cdot m_{\alpha\beta}\cdot \pi_Q^*\eta^\beta\left( [X_H, X_{\Phi^\alpha}] \right) \\ & = -n\cdot m_{\alpha\beta} \cdot \left( X_Hm^{\alpha\beta} + X_{\Phi^\alpha}\xi^\beta - \pi_Q^*d\eta^\alpha(X_H,X_{\Phi^\alpha}) \right) \\ & = -n\cdot m_{\alpha\beta} \cdot \left( dm^{\alpha\beta}(\dot{q}) + d\xi^\beta(W^\alpha) - d\eta^\alpha(\dot{q},W^\beta) \right) \\ & = -n\cdot m_{\alpha\beta}\cdot\left[ \left( di_{W^\beta}\eta^\alpha + i_{W^\beta}d\eta^\alpha\right)(\dot{q}) + d\xi^\beta(W^\alpha) \right] \\ & = -n\cdot m_{\alpha\beta}\cdot \left[ \mathcal{L}_{W^\beta}\eta^\alpha(\dot{q}) + \mathcal{L}_{W^\alpha}\xi^\beta \right] \\ & = -n\cdot\vartheta_{\mathscr{C}}(\dot{q}) - n\cdot\zeta_\mathscr{C}. \end{split} \end{equation*}$

This computation shows that $\mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M) = -n\cdot\vartheta_{\mathscr{C}}(\dot{q})-n\cdot\zeta_\mathscr{C}$ , but $\dot{q}$ cannot be arbitrary as it must lie within $M$ which states that $\eta^\alpha(\dot{q}) + \xi^\alpha = 0$ . Adding multiples of the constraints to the divergence to produce an exact 1-form yields (5.4).

This theorem allows for a straight-forward algorithm to find invariant volumes in nonholonomic systems subject to affine constraints; one only needs to compute the pair $(\vartheta_{\mathscr{C}}, \zeta_\mathscr{C})$ and determine whether or not it can be made exact by appending constraints to it. This procedure will be carried out on multiple examples in §7.

Remark 5.7. In the pure kinetic energy case with linear constraints discussed in ^[24], it is proved that if the system admits an (arbitrary) invariant volume, then one can always find another invariant volume form whose density function depends only on the (reduced) configuration variables. Moreover, systems subjected to affine constraints may possess an invariant volume whose density is not basic, cf. ^[38].

The above shows that "exactness" of $(\vartheta_{\mathscr{C}}, \zeta_\mathscr{C})$ determines the existence of a density depending on configuration. How does this depend on the choice of $\mathscr{C}$ to realize the constraints? It turns out the answer is independent of the choice of realization.

Theorem 5.8. Let $\mathscr{C}_1$ and $\mathscr{C}_2$ both be affine realizations of the constraint manifold $M$ . Suppose that there exist functions $\varphi_\gamma$ such that

$\begin{equation*} \vartheta_{\mathscr{C}_1} - \varphi_\gamma\eta^\gamma_1 = -d\ln\tilde{\rho}_1, \quad \zeta_{\mathscr{C}_1} - \varphi_\gamma\xi^\gamma_1 = 0, \end{equation*}$

then there exists functions $\psi_\gamma$ such that

$\begin{equation*} \vartheta_{\mathscr{C}_2} - \psi_\gamma\eta^\gamma_2 = -d\ln\tilde{\rho}_2, \quad \zeta_{\mathscr{C}_2} - \psi_\gamma\xi^\gamma_2 = 0. \end{equation*}$

Moreover, $(\pi_Q^*\tilde{\rho}_1)\mu_{\mathscr{C}_1} = (\pi_Q^*\tilde{\rho}_2)\mu_{\mathscr{C}_2}$ modulo a constant of motion.

Proof. Let $\eta_2^\gamma = c_\alpha^\gamma\eta_1^\alpha$ and $\xi_2^\gamma = c_\alpha^\gamma \xi_1^\alpha$ where $c_\alpha^\gamma$ is the required coordinate change. For ease of notation, we will drop the subscript "1." The matrices are related via

$\begin{equation*} m_2^{\alpha\beta} = c_\gamma^\alpha c_\delta^\beta m_1^{\gamma\delta}. \end{equation*}$

The density form is

$\begin{equation*} \begin{split} m_2^{\alpha\beta}\vartheta_2 & = \mathcal{L}_{c_\gamma^\alpha W^\gamma} \left(c_\delta^\beta\eta^\delta\right) \\ & = c_\gamma^\alpha c_\delta^\beta \mathcal{L}_{W^\gamma}\eta^\delta + c_\delta^\beta m_1^{\delta\gamma}dc_\gamma^\alpha + c_\gamma^\alpha dc_\delta^\beta(W^\gamma)\eta^\delta \\ & = m_2^{\alpha\beta}\vartheta_1 + m_2^{\alpha\beta}\sigma_\alpha^\gamma dc_\gamma^\alpha + c_\gamma^\alpha dc_\delta^\beta(W^\gamma)\eta^\delta, \end{split} \end{equation*}$

where $(\sigma_\alpha^\gamma) : = (c_\gamma^\alpha)^{-1}$ . By Lemma 5.9 below, we have

$\begin{equation} m_2^{\alpha\beta}\vartheta_2 = m_2^{\alpha\beta}\vartheta_1 + m_2^{\alpha\beta} d\left[ \ln \det(c_\gamma^\alpha)\right] + c_\gamma^\alpha dc_\delta^\beta(W^\gamma)\eta^\delta. \end{equation}$

(5.5)

This shows that $\vartheta_1$ and $\vartheta_2$ differ by something exact and a multiple of the constraining 1-forms. Using the fact that $\vartheta_1 - \varphi_\gamma\eta_1^\gamma = -d \ln\tilde{\rho}_1$ , we see that

$\begin{equation*} \vartheta_2 - \left( \varphi_\gamma + C_\gamma\right) \eta^\gamma = d\left[ -\ln\tilde\rho_1 + \ln\det(c_\gamma^\alpha) \right], \end{equation*}$

where $m_2^{\alpha\beta}C_\gamma = c_\delta^\alpha dc_\gamma^\beta(W^\delta)$ . This provides

$\begin{equation*} \tilde{\rho}_2 = \det(\sigma_\alpha^\gamma)\tilde{\rho}_1, \quad \psi_\gamma = c_\gamma^\delta \left[ \varphi_\delta + C_\delta\right]. \end{equation*}$

These values of $\psi_\gamma$ make $\vartheta_2$ exact. We next check that these values of $\psi_\gamma$ make $\zeta_2$ vanish.

$\begin{equation*} \begin{split} m_2^{\alpha\beta}\left(\zeta_2 - \psi_\gamma c_\delta^\gamma \xi^\delta\right) & = c_\gamma^\alpha c_\delta^\beta \mathcal{L}_{W^\gamma}\xi^\delta + c_\gamma^\alpha dc_\delta^\beta(W^\gamma)\xi^\delta - m_2^{\alpha\beta}\psi_\gamma c_\delta^\gamma \xi^\delta \\ & = m_2^{\alpha\beta}\zeta_1 - m_2^{\alpha\beta}\varphi_\delta\xi^\delta = 0. \end{split} \end{equation*}$

It remains to show that $(\pi_Q^*\tilde{\rho}_1)\mu_{\mathscr{C}_1} = (\pi_Q^*\tilde{\rho}_2)\mu_{\mathscr{C}_2}$ . This is equivalent to $\mu_{\mathscr{C}_1} = (\pi_Q^*\det(c_\gamma^\alpha))\mu_{\mathscr{C}_2}$ . Recalling Definition 4.1, we have $\sigma_1 = d\Phi^1\wedge\ldots d\Phi^k$ and

$\begin{equation*} \sigma_2 = \bigwedge\limits_{i = 1}^k \, d\left( c_\gamma^i \Phi^\gamma\right) = \det(c_\gamma^\alpha) \sigma_1 + \bigwedge\limits_{i = 1}^k \, \Phi^\gamma dc_\gamma^i. \end{equation*}$

Upon applying the constraints, the last term disappears and we obtain

$\begin{equation*} \sigma_1\wedge\varepsilon_1 = \sigma_2\wedge\varepsilon_2 = \det(c_\gamma^\alpha)\sigma_1\wedge\varepsilon_2. \end{equation*}$

This implies the desired result.

5.2.1. Invariant volumes of holonomic systems

A reason why studying $(\vartheta_\mathscr{C}, \zeta_\mathscr{C})$ is insightful is that it immediately demonstrates why holonomic systems systems are measure-preserving, i.e. when $\eta^\alpha = df^\alpha$ and $\xi^\alpha = 0$ , (5.4) can always be solved. This short section idemonstrates how our general theory collapses to the known holonomic case. We start with a helpful lemma.

Lemma 5.9 (^*). $m_{\alpha\beta}\cdot dm^{\alpha\beta} = d\left[ \ln \det \left(m^{\alpha\beta}\right)\right]$ .

^*We thank Dr. Alexander Barvinok for help with this proof.

Proof. It suffices to check along a curve in the manifold. Let $\gamma:I\to Q$ be a curve and let $A(t) = \left(m^{\alpha\beta}\right)\circ\gamma(t)$ be the mass matrix along the curve. Note that $A(t)$ is positive-definite and changes smoothly with $t$ . We have

$\begin{equation*} \frac{d}{dt}\ln\det A(t) = \frac{\frac{d}{dt}\det A(t)}{\det A(t)} = \sum\limits_{i = 1}^m \, \frac{\det A_i(t)}{\det A(t)}, \end{equation*}$

where $A_i(t)$ is obtained from $A(t)$ by differentiating the $i$ -th row and leaving all other rows intact, i.e.

$\begin{equation*} A_i(t) = \left( \begin{array}{ccc} a_{11}(t) & \cdots & a_{1m}(t) \\ \vdots & \ddots & \vdots \\ a_{(i-1)1}(t) & \cdots & a_{(i-1)m}(t) \\ a'_{i1}(t) & \cdots & a'_{im}(t) \\ a_{(i+1)1}(t) & \cdots & a_{(i+1)m}(t) \\ \vdots & \ddots & \vdots \\ a_{m1}(t) & \cdots & a_{mm}(t) \end{array}\right). \end{equation*}$

Expanding $\det A_i(t)$ along the $i$ -th row:

$\begin{equation*} \det A_i(t) = \sum\limits_{j = 1}^m \, (-1)^{i+j-1} a'_{ij}(t)\det A_{ij}(t), \end{equation*}$

where $A_{ij}(t)$ is the $(m-1)\times (m-1)$ matrix obtained from $A_i(t)$ and hence from $A(t)$ by crossing out the $i$ -th row and $j$ -th column.

Next, observe that $(-1)^{i+j-1}\det A_{ij}/\det A(t)$ is the $(j, i)$ -th entry of the inverse matrix $A^{-1}(t) = (b_{ij})(t)$ , and since $A(t)$ is symmetric, is also the $(i, j)$ -th entry of $(b_{ij})(t)$ . Summarizing,

$\begin{equation*} \frac{d}{dt}\ln\det A(t) = \sum\limits_{i,j = 1}^m \, a_{ij}'(t)b_{ij}(t). \end{equation*}$

Proposition 5.10. If the constraints are holonomic, then there exist function $\varphi_\gamma$ such that $\vartheta_{\mathscr{C}}-\varphi_\gamma\eta^\gamma$ is exact. In particular, if $\mathscr{C}$ is chosen such that all $\eta^\alpha$ are closed, $\vartheta_{\mathscr{C}}$ is exact.

Proof. When the constraints are holonomic, the 1-forms $\eta^\alpha$ can be chosen such that they are closed. Then the density form is

which is exact by Lemma 5.9. If a different realization is chosen, Theorem 5.8 states that the resulting (5.4) is still solvable.

6. Connections with the nonholonomic connection

It turns out that for natural Lagrangian systems subject to linear constraints, the divergence - particularly the density form - is encoded in the nonholonomic connection. This interpretation seems to be new.

Throughout this section, let $L:TQ\to\mathbb{R}$ be a natural Lagrangian with Riemannian metric $g$ subject to the linear constraints $\eta^\alpha(v) = 0$ . The nonholonomic connection for this system is given by (cf. §5.3 in ^[20] and ^[39]):

$\begin{equation*} \nabla^\mathscr{C}_XY = \nabla_XY + W^i\cdot m_{ij}\left[ X\left(\eta^j(Y)\right) - \eta^j\left(\nabla_XY\right) \right]. \end{equation*}$

The equations of motion can then be described via

$\begin{equation*} \nabla^\mathscr{C}_{\dot{q}}\dot{q} = F, \end{equation*}$

where $F$ contains the potential and external forces (the constraint forces are contained in the connection).

6.1. Torsion

The nonintegrability of the constraints appears in the torsion of the connection. Computing this, we see

$\begin{equation*} \begin{split} T^\mathscr{C}(X,Y) & = \nabla^\mathscr{C}_XY - \nabla^\mathscr{C}_YX - [X,Y] \\ & = W^i\cdot m_{ij} \left[ X(\eta^j(Y))-Y(\eta^j(X)) - \eta^j(\nabla_XY-\nabla_YX) \right] \\ & = W^j\cdot m_{ij}\left[ X(\eta^j(Y))-Y(\eta^j(X))-\eta^j([X,Y]) \right] \\ & = W^j\cdot m_{ij} \cdot d\eta^j(X,Y). \end{split} \end{equation*}$

The torsion can be written as

$\begin{equation*} T^\mathscr{C} = m_{\alpha\beta}\cdot W^\alpha \otimes d\eta^\beta. \end{equation*}$

Indeed, if the constraining 1-forms $\eta^j$ are all closed (so holonomic) then the torsion vanishes. It is worth pointing out that the torsion is vertical-valued; if $X, Y\in M$ , then $T^\mathscr{C}(X, Y)\in M^\perp$ , i.e. $T(X, Y)$ is orthogonal to the constraint distribution.

Due to the fact that the torsion is a (1, 2)-tensor, its trace will be a (0, 1)-tensor. Therefore, the trace of the nonholonomic torsion will be a 1-form:

$\begin{equation*} \begin{split} \text{tr} T^\mathscr{C} & = m_{\alpha\beta}\cdot i_{W^\alpha}d\eta^\beta. \end{split} \end{equation*}$

Returning to the density form, we see that

$\begin{equation*} \text{tr} T^\mathscr{C} + d\ln\det\left(m^{\alpha\beta}\right) = \vartheta_\mathscr{C}, \end{equation*}$

i.e. the trace of the torsion differs from the density form by something exact. This leads to the following theorem.

Theorem 6.1. A natural nonholonomic system subject to linear constraints has an invariant volume of the form $(\pi_Q^*\rho)\cdot\mu_{\mathscr{C}}$ if and only if there exists functions $\varphi_\gamma$ such that

$\begin{equation*} \text{tr} T^\mathscr{C} + \varphi_\gamma\eta^\gamma \end{equation*}$

is exact.

Remark 6.2. The vanishing of the torsion shows that the constraints are integrable while the integrability of the (trace of the) torsion shows that a volume is preserved.

In the case of nonholonomic systems, the nonholonomic connection is compatible with the metric but has nonzero torsion. This idea extends to arbitrary, metric-compatible connections as the following theorem states.

Theorem 6.3. Let $\tilde{\nabla}$ be an affine connection compatible with the metric with torsion $\tilde{T}$ . There exists an invariant volume with density of the form $\pi_Q^*\rho$ for the geodesic spray if and only if $\text{tr}\tilde{T}$ is exact.

Proof. Consider the volume form on $TQ$ given by

$\begin{equation*} \Omega = \det g \cdot dx^1\wedge\ldots\wedge dx^n\wedge dv^1\wedge\ldots\wedge dv^n. \end{equation*}$

We want to compute $\mathcal{L}_X\Omega$ where $X$ is the geodesic spray given by

$\begin{equation*} X = v^i \frac{\partial}{\partial x^i} - \Gamma^i_{jk}v^jv^k\frac{\partial}{\partial v^i}. \end{equation*}$

The Lie derivative is then

$\begin{equation*} \begin{split} \mathcal{L}_X\Omega & = di_X\Omega \\ & = \left(d\left[ \det g\right](v) - \det g \left(\Gamma_{ik}^i+\Gamma_{ki}^i\right)v^k\right) \cdot \frac{1}{\det g} \Omega, \end{split} \end{equation*}$

and therefore the divergence is given by

$\begin{equation} \mathrm{div}_\Omega(X) = d\left[\ln\det g\right](v) - \left(\Gamma_{ik}^i+\Gamma_{ki}^i\right) v^k. \end{equation}$

(6.1)

We will now use the fact that the connection is compatible with the metric:

$\begin{equation*} \frac{\partial g_{jk}}{\partial x^i} = g_{\ell k}\Gamma^\ell_{ij} + g_{j\ell}\Gamma_{ik}^\ell. \end{equation*}$

This implies that

$\begin{equation*} \begin{split} g^{jk}\frac{\partial g_{jk}}{\partial x^i} & = g^{jk}g_{\ell k}\Gamma_{ij}^\ell + g^{jk}g_{j\ell}\Gamma_{ik}^\ell \\ & = \delta_{\ell}^j\Gamma_{ij}^\ell + \delta_{\ell}^k\Gamma_{ij}^\ell = 2\Gamma_{ik}^k. \end{split} \end{equation*}$

Integrating the left-hand side above gives

$\begin{equation} d\left[ \ln\det g\right](v) = 2\Gamma_{ki}^i v^k. \end{equation}$

(6.2)

Substituting (6.2) into (6.1), we get

$\begin{equation*} \mathrm{div}_\Omega(X) = \left( \Gamma_{ki}^i - \Gamma_{ik}^i\right) v^k. \end{equation*}$

It remains to show that this is the trace of the torsion. Indeed,

$\begin{gather*} \tilde{T} = \left( \Gamma_{ij}^k - \Gamma_{ji}^k\right) \frac{\partial}{\partial x^k}\otimes dx^i \otimes dx^j \\ \implies \text{tr} \tilde{T} = \left(\Gamma_{ki}^i - \Gamma_{ik}^i \right) dx^k. \end{gather*}$

We conclude that

$\begin{equation*} \mathrm{div}_\Omega(X) = \text{tr}\tilde{T}(v). \end{equation*}$

This shows that a way to interpret the torsion of a connection is by measuring how much the geodesic spray fails to preserve volume.

7. Examples

We end this work with applying both Theorem 5.2 and Theorem 5.6 to various nonholonomic systems by calculating the general divergence for nonlinear systems or $\vartheta_{\mathscr{C}}$ and $\zeta_\mathscr{C}$ for affine/linear systems. Examples are taken from ^[18,20,31].

7.1. Affine constraints

We begin by examining multiple nonholonomic systems subject to affine/linear constraints.

7.1.1. The chaplygin sleigh

As an example of Theorem 5.6, we will prove that no invariant basic volumes exist for the Chaplygin sleigh. The Chaplygin sleigh has the configuration space $Q = \mathrm{SE}_2$ , the special Euclidean group, and the following Lagrangian

$\begin{equation*} L = \frac{1}{2}\left( m\dot{x}^2 + m\dot{y}^2 + \left(I+ma^2\right)\dot{\theta}^2 - 2ma\dot{x}\dot{\theta}\sin\theta + 2ma\dot{y}\dot{\theta}\cos\theta \right), \end{equation*}$

where $(x, y)\in\mathbb{R}^2$ is the coordinate of the contact point, $\theta\in\mathrm{SO}_2$ is its orientation, $m$ is the sleigh's mass, $I$ is the moment of inertia about the center of mass, and $a$ is the distance from the center of mass to the contact point (cf. §1.7 in ^[20]).

The nonholonomic constraint is that the sleigh can only slide in the direction it is pointing and is given by

$\begin{equation*} \dot{y}\cos\theta - \dot{x}\sin\theta = 0, \end{equation*}$

which corresponds to the 1-form $\eta = \left(\cos\theta\right)dy - \left(\sin\theta\right)dx$ and function $\xi = 0$ .

To determine the existence of an invariant volume, we only need to compute $\vartheta_{\mathscr{C}}$ as $\zeta_\mathscr{C} = 0$ . The constraining vector field and 1-form are:

$\begin{equation*} W = \frac{ma^2+I}{Im}\left[ \cos\theta \frac{\partial}{\partial y} - \sin\theta\frac{\partial}{\partial x} \right] - \frac{a}{I}\frac{\partial}{\partial\theta}, \quad \eta = (\cos\theta)dy - (\sin\theta)dx. \end{equation*}$

This gives us

$\begin{equation*} \begin{split} \vartheta_{\mathscr{C}} & = \frac{1}{\eta(W)}\mathcal{L}_W\eta \\ & = \frac{ma}{ma^2+I}\left[(\sin\theta)dy+(\cos\theta)dx\right]. \end{split} \end{equation*}$

As a consequence of this, the divergence of the Chaplygin sleigh is given by

$\begin{equation} \mathrm{div}_{\mu_{\mathscr{C}}}(X_H^\mathcal{D}) = -\frac{3mav}{I+ma^2}, \quad v = \dot{x}\cos\theta + \dot{y}\sin\theta. \end{equation}$

(7.1)

We want to show that for any $\tilde{\eta}\in\Gamma(\mathcal{D}^0)$ , $\vartheta_{\mathscr{C}}+\tilde{\eta}$ is not exact. Because there is only one constraint, it suffices to show that there does not exist a smooth $k$ such that $\vartheta_{\mathscr{C}}+k\cdot\eta$ is closed, i.e. it requires the following to be zero:

$\begin{equation*} \begin{split} d\left( \vartheta_{\mathscr{C}} + k\cdot\eta\right) & = \frac{ma}{ma^2+I}\left[ (\cos\theta)d\theta\wedge dy - (\sin\theta)d\theta\wedge dx \right] \\ &\quad + \left(\frac{\partial k}{\partial x}\cos\theta + \frac{\partial k}{\partial y}\sin\theta\right) dx\wedge dy \\ &\quad + \left(\frac{\partial k}{\partial\theta}\cos\theta - k\sin\theta\right) d\theta\wedge dy \\ &\quad -\left(\frac{\partial k}{\partial\theta}\sin\theta + k\cos\theta\right) d\theta\wedge dx. \end{split} \end{equation*}$

Separating the above, we need the following three to vanish:

$\begin{equation} \begin{split} 0 & = \frac{\partial k}{\partial x}\cos\theta + \frac{\partial k}{\partial y}\sin\theta, \\ 0 & = \frac{\partial k}{\partial\theta}\cos\theta - k\sin\theta + \frac{ma}{ma^2+I}\cos\theta, \\ 0 & = \frac{\partial k}{\partial\theta}\sin\theta + k\cos\theta + \frac{ma}{ma^2+I}\sin\theta. \end{split} \end{equation}$

(7.2)

The second two lines of (7.2) are overdetermined for $k$ in the $\theta$ -direction and are inconsistent (unless $a = 0$ and we obtain the trivial solution $k\equiv 0$ ). Therefore, there does not exist a smooth $k$ such that $\vartheta_{\mathscr{C}}+k\cdot\eta$ is closed. We note that this is compatable with the known result that when $a = 0$ , no asymptotically stable dynamics occur.

7.1.2. The falling rolling disk

The next example is that of the falling rolling disk whose configuration space is $Q = \mathrm{SE}_2\times S^1$ . Its Lagrangian is

$\begin{equation*} L = \frac{m}{2}\left[ \left(\xi - R\left(\dot\varphi\sin\theta+\dot\psi\right)\right)^2 + \eta^2\sin^2\theta + \left(\eta\cos\theta + R\dot\theta\right)^2\right], \end{equation*}$

where

$\begin{equation*} \xi = \dot{x}\cos\varphi + \dot{y}\sin\varphi + R\dot\psi, \quad \eta = -\dot{x}\sin\varphi + \dot{y}\cos\varphi. \end{equation*}$

We will consider both the case of the rolling coin on a stationary table (linear constraints) and on a rotating table (affine constraints).

Stationary Table The case of the disk on a stationary table was studied in ^[40]. The constraints are given by the vanishing of the following 1-forms:

$\begin{equation} \begin{split} \eta^1 & = \cos\varphi\cdot dx + \sin\varphi\cdot dy + R\cdot d\psi, \\ \eta^2 & = -\sin\varphi\cdot dx + \cos\varphi\cdot dy. \end{split} \end{equation}$

(7.3)

The corresponding dual vector fields are

$\begin{equation*} \begin{split} W^1 & = \frac{1}{m}\cos\varphi\frac{\partial}{\partial x} + \frac{1}{m}\sin\varphi\frac{\partial}{\partial y} + \frac{R}{I}\frac{\partial}{\partial \psi}, \\ W^2 & = \frac{J+mR^2}{Jm+m^2R^2\sin^2\theta}\left[ -\sin\varphi\frac{\partial}{\partial x} + \cos\varphi\frac{\partial}{\partial y} \right] - \frac{R\cos\theta}{Jm+m^2R^2\sin^2\theta}\frac{\partial}{\partial\theta}. \end{split} \end{equation*}$

Computing $\vartheta_{\mathscr{C}}$ , we obtain

$\begin{equation*} \vartheta_{\mathscr{C}} = -\frac{mR^2\sin(2\theta)}{J+mR^2\sin^2\theta}, \end{equation*}$

which is exact. The resulting invariant volume is

$\begin{equation*} \frac{1}{J+mR^2\sin^2\theta}\mu_{\mathscr{C}}. \end{equation*}$

Rotating Table Although the case of the stationary table has been studied, the authors are unaware of any results for the case of a rotating table. If the falling rolling disk is placed on a table with constant angular velocity $\Omega$ , the constraints become affine with

$\begin{equation*} \begin{split} \eta^1(v) + \xi^1 & = 0, \quad \xi^1 = \Omega\left( y\cos\varphi - x\sin\varphi\right) \\ \eta^2(v) +\xi^2 & = 0, \quad \xi^2 = - \Omega(x\cos\varphi+y\sin\varphi) \end{split} \end{equation*}$

where $\eta^\alpha$ are from (7.3). The volume from the stationary table is still preserved as

$\begin{equation*} \mathcal{L}_{W^1}\xi^1 = \mathcal{L}_{W^2}\xi^2 = 0. \end{equation*}$

7.1.3. The chaplygin sphere

We next consider the case of a non-homogeneous sphere rolling without slipping on a horizontal plane, both stationary and rotating. The center of mass of the sphere is located at its geometric center while its principal moments of inertia are distinct. This example has been studied by Chaplygin ^[41] and, e.g. ^[1,42,43].

The Lagrangian is the kinetic energy,

$\begin{equation*} \begin{split} L & = \frac{1}{2}I_1 \left(\dot{\theta}\cos\psi + \dot{\varphi}\sin\psi\sin\theta\right)^2 + \frac{1}{2}I_2 \left( -\dot{\theta}\sin\psi + \dot{\varphi}\cos\psi\sin\theta\right)^2 \\ &\quad + \frac{1}{2}I_3\left(\dot{\psi}+\dot{\varphi}\cos\theta\right)^2 + \frac{1}{2}M\left(\dot{x}^2+\dot{y}^2\right), \end{split} \end{equation*}$

where it is assumed that the radius is 1.

Stationary Table When the table is stationary, the constraints are given by the two 1-forms

$\begin{equation*} \begin{split} \eta^1 & = dx - \sin\varphi\cdot d\theta + \cos\varphi\sin\theta\cdot d\psi, \\ \eta^2 & = dy + \cos\varphi\cdot d\theta + \sin\varphi\sin\theta \cdot d\psi. \end{split} \end{equation*}$

The density form is given by $\vartheta_{\mathscr{C}} = A\cdot d\theta + B\cdot d\psi$ where

$\begin{equation*} \begin{split} A & = \frac{M\sin(2\theta)\left[J_1+J_2\sin^2\psi \right]} { 2\left( J_3 + J_4\sin^2\theta + J_5\sin^2\theta \sin^2\psi \right)}, \\ B & = \frac{MJ_2\sin(2\psi)\sin^2\theta} { 2\left( J_3 + J_4\sin^2\theta + J_5\sin^2\theta \sin^2\psi \right)}, \end{split} \end{equation*}$

and the constants $J_j$ are

$\begin{equation*} \begin{split} J_1 & = I_1I_2-I_1I_3+I_2M-I_3M, \\ J_2 & = I_1I_3 - I_2I_3 + I_1M - I_2M, \\ J_3 & = I_3M^2 + I_1I_2I_3 + I_1I_3M + I_2I_3M, \\ J_4 & = I_2M^2 -I_3M^2 + I_1I_2M-I_1I_3M, \\ J_5 & = I_1M^2 - I_2M^2 + I_1I_3M - I_2I_3M. \end{split} \end{equation*}$

The density form is exact and the resulting invariant volume is

$\begin{equation*} \begin{gathered} \left(\frac{1+\beta}{1-\beta}\right)^{\frac{1}{2}} \cdot\cos\theta \cdot \left(J_3 + (J_3+J_4)\tan^2\theta\right)^{\frac{1}{2}}\mu_{\mathscr{C}}, \\ \beta = \frac{J_5\sin^2\theta\sin^2\psi}{2J_3+2J_4\sin^2\theta + J_5\sin^2\theta\sin^2\psi}. \end{gathered} \end{equation*}$

Rotating Table When the table is rotating, the constraints becomes affine

$\begin{equation*} \begin{split} \eta^1(v) + \xi^1 & = 0, \quad \xi^1 = \Omega y, \\ \eta^2(v) + \xi^2 & = 0, \quad \xi^2 = -\Omega x. \end{split} \end{equation*}$

Notice that,

$\begin{equation*} \mathcal{L}_{W^1}\xi^1 = \mathcal{L}_{W^2}\xi^2 = 0, \quad \mathcal{L}_{W^2}\xi^1 = -\mathcal{L}_{W^1}\xi^2 = \frac{\Omega}{M}. \end{equation*}$

As the matrix $(m_{\alpha\beta})$ is symmetric, we have that the product

$\begin{equation*} m_{\alpha\beta}\mathcal{L}_{W^\alpha}\xi^\beta = 0. \end{equation*}$

Therefore, the Chaplygin sphere on a rotating table is volume-preserving with the same volume as in the stationary case.

7.1.4. The möbius strip

Theorem 5.2 does not require that $Q$ be orientable. Consider the Möbius strip immersed in $\mathbb{R}^3$ by

$\begin{equation} \begin{split} x & = \left( 1 + v\cdot \cos\left( \frac{u}{2}\right)\right) \cdot \cos(u), \\ y & = \left( 1 + v\cdot \cos\left(\frac{u}{2}\right)\right)\cdot \sin(u), \\ z & = v\cdot\sin\left(\frac{u}{2}\right), \end{split} \end{equation}$

(7.4)

for $0\leq u\leq 2\pi$ and $-1/2 < v < 1/2$ . The Euclidean metric pulled back to the Möbius strip is

$\begin{equation*} g = \left( 4v\cos\left(\frac{u}{2}\right) + 2v^2\cos\left(\frac{u}{2}\right) + \frac{v^2}{4} + 2\right) du\otimes du + 2dv\otimes dv. \end{equation*}$

Notice that the above metric is for the double cover rather than the Möbius strip itself as a consequence of (7.4) only being an immersion. However, the resulting equations of motion will be well-defined on the strip. In two-dimensional systems, any single constraint is automatically holonomic which will make volume-preservation trivial. Let us "thicken" the strip by $w$ with resulting metric

$\begin{equation*} g_{thick} = \left( 4v\cos\left(\frac{u}{2}\right) + 2v^2\cos\left(\frac{u}{2}\right) + \frac{v^2}{4} + 2\right) du\otimes du + 2dv\otimes dv + dw\otimes dw. \end{equation*}$

Consider the linear nonholonomic constraint

$\begin{equation*} \eta = dv + \sin(u)\cdot dw, \quad W = \frac{1}{2}\frac{\partial}{\partial v} + \frac{1}{2}\sin(u)\frac{\partial}{\partial w}. \end{equation*}$

The density form is

$\begin{equation*} \vartheta_\mathscr{C} = \frac{\sin(u)\cos(u)}{1+\sin^2(u)}\, du, \end{equation*}$

which is exact. This produces the invariant volume

$\begin{equation*} \sqrt{ 1 + \sin^2(u)}\cdot \mu_{\mathscr{C}}, \end{equation*}$

which is shown in Figure 2.

Figure 2. A plot of the density corresponding to the invariant volume for the nonholonomic system on the Möbius strip.

DownLoad: Full-Size Img PowerPoint

7.2. Nonlinear constraints

The previous examples consisted of a natural Lagrangian and affine constraints which were describable by Theorem 5.6. The next two examples utilize Theorem 5.2 on systems with natural Lagrangians but nonlinear constraints. These examples can be found in ^[31,44].

7.2.1. Constant kinetic energy

Let $L:T\mathbb{R}^3\to\mathbb{R}$ be the Lagrangian given by

$\begin{equation*} L = \frac{1}{2}m\left( \dot{x}^2 + \dot{y}^2 + \dot{z}^2\right) - mgz, \end{equation*}$

subject to the nonlinear constraint of constant kinetic energy

$\begin{equation*} \Psi = \dot{x}^2 + \dot{y}^2 + \dot{z}^2 - c = 0, \quad c > 0. \end{equation*}$

Transferring to the Hamiltonian side, we have

$\begin{equation*} H = \frac{1}{2m}\left( p_x^2 + p_y^2 + p_z^2\right) + mgz, \quad \Phi = \frac{1}{2}\left( p_x^2 + p_y^2 + p_z^2\right) - c = 0, \end{equation*}$

where $\Phi$ is normalized. As the constraints are nonlinear, (4.3) can be used to determine the divergence. The requisite data to compute the divergence is

$\begin{equation*} \begin{split} \mathcal{C}^*d\Phi & = m\left( p_xdx + p_ydy + p_zdy\right), \\ \mathcal{C}^*d\Phi(X_\Phi) & = m\left( p_x^2+p_y^2+p_z^2\right), \\ \Delta_\mathcal{C}\Phi & = 3m, \\ \phi & = \frac{1}{2m} - \frac{c}{m(p_x^2+p_y^2+p_z^2)}, \\ \{H,\phi\} & = \frac{2gcp_z}{(p_x^2+p_y^2+p_z^2)}. \end{split} \end{equation*}$

Although $\mathcal{C}^*d\Phi(X_\Phi)\ne 0$ everywhere, it does in some tubular neighborhood of $M$ . The divergence of the system is given by

$\begin{equation*} \begin{split} \mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M) & = -3\cdot \mathcal{C}^*d\Phi\left( [X_H,X_\phi]\right) - 3\{H,\phi\}\mathcal{M}\\ & = -\frac{18mgcp_z}{(p_x^2+p_y^2+p_z^2)^2} - \frac{18mgcp_z}{(p_x^2+p_y^2+p_z^2)^2} \\ & = -\frac{9mgp_z}{c}, \end{split} \end{equation*}$

where we used the fact that $p_x^2+p_y^2+p_z^2 = 2c$ . The divergence does not vanish so $\mu_{\mathscr{C}}$ is not preserved. However, there exists an exact 1-form which produces the divergence:

$\begin{equation*} \frac{9mg}{c}dz\left( X_H^M\right) = -\mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M). \end{equation*}$

Therefore, the following volume-form is preserved

$\begin{equation*} \exp\left( \frac{9mgz}{c} \right)\cdot \mu_{\mathscr{C}}. \end{equation*}$

7.2.2. Appel's example

The other nonlinear constraint example we will examine is Appel's example. The Lagrangian is the same as the constant kinetic energy case,

$\begin{equation*} L = \frac{1}{2}m\left( \dot{x}^2+\dot{y}^2+\dot{z}^2\right) - mgz, \end{equation*}$

while the nonlinear constraint is now

$\begin{equation*} \Psi = a^2(\dot{x}^2+\dot{y}^2) - \dot{z}^2 = 0. \end{equation*}$

The data on the Hamiltonian side becomes

$\begin{equation*} H = \frac{1}{2m}\left( p_x^2+p_y^2 + p_z^2\right) + mgz, \quad \Phi = \frac{a^2}{2}\left(p_x^2+p_y^2\right) - \frac{1}{2}p_z^2 = 0, \end{equation*}$

where, again, the constraint is normalized. As before, the constraints are nonlinear so (4.3) can be used. The requisite data is

$\begin{equation*} \begin{split} \mathcal{C}^*d\Phi & = a^2mp_xdx + a^2mp_ydy - mp_zdz, \\ \mathcal{C}^*d\Phi(X_\Phi) & = ma^4(p_x^2+p_y^2) +mp_z^2, \\ \Delta_\mathcal{C}\Phi & = m(2a^2-1),\\ \phi & = \frac{a^2(p_x^2+p_y^2)-p_z^2}{2ma^4(p_x^2+p_y^2)+2mp_z^2},\\ \{H,\phi\} & = - \frac{a^2(a^2+1)gp_z(p_x^2+p_y^2)}{(a^4p_x^2+a^4p_y^2+p_z^2)^2}. \end{split} \end{equation*}$

The constraint is not admissible as $\mathcal{C}^*d\Phi(X_\Phi)$ vanishes at $p_x = p_y = p_z = 0$ which is in the constraint manifold. As long as $a\ne 1$ , this is the only place where this degeneracy occurs. The divergence is

$\begin{equation*} \begin{split} \mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M) & = -3\cdot\mathcal{C}^*d\Phi\left( [X_H,X_\phi]\right) - 3\{H,\phi\}\mathcal{M} \\ & = \frac{12a^2(a^2+1)mgp_z(p_x^2+p_y^2)(a^6p_x^2+a^6p_y^2-p_z^2)}{(a^4p_x^2+a^4p_y^2+p_z^2)^3}. \end{split} \end{equation*}$

To simplify the divergence, notice that the constraint makes $p_z^2 = a^2(p_x^2+p_y^2)$ . Substituting this, the divergence becomes

$\begin{equation*} \mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M) = \frac{12(a^2-1)mg}{(a^2+1)p_z}. \end{equation*}$

The following exact 1-form solves the cohomology equation

$\begin{equation*} \frac{12(a^2-1)}{a^2p_z} dp_z\left(X_H^M\right) = -\mathrm{div}_{\mu_{\mathscr{C}}}(X_H^M), \end{equation*}$

$\begin{equation*} \dot{p}_z = -\frac{a^2mg}{1+a^2}. \end{equation*}$

Therefore, the following form is preserved

$\begin{equation*} p_z^K\cdot\mu_{\mathscr{C}}, \quad K = \frac{12(a^2-1)}{a^2}. \end{equation*}$

Unfortunately, this form is not a volume-form as it vanishes when $p_z = 0$ .

7.3. Non-Mechanical lagrangians

All of the examples examined so far have been for systems whose Lagrangian is natural. We conclude this work with two examples whose Lagrangian is not given by a Riemannian metric.

7.3.1. Higher-Order lagrangian

Consider the Lagrangian

$\begin{equation*} L = \frac{1}{4}\left( \dot{x}^4 + \dot{y}^4 + \dot{z}^4 \right), \end{equation*}$

subject to the nonintegrable constraint $\dot{z} = x\dot{y}$ . The Hamiltonian is

$\begin{equation*} H = \frac{3}{4}\left( p_x^{4/3} + p_y^{4/3} + p_z^{4/3} \right), \end{equation*}$

and the constraint becomes

$\begin{equation*} \Phi = p_z^{1/3} - xp_y^{1/3}. \end{equation*}$

Notice that singularities appear when the momentum vanishes. Continuing with the computations, we have

$\begin{equation*} \begin{split} \mathcal{C}^*d\Phi & = dz - xdy \\ \mathcal{C}^*d\Phi(X_\Phi) & = \frac{x^2}{3p_y^{2/3}} + \frac{1}{3p_z^{2/3}} \\ \phi & = 3p_z - \frac{3x^2p_z + 3xp_y}{x^2 + \nu^{2/3}}, \quad \nu = \frac{p_y}{p_z} \\ \Delta_\mathcal{C}\Phi & = 0 \end{split} \end{equation*}$

Applying the constraint, the divergence simplifies to

$\begin{equation*} \begin{split} \mathrm{div}_{\mu_\mathscr{C}}(X_H^M) & = -3p_x^{1/3}\cdot \frac{x^4-2}{x(1+x^4)} \\ & = -3\frac{x^4-2}{x(1+x^4)}dx(X_H^M) \end{split} \end{equation*}$

The following form is exact

$\begin{equation*} \alpha = \frac{3x^4-6}{x(1+x^4)} dx, \end{equation*}$

and an invariant form is

$\begin{equation*} \frac{(1+x^4)^{9/4}}{x^6}\mu_\mathscr{C}. \end{equation*}$

Notice the singularity at $x = 0$ .

7.3.2. Relativistic lagrangian

Suppose we have the relativistic Lagrangian

$\begin{equation*} L = -m_0c^2\sqrt{ 1 - \frac{v^2}{c^2}}, \quad v^2 = \dot{x}^2 + \dot{y}^2 + \dot{z}^2. \end{equation*}$

With the same constraints $\dot{z} = x\dot{y}$ . The Hamiltonian and constraint are

$\begin{equation*} H = c\sqrt{ p^2 + m^2c^2}, \quad \Phi = \frac{c\left( p_z - xp_y\right)}{\sqrt{p^2+m^2c^2}}. \end{equation*}$

The computation yields

$\begin{equation*} \begin{split} \mathcal{C}^*d\Phi & = dz - xdy \\ \mathcal{C}^*d\Phi(X_\Phi) & = \frac{c\left( (m^2c^2+p_x^2) (1+x^2) +(xp_z + p_y)^2\right)}{\left( m^2c^2 + p^2\right)^{3/2}} \\ \phi & = \frac{ (p^2+m^2c^2)(p_z - xp_y)}{(m^2c^2+p_x^2) (1+x^2) +(xp_z + p_y)^2} \\ \Delta_\mathcal{C}\Phi & = 0 \end{split} \end{equation*}$

Applying the constraints, the divergence simplifies to

$\begin{equation*} \begin{split} \mathrm{div}_{\mu_\mathscr{C}}(X_H^M) & = \frac{3xp_x}{(1+x^2)\sqrt{m^2c^2 + p_x^2 + (1+x^2)p_y^2}} \\ & = \frac{3x}{1+x^2}dx(X_H^M) \end{split} \end{equation*}$

Therefore, an invariant volume is given by

$\begin{equation*} (1+x^2)^{-3/2}\mu_\mathscr{C}. \end{equation*}$

Acknowledgments

We thank Dr. J.C. Marrero for pointing us to existing work in this field as well as the reviewers for helpful comments.

References

[1]	Z. Zhang, X. Jia, T. Yang, Y. Gu, W. Wang, L. Chen, Multi-objective optimization of lubricant volume in an ELSD considering thermal effects, Int. J. Therm. Sci., 164 (2021), 106884. https://doi.org/10.1016/j.ijthermalsci.2021.106884 doi: 10.1016/j.ijthermalsci.2021.106884
[2]	C. Yu, J. Liu, J. Zhang, K. Xue, S. Zhang, J. Liao, et al., Design and optimization and experimental verification of a segmented double-helix blade roller for straw returning cultivators, J. Chin. Inst. Eng., 44 (2021), 379–387. https://doi.org/10.1080/02533839.2021.1903342 doi: 10.1080/02533839.2021.1903342
[3]	Y. Cao, J. Yao, J. Li, X. Chen, J. Wu, Optimization of microbial oils production from kitchen garbage by response surface methodology, J. Renew. Sustain. Ener., 5 (2013), 053105. https://doi.org/10.1063/1.4821218 doi: 10.1063/1.4821218
[4]	K. Cao, Z. Li, Y. Gu, L. Zhang, L. Chen, The control design of transverse interconnected electronic control air suspension based on seeker optimization algorithm, P. I. Mech. Eng. D-J. Aut., 235 (2021), 2200–2211. https://doi.org/10.1177/0954407020984667 doi: 10.1177/0954407020984667
[5]	G. Wang, H. Yuan, Stability and critical spinning speed of a flexible liquid-filled rotor in thermal environment with nonlinear variable-temperature, Appl. Math. Model., 95 (2021), 143–158. https://doi.org/10.1016/j.apm.2021.01.056 doi: 10.1016/j.apm.2021.01.056
[6]	G. Wang, H. Yuan, Dynamics and stability analysis of an axially functionally graded hollow rotor partially filled with liquid, Compos. Struct., 266 (2021), 113821. https://doi.org/10.1016/j.compstruct.2021.113821 doi: 10.1016/j.compstruct.2021.113821
[7]	G. Wang, W. Yang, H. Yuan, Dynamics and stability analysis of a flexible liquid-filled rotor in a constant thermal environment, J. Vib. Control, 28 (2021), 2913–2924. https://doi.org/10.1177/10775463211022489 doi: 10.1177/10775463211022489
[8]	D. Zhang, N. Zhang, N. Ye, J. Fang, X. Han, Hybrid learning algorithm of radial basis function networks for reliability analysis, IEEE Trans. Reliab., 70 (2021), 887–900. https://doi.org/10.1109/TR.2020.3001232 doi: 10.1109/TR.2020.3001232
[9]	X. Du, H. Xu, F. Zhu, A data mining method for structure design with uncertainty in design variables, Comput. Struct., 244 (2021), 106457. https://doi.org/10.1016/j.compstruc.2020.106457 doi: 10.1016/j.compstruc.2020.106457
[10]	E. Acar, G. Bayrak, Y. Jung, I. Lee, P. Ramu, S. S. Ravichandran, Modeling, analysis and optimization under uncertainties: A review, Struct. Multidiscip. Optimiz., 64 (2021), 2909–2945. https://doi.org/10.1007/s00158-021-03026-7 doi: 10.1007/s00158-021-03026-7
[11]	M. Zimmermann, S. Königs, C. Niemeyer, J. Fender, C. Zeherbauer, R. Vitale, et al., On the design of large systems subject to uncertainty, J. Eng. Design, 28 (2017), 233–254. https://doi.org/10.1080/09544828.2017.1303664 doi: 10.1080/09544828.2017.1303664
[12]	Y. Noh, K. K. Choi, I. Lee, Reduction of ordering effect in reliability-based design optimizatioin using dimension reduction method, AIAA J. 47 (2009), 994–1004. https://doi.org/10.2514/1.40224
[13]	U. Lee, N. Kang, I. Lee, Selection of optimal target reliability in RBDO through reliability-based design for market systems (RBDMS) and application to electric vehicle design, Struct. Multidiscip. Optimiz., 60 (2019), 949–963. https://doi.org/10.1007/s00158-019-02245-3 doi: 10.1007/s00158-019-02245-3
[14]	K. Wang, P. Wang, X. Chen, L. T. Zhao, Multiobjective optimization design of toll plaza, Math. Probl. Eng., 2020 (2020), 2324894. https://doi.org/10.1155/2020/2324894 doi: 10.1155/2020/2324894
[15]	C. Yu, D. Zhu, Y. Gao, K. Xue, S. Zhang, J. Liao, et al., Optimization and experiment of counter-rotating straw returning cultivator based on discrete element method, J. Adv. Mech. Des. Syst., 14 (2020). https://doi.org/10.1299/jamdsm.2020jamdsm0097
[16]	S. Liu, J. Sun, H. Zhou, F. Wei, M. Lu, M. Lei, Experimental and numerical study on fatigue performance for TIG welding and EB welding of RAFM steel plate, Fusion Eng. Des., 146 (2019), 2663–2666. https://doi.org/10.1016/j.fusengdes.2019.04.076 doi: 10.1016/j.fusengdes.2019.04.076
[17]	J. Mendoza, E. Bismut, D. Straub, J. Köhler, Optimal life-cycle mitigation of fatigue failure risk for structural systems, Reliab. Eng. Syst. Safe., 222 (2022), 108390. https://doi.org/10.1016/j.ress.2022.108390 doi: 10.1016/j.ress.2022.108390
[18]	C. A. Castiglioni, R. Pucinotti, Failure criteria and cumulative damage models for steel components under cyclic loading, J. Constr. Steel Res., 65 (2009), 751–765. https://doi.org/10.1016/j.jcsr.2008.12.007 doi: 10.1016/j.jcsr.2008.12.007
[19]	D. M. Harland, R. D. Lorenz, Space systems failures: Disasters and rescues of satellites, rockets and space probes, in Springer Praxis Books, Springer Praxis, 2005. https://doi.org/10.1007/978-0-387-27961-9
[20]	A. D. Kiureghian, O. Ditlevsen, Aleatory or epistemic? Does it matter?, Struct. Safe., 31 (2009), 105–112. https://doi.org/10.1016/j.strusafe.2008.06.020 doi: 10.1016/j.strusafe.2008.06.020
[21]	M. A. Hariri-Ardebili, F. Pourkamali-Anaraki, Structural uncertainty quantification with partial information, Expert Syst. Appl., 198 (2022), 116736. https://doi.org/10.1016/j.eswa.2022.116736 doi: 10.1016/j.eswa.2022.116736
[22]	K. Bowcutt. A perspective on the future of aerospace vehicle design, in 12th AIAA International Space Planes and Hypersonic Systems and Technologies: American Institute of Aeronautics and Astronautics, 2003.
[23]	J. Fang, Y. Gao, G. Sun, Y. Zhang, Q. Li, Crashworthiness design of foam-filled bitubal structures with uncertainty, Int. J. Non-Lin. Mech., 67 (2014), 120–132. https://doi.org/10.1016/j.ijnonlinmec.2014.08.005 doi: 10.1016/j.ijnonlinmec.2014.08.005
[24]	S. A. Latifi Rostami, A. Kolahdooz, J. Zhang, Robust topology optimization under material and loading uncertainties using an evolutionary structural extended finite element method, Eng. Anal. Bound. Elem., 133 (2021), 61–70. https://doi.org/10.1016/j.enganabound.2021.08.023 doi: 10.1016/j.enganabound.2021.08.023
[25]	X. Wang, Z. Meng, B. Yang, C. Cheng, K. Long, J. Li, Reliability-based design optimization of material orientation and structural topology of fiber-reinforced composite structures under load uncertainty, Compos. Struct., 2022 (2022), 115537. https://doi.org/10.1016/j.compstruct.2022.115537 doi: 10.1016/j.compstruct.2022.115537
[26]	M. E. Riley, R. V. Grandhi, Quantification of model-form and predictive uncertainty for multi-physics simulation, Comput. Struct., 89 (2011), 1206–1213. https://doi.org/10.1016/j.compstruc.2010.10.004 doi: 10.1016/j.compstruc.2010.10.004
[27]	B. Do, M. Ohsaki, M. Yamakawa, Bayesian optimization for robust design of steel frames with joint and individual probabilistic constraints, Eng. Struct., 245 (2021), 112859. https://doi.org/10.1016/j.engstruct.2021.112859 doi: 10.1016/j.engstruct.2021.112859
[28]	Z. Meng, Y. Pang, Y. Pu, X. Wang, New hybrid reliability-based topology optimization method combining fuzzy and probabilistic models for handling epistemic and aleatory uncertainties, Comput. Method. Appl. Mech., 363 (2020), 112886. https://doi.org/10.1016/j.cma.2020.112886 doi: 10.1016/j.cma.2020.112886
[29]	C. Yang, H. Ouyang, A novel load-dependent sensor placement method for model updating based on time-dependent reliability optimization considering multi-source uncertainties, Mech. Syst. Signal Proces., 165 (2022), 108386. https://doi.org/10.1016/j.ymssp.2021.108386 doi: 10.1016/j.ymssp.2021.108386
[30]	G. A. da Silva, E. L. Cardoso, A. T. Beck, Comparison of robust, reliability-based and non-probabilistic topology optimization under uncertain loads and stress constraints, Probabilist. Eng. Mech., 59 (2020), 103039. https://doi.org/10.1016/j.probengmech.2020.103039 doi: 10.1016/j.probengmech.2020.103039
[31]	J. M. King, R. V. Grandhi, T. W. Benanzer, Quantification of epistemic uncertainty in re-usable launch vehicle aero-elastic design, Eng. Optimiz., 44 (2012), 489–504. https://doi.org/10.1080/0305215X.2011.588224 doi: 10.1080/0305215X.2011.588224
[32]	Y. C. Tsao, V. V. Thanh, A multi-objective fuzzy robust optimization approach for designing sustainable and reliable power systems under uncertainty, Appl. Soft Comput., 92 (2020), 106317. https://doi.org/10.1016/j.asoc.2020.106317 doi: 10.1016/j.asoc.2020.106317
[33]	L. Wang, B. Ni, X. Wang, Z. Li, Reliability-based topology optimization for heterogeneous composite structures under interval and convex mixed uncertainties, Appl. Math. Model., 99 (2021), 628–652. https://doi.org/10.1016/j.apm.2021.06.014 doi: 10.1016/j.apm.2021.06.014
[34]	L. A. Zadeh, Fuzzy sets as a basis for a theory of possibility, Fuzzy Set. Syst., 1 (1978), 3–28. https://doi.org/10.1016/0165-0114(78)90029-5 doi: 10.1016/0165-0114(78)90029-5
[35]	H. Lü, K. Yang, X. Huang, H. Yin, W. B. Shangguan, D. Yu, An efficient approach for the design optimization of dual uncertain structures involving fuzzy random variables, Comput. Method. Appl. Mech., 371 (2020), 113331. https://doi.org/10.1016/j.cma.2020.113331 doi: 10.1016/j.cma.2020.113331
[36]	R. E. Moore, Methods and Application of Interval Analysis, SIAM: Philadelphia, USA, 1979. https://doi.org/10.1137/1.9781611970906
[37]	S. P. Gurav, J. F. L. Goosen, F. vanKeulen, Bounded-But-Unknown uncertainty optimization using design sensitivities and parallel computing: Application to MEMS, Comput. Struct., 83 (2005), 1134–1149. https://doi.org/10.1016/j.compstruc.2004.11.021 doi: 10.1016/j.compstruc.2004.11.021
[38]	R. Yuan, D. Ma, H. Zhang, Flow characteristics and grain size distribution of granular gangue mineral by compaction treatment, Adv. Mater. Sci. Eng., 2017 (2017), 2509286. https://doi.org/10.1155/2017/2509286 doi: 10.1155/2017/2509286
[39]	G. Sun, T. Pang, J. Fang, G. Li, Q. Li, Parameterization of criss-cross configurations for multiobjective crashworthiness optimization, Int. J. Mech. Sci., 124–125 (2017), 145–157. https://doi.org/10.1016/j.ijmecsci.2017.02.027
[40]	J. Fang, G. Sun, N. Qiu, T. Pang, S. Li, Q. Li, On hierarchical honeycombs under out-of-plane crushing, Int. J. Solids Struct., 135 (2018), 1–13. https://doi.org/10.1016/j.ijsolstr.2017.08.013. doi: 10.1016/j.ijsolstr.2017.08.013
[41]	J. Fu, Q. Liu, K. Liufu, Y. Deng, J. Fang, Q. Li, Design of bionic-bamboo thin-walled structures for energy absorption, Thin-Wall. Struct., 135 (2019), 400–413. https://doi.org/10.1016/j.tws.2018.10.003 doi: 10.1016/j.tws.2018.10.003
[42]	J. Fang, Y. Gao, G. Sun, G. Zheng, Q. Li, Dynamic crashing behavior of new extrudable multi-cell tubes with a functionally graded thickness, Int. J. Mech. Sci., 103 (2015), 63–73. https://doi.org/10.1016/j.ijmecsci.2015.08.029 doi: 10.1016/j.ijmecsci.2015.08.029
[43]	J. Fang, Y. Gao, G. Sun, N. Qiu, Q. Li, On design of multi-cell tubes under axial and oblique impact loads, Thin-Wall. Struct., 95 (2015), 115–126. https://doi.org/10.1016/j.tws.2015.07.002 doi: 10.1016/j.tws.2015.07.002
[44]	J. Fang, G. Sun, N. Qiu, N. H. Kim, Q. Li, On design optimization for structural crashworthiness and its state of the art, Struct. Multidiscip. Optimiz., 55 (2017), 1091–1119. https://doi.org/10.1007/s00158-016-1579-y doi: 10.1007/s00158-016-1579-y
[45]	N. Qiu, Y. Gao, J. Fang, Z. Feng, G. Sun, Q. Li, Crashworthiness analysis and design of multi-cell hexagonal columns under multiple loading cases, Finite Elem. Anal. Des., 104 (2015), 89–101. https://doi.org/10.1016/j.finel.2015.06.004 doi: 10.1016/j.finel.2015.06.004
[46]	S. Kodiyalam, High performance computing for multidisciplinary design optimization and robustness of vehicle structures, in Computational Fluid and Solid Mechanics 2003, Elsevier Science Ltd, (2003), 2305–2307. https://doi.org/10.1016/B978-008044046-0.50566-2
[47]	K. Hamza, K. Saitou, Design Optimization of Vehicle Structures for Crashworthiness Using Equivalent Mechanism Approximations, J. Mech. Des., 127 (2004), 485–492. https://doi.org/10.1115/1.1862680 doi: 10.1115/1.1862680
[48]	X. Gu, G. Sun, G. Li, X. Huang, Y. Li, Q. Li, Multiobjective optimization design for vehicle occupant restraint system under frontal impact, Struct. Multidiscip. Optimiz., 47 (2013), 465–477. https://doi.org/10.1007/s00158-012-0811-7 doi: 10.1007/s00158-012-0811-7
[49]	L. Chen, W. Li, Y. Yang, W. Miao, Evaluation and optimization of vehicle pedal comfort based on biomechanics, Proc. Inst. Mech. Eng. Part D J. Automob. Eng., 234 (2019), 1402–1412. https://doi.org/10.1177/0954407019878355 doi: 10.1177/0954407019878355
[50]	X. L. Zhang, T. Wu, Y. Shao, J. Song, Structure optimization of wheel force transducer based on natural frequency and comprehensive sensitivity, Chin. J. Mech. Eng., 30 (2017), 973–981. https://doi.org/10.1007/s10033-017-0149-6 doi: 10.1007/s10033-017-0149-6
[51]	X. Xu, X. Chen, Z. Liu, Y. Zhang, Y. Xu, J. Fang, et al., A feasible identification method of uncertainty responses for vehicle structures, Struct. Multidiscip. Optimiz., 64 (2021), 3861–3876. https://doi.org/10.1007/s00158-021-03065-0 doi: 10.1007/s00158-021-03065-0
[52]	X. Liu, X. Liu, Z. Zhou, L. Hu, An efficient multi-objective optimization method based on the adaptive approximation model of the radial basis function, Struct. Multidiscip. Optimiz., 63 (2021), 1385–1403. https://doi.org/10.1007/s00158-020-02766-2 doi: 10.1007/s00158-020-02766-2
[53]	J. Fang, Y. Gao, G. Sun, C. Xu, Q. Li, Multiobjective robust design optimization of fatigue life for a truck cab, Reliab. Eng. Syst. Safe., 135 (2015), 1–8. https://doi.org/10.1016/j.ress.2014.10.007 doi: 10.1016/j.ress.2014.10.007
[54]	X. Gu, G. Sun, G. Li, L. Mao, Q. Li, A Comparative study on multiobjective reliable and robust optimization for crashworthiness design of vehicle structure, Struct. Multidiscip. Optimiz., 48 (2013), 669–684. https://doi.org/10.1007/s00158-013-0921-x doi: 10.1007/s00158-013-0921-x
[55]	J. Zhou, F. Lan, J. Chen, F. Lai, Uncertainty optimization design of a vehicle body structure considering random deviations, Automot. Innov., 1 (2018), 342–351. https://doi.org/10.1007/s42154-018-0041-9 doi: 10.1007/s42154-018-0041-9
[56]	J. Zhu, X. Wang, H. Zhang, Y. Li, R. Wang, Z. Qiu, Six sigma robust design optimization for thermal protection system of hypersonic vehicles based on successive response surface method, Chinese J. Aeronaut., 32 (2019), 2095–2108. https://doi.org/10.1016/j.cja.2019.04.009 doi: 10.1016/j.cja.2019.04.009
[57]	X. Wang, L. Shi, A new metamodel method using Gaussian process based bias function for vehicle crashworthiness design, Int. J. Crashworthines., 19 (2014), 311–321. https://doi.org/10.1080/13588265.2014.898932 doi: 10.1080/13588265.2014.898932
[58]	N. Qiu, J. Zhang, F. Yuan, Z. Jin, Y. Zhang, J. Fang, Mechanical performance of triply periodic minimal surface structures with a novel hybrid gradient fabricated by selective laser melting, Eng. Struct., 263 (2022), 114377. https://doi.org/10.1016/j.engstruct.2022.114377 doi: 10.1016/j.engstruct.2022.114377
[59]	L. Chen, P. Ma, J. Tian, X. Liang, Prediction and optimization of lubrication performance for a transfer case based on computational fluid dynamics, Eng. Appl. Comp. Fluid, 13 (2019), 1013–1023. https://doi.org/10.1080/19942060.2019.1663765 doi: 10.1080/19942060.2019.1663765
[60]	X. Xu, Y. Zhang, X. Wang, J. Fang, J. Chen, J. Li, Searching superior crashworthiness performance by constructing variable thickness honeycombs with biomimetic cells, Int. J. Mech. Sci., 235 (2022), 107718. https://doi.org/10.1016/j.ijmecsci.2022.107718 doi: 10.1016/j.ijmecsci.2022.107718
[61]	X. Xu, Y. Zhang, J. Wang, F. Jiang, C. H. Wang, Crashworthiness design of novel hierarchical hexagonal columns, Compos. Struct., 194 (2018), 36–48. https://doi.org/10.1016/j.compstruct.2018.03.099 doi: 10.1016/j.compstruct.2018.03.099
[62]	X. Song, L. Lai, S. Xiao, Y. Tang, M. Song, J. Zhang, et al., Bionic design and multi-objective optimization of thin-walled structures inspired by conchs, Electron. Res. Arch., 31 (2023), 575–598. https://doi.org/10.3934/era.2023028 doi: 10.3934/era.2023028
[63]	J. Fang, Y. Gao, G. Sun, Q. Li, Multiobjective reliability-based optimization for design of a vehicledoor, Finite Elem. Anal. Des., 67 (2013), 13–21. https://doi.org/10.1016/j.finel.2012.11.007 doi: 10.1016/j.finel.2012.11.007
[64]	N. Qiu, Z. Jin, J. Liu, L. Fu, Z. Chen, N. H. Kim, Hybrid multi-objective robust design optimization of a truck cab considering fatigue life, Thin-Wall. Struct., 162 (2021), 107545. https://doi.org/10.1016/j.tws.2021.107545 doi: 10.1016/j.tws.2021.107545
[65]	G. Sun, H. Zhang, J. Fang, G. Li, Q. Li, A new multi-objective discrete robust optimization algorithm for engineering design, Appl. Math. Model., 53 (2018), 602–621. https://doi.org/10.1016/j.apm.2017.08.016 doi: 10.1016/j.apm.2017.08.016
[66]	F. Lei, X. Lv, J. Fang, T. Pang, Q. Li, G. Sun, Injury biomechanics-based nondeterministic optimization of front-end structures for safety in pedestrian-vehicle impact, Thin-Wall. Struct., 167 (2021). https://doi.org/10.1016/j.tws.2021.108087
[67]	H. Lu, D. Yu, Brake squeal reduction of vehicle disc brake system with interval parameters by uncertain optimization, J. Sound Vib., 333 (2014), 7313–7325. https://doi.org/10.1016/j.jsv.2014.08.027 doi: 10.1016/j.jsv.2014.08.027
[68]	J. Wu, Z. Luo, Y. Zhang, N. Zhang, An interval uncertain optimization method for vehicle suspensions using Chebyshev metamodels, Appl. Math. Model., 38 (2014), 3706–3723. https://doi.org/10.1016/j.apm.2014.02.012 doi: 10.1016/j.apm.2014.02.012
[69]	A. Jamali, M. Salehpour, N. Nariman-zadeh, Robust Pareto active suspension design for vehicle vibration model with probabilistic uncertain parameters, Multibody Syst. Dyn., 30 (2013), 265–285. https://doi.org/10.1007/s11044-012-9337-4 doi: 10.1007/s11044-012-9337-4
[70]	X. Gu, W. Wang, L. Xia, P. Jiang, A system optimisation design approach to vehicle structure under frontal impact based on SVR of optimised hybrid kernel function, Int. J. Crashworthines., 26 (2021), 1–15. https://doi.org/10.1080/13588265.2019.1634335 doi: 10.1080/13588265.2019.1634335
[71]	Z. Liu, C. Jiang, Y. Li, Y. Bai, Fatigue life optimization for spot-welded Structures of vehicle body considering uncertainty of welding spots, China Mech. Eng., 26 (2015), 2544–2549. https://doi.org/10.3969/j.issn.1004-132X.2015.18.024 doi: 10.3969/j.issn.1004-132X.2015.18.024
[72]	L. Farkas, D. Moens, S. Donders, D. Vandepitte, Optimisation study of a vehicle bumper subsystem with fuzzy parameters, Mech. Syst. Signal Pr., 32 (2012), 59–68. https://doi.org/10.1016/j.ymssp.2011.11.014 doi: 10.1016/j.ymssp.2011.11.014
[73]	E. Acar, K. Solanki, System reliability based vehicle design for crashworthiness and effects of various uncertainty reduction measures, Struct. Multidiscip. Optimiz., 39 (2009), 311–325. https://doi.org/10.1007/s00158-008-0327-3 doi: 10.1007/s00158-008-0327-3
[74]	Q. Zhao, H. Zhang, Z. Zhu, R. Jiang, L. Yuan, Reliability-based topology optimization for vehicle suspension control arm, Aut. Eng., 40(2018), 313–319. https://doi.org/10.19562/j.chinasae.qcgc.2018.03.011 doi: 10.19562/j.chinasae.qcgc.2018.03.011
[75]	M. Grujicic, G. Arakere, W. C. Bell, H. Marvi, H. V. Yalavarthy, B. Pandurangan, et al., Reliability-based design optimization for durability of ground vehicle suspension system components, J. Mater. Eng. Perform., 19 (2010), 301–313. https://doi.org/10.1007/s11665-009-9482-y doi: 10.1007/s11665-009-9482-y
[76]	M. Rais-Rohani, K. N. Solanki, E. Acar, C. D. Eamon, Shape and sizing optimisation of automotive structures with deterministic and probabilistic design constraints, Int. J. Vehicle Des., 54 (2010), 309–338. https://doi.org/10.1504/IJVD.2010.036839 doi: 10.1504/IJVD.2010.036839
[77]	X. Xu, J. Chen, Z. Lin, Y. Qiao, X. Chen, Y. Zhang, et al., Optimization design for the planetary gear train of an electric vehicle under uncertainties, Actuators, 11 (2022). https://doi.org/10.3390/act11020049
[78]	G. Sun, H. Zhang, R. Wang, X. Lv, Q. Li, Multiobjective reliability-based optimization for crashworthy structures coupled with metal forming process, Struct. Multidiscip. Optimiz., 56 (2017), 1571–1587. https://doi.org/10.1007/s00158-017-1825-y doi: 10.1007/s00158-017-1825-y
[79]	Y. Xu, Y. Gao, C. Wu, J. Fang, Q. Li, Robust topology optimization for multiple fiber-reinforced plastic (FRP) composites under loading uncertainties, Struct. Multidiscip. Optimiz., 59 (2019), 695–711. https://doi.org/10.1007/s00158-018-2175-0 doi: 10.1007/s00158-018-2175-0
[80]	C. van Mierlo, L. Burmberger, M. Daub, F. Duddeck, M. G. R. Faes, D. Moens, Interval methods for lack-of-knowledge uncertainty in crash analysis, Mech. Syst. Signal Proces., 168 (2022). https://doi.org/10.1016/j.ymssp.2021.108574
[81]	C. Lin, F. Gao, Y. Bai, Multiobjective reliability-based design optimisation for front structure of an electric vehicle using hybrid metamodel accuracy improvement strategy-based probabilistic sufficiency factor method, Int. J. Crashworthines., 23 (2018), 290–301. https://doi.org/10.1080/13588265.2017.1317466 doi: 10.1080/13588265.2017.1317466
[82]	F. Y. Li, G. Y. Li, Interval-based uncertain multi-objective optimization design of vehicle crashworthiness, CMC-Comput. Mater. Con., 15 (2010), 199–219. https://doi.org/10.2478/s11533-009-0061-0 doi: 10.2478/s11533-009-0061-0
[83]	J. Li, Y. Fang, Z. Zhan, Y. Jiang, An enhanced surrogate model based vehicle robust design method under materials and manufacturing uncertainties, in ASME International Mechanical Engineering Congress and Exposition, 2016. https://doi.org/10.1115/IMECE2016-67714
[84]	R. J. Yang, L. Gu, C. H. Tho, K. K. Choi, B. Youn, Reliability-based multidisciplinary design optimization of vehicle structures. in 43rd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, 2002. https://doi.org/10.2514/6.2002-1758
[85]	L. Cao, J. Liu, L. Xie, C. Jiang, R. Bi, Non-probabilistic polygonal convex set model for structural uncertainty quantification, Appl. Math. Model., 89 (2021), 504–518. https://doi.org/10.1016/j.apm.2020.07.025 doi: 10.1016/j.apm.2020.07.025
[86]	X. Meng, J. Liu, L. Cao, Z. Yu, D. Yang, A general frame for uncertainty propagation under multimodally distributed random variables, Comput. Method. Appl. Mech. Eng., 367 (2020). https://doi.org/10.1016/j.cma.2020.113109
[87]	G. Chen, D. Yang, A unified analysis framework of static and dynamic structural reliabilities based on direct probability integral method, Mech. Syst. Signal Proces., 158 (2021). https://doi.org/10.1016/j.ymssp.2021.107783
[88]	X. Y. Long, D. L. Mao, C. Jiang, F. Y. Wei, G. J. Li, Unified uncertainty analysis under probabilistic, evidence, fuzzy and interval uncertainties, Comput. Method. Appl. Mech. Eng., 355 (2019), 1–26. https://doi.org/10.1016/j.cma.2019.05.041
[89]	X. Xu, X. Chen, Z. Liu, J. Yang, Y. Xu, Y. Zhang, et al., Multi-objective reliability-based design optimization for the reducer housing of electric vehicles, Eng. Optimiz., 54 (2022), 1324–1340. https://doi.org/10.1080/0305215X.2021.1923704 doi: 10.1080/0305215X.2021.1923704
[90]	N. Qiu, C. Park, Y. Gao, J. Fang, G. Sun, N. H. Kim, Sensitivity-based parameter calibration and model validation under model error, J. Mech. Des., 140 (2017). https://doi.org/10.1115/1.4038298
[91]	Z. L. Huang, C. Jiang, Y. S. Zhou, Z. Luo, Z. Zhang, An incremental shifting vector approach for reliability-based design optimization, Struct. Multidiscip. Optimiz., 53 (2016), 523–543. https://doi.org/10.1007/s00158-015-1352-7 doi: 10.1007/s00158-015-1352-7
[92]	W. Gao, D. Wu, K. Gao, X. Chen, F. Tin-Loi, Structural reliability analysis with imprecise random and interval fields, Appl. Math. Model., 55 (2018), 49–67. https://doi.org/10.1016/j.apm.2017.10.029 doi: 10.1016/j.apm.2017.10.029
[93]	N. Qiu, Y. Gao, J. Fang, G. Sun, Q. Li, N. H. Kim, Crashworthiness optimization with uncertainty from surrogate model and numerical error, Thin-Wall. Struct., 129 (2018), 457–472. https://doi.org/10.1016/j.tws.2018.05.002 doi: 10.1016/j.tws.2018.05.002
[94]	G. S. Song, M. C. Kim, Theoretical approach for uncertainty quantification in probabilistic safety assessment using sum of lognormal random variables, Nucl. Eng. Technol., 54 (2022), 2084–2093. https://doi.org/10.1016/j.net.2021.12.033 doi: 10.1016/j.net.2021.12.033
[95]	N. A. W. Walton, R. Crowder, N. Satvat, N. R. Brown, V. Sobes, Demonstration of a random sampling approach to uncertainty propagation for generic pebble-bed fluoride-salt-cooled high temperature reactor (gFHR), Nucl. Eng. Des., 395 (2022), 111843. https://doi.org/10.1016/j.nucengdes.2022.111843 doi: 10.1016/j.nucengdes.2022.111843
[96]	X. P. Du, W. Chen, Sequential optimization and reliability assessment method for efficient probabilistic design, J. Mech. Des., 126 (2004), 225–233. https://doi.org/10.1115/1.1649968 doi: 10.1115/1.1649968
[97]	Y. Lei, N. Yang, D. Xia, Probabilistic structural damage detection approaches based on structural dynamic response moments, Smart Struct. Syst., 20 (2017), 207–217. https://doi.org/10.12989/sss.2017.20.2.207 doi: 10.12989/sss.2017.20.2.207
[98]	P. Hoang-Anh, T. Viet-Hung, V. Tien-Chuong, Fuzzy finite element analysis for free vibration response of functionally graded semi-rigid frame structures, Appl. Math. Model., 88 (2020), 852–869. https://doi.org/10.1016/j.apm.2020.07.014 doi: 10.1016/j.apm.2020.07.014
[99]	Z. Chen, Z. Lu, C. Ling, K. Feng, Reliability analysis model of time-dependent multi-mode system under fuzzy uncertainty: Applied to undercarriage structures, Aerosp. Sci. Technol., 120 (2022), 107278. https://doi.org/10.1016/j.ast.2021.107278 doi: 10.1016/j.ast.2021.107278
[100]	P. Wang, H. Zhu, H. Tian, G. Cai, Analytic target cascading with fuzzy uncertainties based on global sensitivity analysis for overall design of launch vehicle powered by hybrid rocket motor, Aerosp. Sci. Technol., 114 (2021), 106680. https://doi.org/10.1016/j.ast.2021.106680 doi: 10.1016/j.ast.2021.106680
[101]	S. Yin, D. Yu, H. Yin, B. Xia, A new evidence-theory-based method for response analysis of acoustic system with epistemic uncertainty by using Jacobi expansion, Comput. Method. Appl. Mech. Eng., 322 (2017), 419–440. https://doi.org/10.1016/j.cma.2017.04.020 doi: 10.1016/j.cma.2017.04.020
[102]	H. Tang, D. Li, J. Li, S. Xue, Epistemic uncertainty quantification in metal fatigue crack growth analysis using evidence theory, Int. J. Fatigue, 99 (2017), 163–174. https://doi.org/10.1016/j.ijfatigue.2017.03.004 doi: 10.1016/j.ijfatigue.2017.03.004
[103]	Z. Zhang, X. X. Ruan, M. F. Duan, C. Jiang, An efficient epistemic uncertainty analysis method using evidence theory, Comput. Method. Appl. Mech. Eng., 339 (2018), 443–466. https://doi.org/10.1016/j.cma.2018.04.033 doi: 10.1016/j.cma.2018.04.033
[104]	X. Tang, K. Yuan, N. Gu, P. Li, R. Peng, An interval quantification-based optimization approach for wind turbine airfoil under uncertainties, Energy, 244 (2022), 122623. https://doi.org/10.1016/j.energy.2021.122623 doi: 10.1016/j.energy.2021.122623
[105]	Z. Niu, H. Zhu, X. Huang, A. Che, S. Fu, S. Meng, et al., Uncertainty quantification method for elastic wave tomography of concrete structure using interval analysis, Measurement, 205 (2022), 112160. https://doi.org/10.1016/j.measurement.2022.112160 doi: 10.1016/j.measurement.2022.112160
[106]	B. Y. Ni, C. Jiang, P. G. Wu, Z. H. Wang, W. Y. Tian, A sequential simulation strategy for response bounds analysis of structures with interval uncertainties, Comput. Struct., 266 (2022), 106785. https://doi.org/10.1016/j.compstruc.2022.106785 doi: 10.1016/j.compstruc.2022.106785
[107]	T. Tang, H. Luo, Y. Song, H. Fang, J. Zhang, Chebyshev inclusion function based interval kinetostatic modeling and parameter sensitivity analysis for Exechon-like parallel kinematic machines with parameter uncertainties, Mech. Mach. Theory, 157 (2021). https://doi.org/10.1016/j.mechmachtheory.2020.104209
[108]	C. Viegas, D. Daney, M. Tavakoli, A. T. de Almeida, Performance analysis and design of parallel kinematic machines using interval analysis, Mech. Mach. Theory, 115 (2017), 218–236. https://doi.org/10.1016/j.mechmachtheory.2017.05.003 doi: 10.1016/j.mechmachtheory.2017.05.003
[109]	M. Ma, L. Wang, Reliability-based topology optimization framework of two-dimensional phononic crystal band-gap structures based on interval series expansion and mapping conversion method, Int. J. Mech. Sci., 196 (2021). https://doi.org/10.1016/j.ijmecsci.2020.106265
[110]	Z. Qiu, X. Li, A new model for the eigenvalue buckling analysis with unknown-but-bounded parameters, Aerosp. Sci. Technol., 113 (2021). https://doi.org/10.1016/j.ast.2021.106634
[111]	B. Xia, H. Lu, D. Yu, C. Jiang, Reliability-based design optimization of structural systems under hybrid probabilistic and interval model, Comput. Struct., 160 (2015), 126–134. https://doi.org/10.1016/j.compstruc.2015.08.009 doi: 10.1016/j.compstruc.2015.08.009
[112]	L. Wang, Z. Chen, G. Yang, Q. Sun, J. Ge, An interval uncertain optimization method using back-propagation neural network differentiation, Comput. Method. Appl. Mech. Eng., 366 (2020). https://doi.org/10.1016/j.cma.2020.113065
[113]	S. H. Chen, L. Ma, G. W. Meng, R. Guo, An efficient method for evaluating the natural frequencies of structures with uncertain-but-bounded parameters, Comput. Struct., 87 (2009), 582–590. https://doi.org/10.1016/j.compstruc.2009.02.009 doi: 10.1016/j.compstruc.2009.02.009
[114]	N. Impollonia, G. Muscolino, Interval analysis of structures with uncertain-but-bounded axial stiffness, Comput. Method. Appl. Mech. Eng., 200 (2011), 1945–1962. https://doi.org/10.1016/j.cma.2010.07.019 doi: 10.1016/j.cma.2010.07.019
[115]	G. Zhao, G. Wen, J. Liu, A novel analysis method for vibration systems under time-varying uncertainties based on interval process model, Probabilistic Eng. Mech., 70 (2022), 103363. https://doi.org/10.1016/j.probengmech.2022.103363 doi: 10.1016/j.probengmech.2022.103363
[116]	L. Wang, J. Liu, C. Yang, D. Wu, A novel interval dynamic reliability computation approach for the risk evaluation of vibration active control systems based on PID controllers, Appl. Math. Model., 92 (2021), 422–446. https://doi.org/10.1016/j.apm.2020.11.007 doi: 10.1016/j.apm.2020.11.007
[117]	W. Gao, D. Wu, C. Song, F. Tin-Loi, X. Li, Hybrid probabilistic interval analysis of bar structures with uncertainty using a mixed perturbation Monte-Carlo method, Finite Elem. Anal. Des., 47 (2011), 643–652. https://doi.org/10.1016/j.finel.2011.01.007 doi: 10.1016/j.finel.2011.01.007
[118]	Z. P. Qiu, X. J. Wang, Solution theorems for the standard eigenvalue problem of structures with uncertain-but-bounded parameters, J. Sound Vib., 282 (2005), 381–399. https://doi.org/10.1016/j.jsv.2004.02.024 doi: 10.1016/j.jsv.2004.02.024
[119]	F. Li, Z. Luo, J. Rong, N. Zhang, Interval multi-objective optimisation of structures using adaptive Kriging approximations, Comput. Struct., 119 (2013), 68–84. https://doi.org/10.1016/j.compstruc.2012.12.028 doi: 10.1016/j.compstruc.2012.12.028
[120]	J. A. Fernandez-Prieto, J. Canada-Bago, M. A. Gadeo-Martos, J. R. Velasco, Optimisation of control parameters for genetic algorithms to test computer networks under realistic traffic loads, Appl. Soft Comput., 12 (2012), 1875–1883. https://doi.org/10.1016/j.asoc.2012.04.018 doi: 10.1016/j.asoc.2012.04.018
[121]	Y. G. Xu, G. R. Li, Z. P. Wu, A novel hybrid genetic algorithm using local optimizer based on heuristic pattern move, Appl. Artif. Intell., 15 (2001), 601–631. https://doi.org/10.1080/088395101750363966 doi: 10.1080/088395101750363966
[122]	C. Jiang, X. Han, G. P. Liu, A sequential nonlinear interval number programming method for uncertain structures, Comput. Meth. Appl. Mech. Eng., 197 (2008), 4250–4265. https://doi.org/10.1016/j.cma.2008.04.027 doi: 10.1016/j.cma.2008.04.027
[123]	J. Cheng, M. Y. Tang, Z. Y. Liu, J. R. Tan, Direct reliability-based design optimization of uncertain structures with interval parameters, J. Zhejiang Univ-Sci. A, 17 (2016), 841–854. https://doi.org/10.1631/jzus.A1600143 doi: 10.1631/jzus.A1600143
[124]	Y. P. Li, G. H. Huang, P. Guo, Z. F. Yang, S. L. Nie, A dual-interval vertex analysis method and its application to environmental decision making under uncertainty, Eur. J. Oper. Res., 200 (2010), 536–550. https://doi.org/10.1016/j.ejor.2009.01.013 doi: 10.1016/j.ejor.2009.01.013
[125]	C. Wang, H. G. Matthies, Non-probabilistic interval process model and method for uncertainty analysis of transient heat transfer problem, Int. J. Therm. Sci., 144 (2019), 147–157. https://doi.org/10.1016/j.ijthermalsci.2019.06.002 doi: 10.1016/j.ijthermalsci.2019.06.002
[126]	S. Nayak, S. Chakraverty, Non-probabilistic approach to investigate uncertain conjugate heat transfer in an imprecisely defined plate, Int. J. Heat Mass Tran., 67 (2013), 445–454. https://doi.org/10.1016/j.ijheatmasstransfer.2013.08.036 doi: 10.1016/j.ijheatmasstransfer.2013.08.036
[127]	Z. Qiu, X. Wang, J. Chen, Exact bounds for the static response set of structures with uncertain-but-bounded parameters, Int. J. Solids Struct., 43 (2006), 6574–6593. https://doi.org/10.1016/j.ijsolstr.2006.01.012 doi: 10.1016/j.ijsolstr.2006.01.012
[128]	Z. Qiu, Y. Xia, H. Yang, The static displacement and the stress analysis of structures with bounded uncertainties using the vertex solution theorem, Comput. Method. Appl. Mech. Eng., 196 (2007), 4965–4984. https://doi.org/10.1016/j.cma.2007.06.022 doi: 10.1016/j.cma.2007.06.022
[129]	Z. P. Qiu, S. H. Chen, I. Elishakoff, Bounds of eigenvalues for structures with an interval description of uncertain-but-non-random parameters, Chaos Solition. Fract., 7 (1996), 425–434. https://doi.org/10.1016/0960-0779(95)00065-8 doi: 10.1016/0960-0779(95)00065-8
[130]	Z. P. Qiu, X. J. Wang, Parameter perturbation method for dynamic responses of structures with uncertain-but-bounded parameters based on interval analysis, Int. J. Solids Struct., 42 (2005), 4958–4970. https://doi.org/10.1016/j.ijsolstr.2005.02.023 doi: 10.1016/j.ijsolstr.2005.02.023
[131]	Z. Qiu, L. Ma, X. Wang, Non-probabilistic interval analysis method for dynamic response analysis of nonlinear systems with uncertainty, J. Sound Vib., 319 (2009), 531–540. https://doi.org/10.1016/j.jsv.2008.06.006 doi: 10.1016/j.jsv.2008.06.006
[132]	D. Meng, T. Xie, P. Wu, C. He, Z. Hu, Z. Lv, An uncertainty-based design optimization strategy with random and interval variables for multidisciplinary engineering systems, Structures, 32 (2021), 997–1004. https://doi.org/10.1016/j.istruc.2021.03.020 doi: 10.1016/j.istruc.2021.03.020
[133]	X. Liu, X. Yu, J. Tong, X. Wang, X. Wang, Mixed uncertainty analysis for dynamic reliability of mechanical structures considering residual strength, Reliab. Eng. Syst. Safe., 209 (2021), 107472. https://doi.org/10.1016/j.ress.2021.107472 doi: 10.1016/j.ress.2021.107472
[134]	J. Wang, Z. Lu, Probabilistic safety model and its efficient solution for structure with random and interval mixed uncertainties, Mech. Mach. Theory, 147 (2020), 103782. https://doi.org/10.1016/j.mechmachtheory.2020.103782 doi: 10.1016/j.mechmachtheory.2020.103782
[135]	D. M. Do, K. Gao, W. Yang, C. Q. Li, Hybrid uncertainty analysis of functionally graded plates via multiple-imprecise-random-field modelling of uncertain material properties, Comput. Method. Appl. Mech. Eng., 368 (2020), 113116. https://doi.org/10.1016/j.cma.2020.113116 doi: 10.1016/j.cma.2020.113116
[136]	J. Wang, Z. Lu, M. Zhong, T. Wang, C. Sun, H. Li, Coupled thermal–structural analysis and multi-objective optimization of a cutting-type energy-absorbing structure for subway vehicles, Thin-Wall. Struct., 141 (2019), 360-373. https://doi.org/10.1016/j.tws.2019.04.026 doi: 10.1016/j.tws.2019.04.026
[137]	M. Luo, Y. Chen, D. Gao, L. Wang, Inversion study of vehicle frontal collision and front bumper collision, Electron. Res. Arch. 31 (2023), 776–792. https://doi.org/10.3934/era.2023039
[138]	W. Li, J. Wang, Z. Du, H. Ma, L. Zhang, L. Duan, Lightweight design method and application of MEWP bracket based on multi-level optimization, Electron. Res. Arch., 30 (2022), 4416–4435. https://doi.org/10.3934/era.2022224 doi: 10.3934/era.2022224
[139]	X. Xu, G. Xu, J. Chen, Z. Liu, X. Chen, Y. Zhang, et al., Multi-objective design optimization using hybrid search algorithms with interval uncertainty for thin-walled structures, Thin-Wall. Struct., 175 (2022), 109218. https://doi.org/10.1016/j.tws.2022.109218 doi: 10.1016/j.tws.2022.109218
[140]	Y. Wu, W. Li, J. Fang, Q. Lan, Multi-objective robust design optimization of fatigue life for a welded box girder, Eng. Optimiz., 50 (2018), 1252–1269. https://doi.org/10.1080/0305215X.2017.1395023 doi: 10.1080/0305215X.2017.1395023
[141]	Y. Wu, L. Sun, P. Yang, J. Fang, W. Li, Energy absorption of additively manufactured functionally bi-graded thickness honeycombs subjected to axial loads, Thin-Wall. Struct., 164 (2021), 107810. https://doi.org/10.1016/j.tws.2021.107810 doi: 10.1016/j.tws.2021.107810
[142]	Y. Su, H. Tang, S. Xue, D. Li, Multi-objective differential evolution for truss design optimization with epistemic uncertainty, Adv. Struct. Eng., 19 (2016), 1403–1419. https://doi.org/10.1177/1369433216643250 doi: 10.1177/1369433216643250
[143]	X. Liu, Q. Fu, N. Ye, L. Yin, The multi-objective reliability-based design optimization for structure based on probability and ellipsoidal convex hybrid model, Struct. Safe., 77 (2019), 48–56. https://doi.org/10.1016/j.strusafe.2018.11.004 doi: 10.1016/j.strusafe.2018.11.004
[144]	T. Vo-Duy, D. Duong-Gia, V. Ho-Huu, T. Nguyen-Thoi, An effective couple method for reliability-based multi-objective optimization of truss structures with static and dynamic constraints, Int. J. Comput. Method., 17 (2020). https://doi.org/10.1142/S0219876219500166
[145]	F. S. Lobato, M. A. da Silva, A. A. Cavalini, V. Steffen, Reliability-based robust multi-objective optimization applied to engineering system design, Eng. Optimiz., 52 (2020), 1–21. https://doi.org/10.1080/0305215X.2019.1577413 doi: 10.1080/0305215X.2019.1577413
[146]	P. Wang, S. Zheng, G. Wu, Multidisciplinary design optimization of vehicle body structure based on collaborative optimization and multi-objective genetic algorithm, Chinese J. Mech. Eng., 47 (2011), 102–108. https://doi.org/10.3901/JME.2011.02.102 doi: 10.3901/JME.2011.02.102
[147]	H. Yin, H. Fang, Y. Xiao, G. Wen, Q. Qing, Multi-objective robust optimization of foam-filled tapered multi-cell thin-walled structures, Struct. Multidiscip. Optimiz., 52 (2015), 1051–1067. https://doi.org/10.1007/s00158-015-1299-8 doi: 10.1007/s00158-015-1299-8
[148]	A. Khakhali, N. Nariman-zadeh, A. Darvizeh, A. Masoumi, B. Notghi, Reliability-based robust multi-objective crashworthiness optimisation of S-shaped box beams with parametric uncertainties, Int. J. Crashworthines., 15 (2010), 443–456. https://doi.org/10.1080/13588261003696458 doi: 10.1080/13588261003696458
[149]	A. U. Ebenuwa, K. F. Tee, Y. Zhang, Fuzzy-based multi-objective design optimization of buried pipelines, Int. J. Uncertain. Fuzz., 29 (2021), 209–229. https://doi.org/10.1142/S0218488521500104 doi: 10.1142/S0218488521500104
[150]	E. Untiedt. A Parametrized Model for Optimization with Mixed Fuzzy and Possibilistic Uncertainty, Springer, Berlin, 2010. https://doi.org/10.1007/978-3-642-13935-2_9
[151]	X. Liu, Z. Zhang, L. Yin, A multi-objective optimization method for uncertain structures based on nonlinear interval number programming method, Mech. Based Des. Struct. Mach., 45 (2017), 25–42. https://doi.org/10.1080/15397734.2016.1141365 doi: 10.1080/15397734.2016.1141365
[152]	T. Xie, Q. Zhou, J. Hu, L. Shu, P. Jiang, A sequential multi-objective robust optimization approach under interval uncertainty based on support vector machines, in 2017 IEEE International Conference on Industrial Engineering and Engineering Management (IEEM), 2017, 2088–2092. https://doi.org/10.1109/IEEM.2017.8290260
[153]	H. Chagraoui, M. Soula, M. Guedri, Robust multi-objective collaborative optimization of complex structures. Adv. Acous. Vib., 5 (2017), 247–258. https://doi.org/10.1007/978-3-319-41459-1_24 doi: 10.1007/978-3-319-41459-1_24
[154]	F. Li, G. Li, G. Sun, Z. Luo, Z. Zhang, Multi-disciplinary optimization for multi-objective uncertainty design of thin walled beams, CMC-Comput. Mater. Con., 19 (2010), 37–56. http://hdl.handle.net/10453/22026
[155]	X. Zhang, W. Sun, J. Huo, X. Ding, Interval multi-objective optimization design based on physical programming, Przeglad Elektrotechiczhy, 88 (2012), 379–381.
[156]	J. Cheng, G. Duan, Z. Liu, X. Li, Y. Feng, X. Chen, Interval multiobjective optimization of structures based on radial basis function, interval analysis, and NSGA-Ⅱ, J. Zhejiang Univ-Sci. A, 15 (2014), 774–788. https://doi.org/10.1631/jzus.A1300311
[157]	X. Feng, J. Wu, Y. Zhang, M. Jiang, Suspension kinematic/compliance uncertain optimization using a chebyshev polynomial approach, SAE Int. J. Mater. Manu., 8 (2015), 257–262. https://doi.org/10.4271/2015-01-0432 doi: 10.4271/2015-01-0432
[158]	X. Li, C. Jiang, X. Han, An uncertainty multi-objective optimization based on interval analysis and its application, China Mech. Eng., 22 (2011), 1100–1106. http://www.cmemo.org.cn/EN/Y2011/V22/I9/1100
[159]	C. Jiang, H. C. Xie, Z. G. Zhang, X. Han, A new interval optimization method considering tolerance design, Eng. Optimiz., 47 (2015), 1637–1650. https://doi.org/10.1080/0305215X.2014.982632 doi: 10.1080/0305215X.2014.982632

This article has been cited by:

Luis C. García-Naranjo, Rafael Ortega, Antonio J. Ureña, Invariant Measures as Obstructions to Attractors in Dynamical Systems and Their Role in Nonholonomic Mechanics, 2024, 29, 1560-3547, 751, 10.1134/S156035472456003X

Reader Comments

Your name:*

Email:*
© 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)