Unilateral contact
   HOME

TheInfoList



OR:

In
contact mechanics Contact mechanics is the study of the deformation of solids that touch each other at one or more points.Johnson, K. L, 1985, Contact mechanics, Cambridge University Press.Popov, Valentin L., 2010, ''Contact Mechanics and Friction. Physical P ...
, the term unilateral contact, also called unilateral constraint, denotes a mechanical constraint which prevents penetration between two rigid/flexible bodies. Constraints of this kind are omnipresent in non-smooth multibody dynamics applications, such as granular flows,
legged robot Legged robots are a type of mobile robot which use articulated limbs, such as leg mechanisms, to provide locomotion. They are more versatile than wheeled robots and can traverse many different terrains, though these advantages require increased ...
,
vehicle dynamics For motorized vehicles, such as automobiles, aircraft, and watercraft, vehicle dynamics is the study of vehicle motion, e.g., how a vehicle's forward movement changes in response to driver inputs, propulsion system outputs, ambient conditions, air ...
, particle damping, imperfect joints, or rocket landings. In these applications, the unilateral constraints result in impacts happening, therefore requiring suitable methods to deal with such constraints.


Modelling of the unilateral constraints

There are mainly two kinds of methods to model the unilateral constraints. The first kind is based on smooth contact dynamics, including methods using Hertz's models, penalty methods, and some regularization force models, while the second kind is based on the non-smooth contact dynamics, which models the system with unilateral contacts as
variational inequalities In mathematics, a variational inequality is an inequality involving a functional, which has to be solved for all possible values of a given variable, belonging usually to a convex set. The mathematical theory of variational inequalities was initi ...
.


Smooth contact dynamics

In this method, normal forces generated by the unilateral constraints are modelled according to the local material properties of bodies. In particular, contact force models are derived from continuum mechanics, and expressed as functions of the gap and the impact velocity of bodies. As an example, an illustration of the classic Hertz contact model is shown in the figure on the right. In such model, the contact is explained by the local deformation of bodies. More contact models can be found in some review scientific works or in the article dedicated to
contact mechanics Contact mechanics is the study of the deformation of solids that touch each other at one or more points.Johnson, K. L, 1985, Contact mechanics, Cambridge University Press.Popov, Valentin L., 2010, ''Contact Mechanics and Friction. Physical P ...
.


Non-smooth contact dynamics

In non-smooth method, unilateral interactions between bodies are fundamentally modelled by the Signorini condition for non-penetration, and impact laws are used to define the impact process. The Signorini condition can be expressed as the complementarity problem: g \geq 0, \quad \lambda \geq 0, \quad \lambda \perp g , where g denotes the distance between two bodies and \lambda denotes the contact force generated by the unilateral constraints, as shown in the figure below. Moreover, in terms of the concept of proximal point of convex theory, the Signorini condition can be equivalently expressed as: \lambda =_(\lambda -\rho g ), where \rho>0 denotes an auxiliary parameter, and _(x) represents the proximal point in the set C to the variable x, defined as: _(x)=_\, y-x\, . Both the expressions above represent the dynamic behaviour of unilateral constraints: on the one hand, when the normal distance g_ is above zero, the contact is open, which means that there is no contact force between bodies, \lambda =0 ; on the other hand, when the normal distance g_ is equal to zero, the contact is closed, resulting in \lambda \geq0. When implementing non-smooth theory based methods, the velocity Signorini condition or the acceleration Signorini condition are actually employed in most cases. The velocity Signorini condition is expressed as: U_^\geq 0,\quad \lambda \geq0,\quad U^\lambda =0, where U_^ denotes the relative normal velocity after impact. The velocity Signorini condition should be understood together with the previous conditions g \geq 0,\;\lambda \geq 0,\;\lambda \perp g . The acceleration Signorini condition is considered under closed contact (g =0, U_^=0), as: \ddot g \geq 0,\quad \lambda \geq0,\quad \ddot g \lambda =0, where the overdots denote the second-order derivative with respect to time. When using this method for unilateral constraints between two rigid bodies, the Signorini condition alone is not enough to model the impact process, so impact laws, which give the information about the states before and after the impact, are also required. For example, when the Newton restitution law is employed, a
coefficient of restitution The coefficient of restitution (COR, also denoted by ''e''), is the ratio of the final to initial relative speed between two objects after they collide. It normally ranges from 0 to 1 where 1 would be a perfectly elastic collision. A perfectl ...
will be defined as: e=-/, where U_^denotes the relative normal velocity before impact.


Frictional unilateral constraints

For frictional unilateral constraints, the normal contact forces are modelled by one of the methods above, while the friction forces are commonly described by means of Coulomb's friction law. Coulomb's friction law can be expressed as follows: when the tangential velocity U_ is not equal to zero, namely when the two bodies are sliding, the friction force \lambda_ is proportional to the normal contact force \lambda; when instead the tangential velocity U_ is equal to zero, namely when the two bodies are relatively steady, the friction force \lambda_ is no more than the maximum of the static friction force. This relationship can be summarised using the maximum dissipation principle, as \lambda_ \in D(\mu \lambda)~~~~~~\forall S\in D(\mu \lambda)~~~~~~(S-\lambda_)U_\geq 0, where D(\mu \lambda)=\ represents the friction cone, and \mu denotes the kinematic friction coefficient. Similarly to the normal contact force, the formulation above can be equivalently expressed in terms of the notion of proximal point as: \lambda_=_(\lambda_T-\rho U_), where \rho>0 denotes an auxiliary parameter.


Solution techniques

If the unilateral constraints are modelled by the continuum mechanics based contact models, the contact forces can be computed directly through an explicit mathematical formula, that depends on the contact model of choice. If instead the non-smooth theory based method is employed, there are two main formulations for the solution of the Signorini conditions: the
nonlinear In mathematics and science, a nonlinear system is a system in which the change of the output is not proportional to the change of the input. Nonlinear problems are of interest to engineers, biologists, physicists, mathematicians, and many other ...
/
linear complementarity problem In mathematical optimization theory, the linear complementarity problem (LCP) arises frequently in computational mechanics and encompasses the well-known quadratic programming as a special case. It was proposed by Cottle and Dantzig in 1968. ...
(N/LCP) formulation and the augmented Lagrangian formulation. With respect to the solution of contact models, the non-smooth method is more tedious, but less costly from the computational viewpoint. A more detailed comparison of solution methods using contact models and non-smooth theory was carried out by Pazouki et al.


N/LCP formulations

Following this approach, the solution of dynamics equations with unilateral constraints is transformed into the solution of N/LCPs. In particular, for frictionless unilateral constraints or unilateral constraints with planar friction, the problem is transformed into LCPs, while for frictional unilateral constraints, the problem is transformed into NCPs. To solve LCPs, the pivoting algorithm, originating from the algorithm of Lemek and Dantzig, is the most popular method. Unfortunately, however, numerical experiments show that the pivoting algorithm may fail when handling systems with a large number of unilateral contacts, even using the best optimizations. For NCPs, using a polyhedral approximation can transform the NCPs into a set of LCPs, which can then be solved by the LCP solver. Other approaches beyond these methods, such NCP-functions or cone complementarity problems (CCP) based methods are also employed to solve NCPs.


Augmented Lagrangian formulation

Different from the N/LCP formulations, the augmented Lagrangian formulation uses the proximal functions described above, \lambda=_(\lambda-\rho g). Together with dynamics equations, this formulation is solved by means of
root-finding algorithms In mathematics and computing, a root-finding algorithm is an algorithm for finding zeros, also called "roots", of continuous functions. A zero of a function , from the real numbers to real numbers or from the complex numbers to the complex numbers ...
. A comparative study between LCP formulations and the augmented Lagrangian formulation was carried out by Mashayekhi et al.


See also

* * * * * * Collision response *
Variational inequalities In mathematics, a variational inequality is an inequality involving a functional, which has to be solved for all possible values of a given variable, belonging usually to a convex set. The mathematical theory of variational inequalities was initi ...


References


Further reading


Open-source software

Open-source codes and non-commercial packages using the non-smooth based method: * {{annotated link, Siconos
Chrono
an open source multi-physics simulation engine, see also projec
website


Books and articles

*Acary V., Brogliato B. Numerical Methods for Nonsmooth Dynamical Systems. Applications in Mechanics and Electronics. ''Springer Verlag, LNACM 35, Heidelberg, 2008.'' *Brogliato B. Nonsmooth Mechanics. ''Communications and Control Engineering Series'' Springer-Verlag, London, 1999 (2dn Ed.) * Demyanov, V.F., Stavroulakis, G.E., Polyakova, L.N., Panagiotopoulos, P.D. "Quasidifferentiability and Nonsmooth Modelling in Mechanics, Engineering and Economics" Springer 1996 *Glocker, Ch. ''Dynamik von Starrkoerpersystemen mit Reibung und Stoessen'', volume 18/182 of ''VDI Fortschrittsberichte Mechanik/Bruchmechanik.'' VDI Verlag, Düsseldorf, 1995 *Glocker Ch. and Studer C. Formulation and preparation for Numerical Evaluation of Linear Complementarity Systems. ''Multibody System Dynamics'' 13(4):447-463, 2005 *Jean M. The non-smooth contact dynamics method. ''Computer Methods in Applied mechanics and Engineering'' 177(3-4):235-257, 1999 *Moreau J.J. '' Unilateral Contact and Dry Friction in Finite Freedom Dynamics,'' volume 302 of '' Non-smooth Mechanics and Applications, CISM Courses and Lectures''. Springer, Wien, 1988 *Pfeiffer F., Foerg M. and Ulbrich H. Numerical aspects of non-smooth multibody dynamics. ''Comput. Methods Appl. Mech. Engrg'' 195(50-51):6891-6908, 2006 *Potra F.A., Anitescu M., Gavrea B. and Trinkle J. A linearly implicit trapezoidal method for integrating stiff multibody dynamics with contacts, joints and friction. ''Int. J. Numer. Meth. Engng'' 66(7):1079-1124, 2006 *Stewart D.E. and Trinkle J.C. An Implicit Time-Stepping Scheme for Rigid Body Dynamics with Inelastic Collisions and Coulomb Friction. ''Int. J. Numer. Methods Engineering'' 39(15):2673-2691, 1996 *Studer C. ''Augmented time-stepping integration of non-smooth dynamical systems'', PhD Thesis ETH Zurich, ETH E-Collection, to appear 2008 *Studer C. ''Numerics of Unilateral Contacts and Friction -- Modeling and Numerical Time Integration in Non-Smooth Dynamics'', Lecture Notes in Applied and Computational Mechanics, Volume 47, Springer, Berlin, Heidelberg, 2009 Mechanics