Beyond the classical Cauchy-Born rule Andrea Braides SISSA via Bonomea 265 Trieste Italy

2025-04-27 0 0 2.11MB 117 页 10玖币
侵权投诉
Beyond the classical Cauchy-Born rule
Andrea Braides
SISSA, via Bonomea 265, Trieste, Italy
Andrea Causin and Margherita Solci
DADU, Universit`a di Sassari
piazza Duomo 6, 07041 Alghero (SS), Italy
Lev Truskinovsky
PMMH, CNRS - UMR 7636 PSL-ESPCI,
10 Rue Vauquelin, 75005 Paris, France
Abstract
Physically motivated variational problems involving non-convex energies are often for-
mulated in a discrete setting and contain boundary conditions. The long-range interactions
in such problems, combined with constraints imposed by lattice discreteness, can give rise
to the phenomenon of geometric frustration even in a one-dimensional setting. While non-
convexity entails the formation of microstructures, incompatibility between interactions
operating at different scales can produce nontrivial mixing effects which are exacerbated
in the case of incommensuration between the optimal microstructures and the scale of the
underlying lattice. Unraveling the intricacies of the underlying interplay between non-
convexity, non-locality and discreteness, represents the main goal of this study. While in
general one cannot expect that ground states in such problems possess global properties,
such as periodicity, in some cases the appropriately defined ‘global’ solutions exist, and
are sufficient to describe the corresponding continuum (homogenized) limits. We interpret
those cases as complying with a Generalized Cauchy-Born (GCB) rule, and present a new
class of problems with geometrical frustration which comply with GCB rule in one range
of (loading) parameters while being strictly outside this class in a complimentary range.
A general approach to problems with such ‘mixed’ behavior is developed.
1 Introduction
Variational problems emerging from applications are often both discrete and non-convex. Im-
portant examples include one-dimensional boundary-value problems with translation-invariant
energy densities describing pairwise interactions. Such problems constitute the main subject
of this paper.
1
arXiv:2210.06147v2 [math.AP] 15 Oct 2022
The representative energies for this class of problems can be written in the following generic
form
F(w;k) = mink
X
i,j=0
fij(uiuj) : u0= 0, uk=w,(1.1)
where for every nnatural number fnis a potentially nonconvex energy governing interactions
between the lattice points at distance n, and the minimum is searched among k+ 1-arrays
(u0, . . . , uk). We may assume that f0= 0. As the parameter kincreases and more interactions
are taken into account, a question arises about the behavior of minimal arrays (uk
0, . . . , uk
k)
and of the corresponding minimal energy. One of the most important issues concerns the
existence of a continuum limit of the type Fhom(u) = RIfhom(u0)dt, with Ian interval in
which the nodes iin (1.1) are identified as a discrete subset (e.g., I= [0,1] where the discrete
subset is 1
kZ[0,1]). The single function fhom is expected to carry, in a condensed way, all
the relevant information about the infinite set of functions fnfrom (1.1).
To track the asymptotic behavior of the minimum values in (1.1) we can use the average
derivative z=w/k as a parameter, and scale the energy by k. Then, under assumptions on a
suitably fast decay of fnwith respect to n, it can be shown that the limiting energy density
fhom exists and can be expressed by the formula
fhom(z) = lim
k+
1
kmink
X
i,j=0
fij(uiuj) : u0= 0, uk=kz.(1.2)
Moreover, it can be shown that the function fhom is convex in the parameter z. This result
represents a particular case of a more general variational theory for limits of lattice energies
(see e.g. [3]); it can be also seen as a zero-temperature limit of the analogous result in Statistical
Physics ([81, 79]). However, formula (1.2) is only a formal homogenization result in a discrete-
to-continuum setting which is usually non-constructive. In this paper we are raising the issue
of the actual computability of fhom(z).
Explicit formulas for fhom(z) in terms of fnare known only in few cases, most of which are
mentioned below. In general, it is known that the behavior of minimizing arrays (uk
0, . . . , uk
k) at
fixed z, may be complex, including equi-distribution (‘crystallization’; see e.g. [66]), periodic
oscillations [24, 49], development of discontinuities (fracture in lattice models [84, 27]) or
defects (internal boundary layers [22]).
A robust approach to the computation of fhom(z) is known under the name of Cauchy-
Born (CB) rule and is applicable under some restrictive conditions ([41, 15]). It is based on
the assumption that the homogenized energy can be computed using the affine interpolations
uj=zj and relying exclusively on problems with finite k. Various sufficient conditions for the
validity of the Cauchy-Born rule have been obtained by a number of authors mostly in the
context of local minimizers [40, 58, 67, 76, 85, 34, 86]. While those results are usually valid
only for subsets of loading parameters, they are often applicable for dimensions higher than
one. They are of considerable interest, first of all, for the development of numerical methods
2
because the applicability of the classical CB rule makes such methods extremely efficient, even
if for a limited set of boundary conditions. The difference of our approach to (1.2) is that
we are interested in global minimization (viewed as a zero temperature limit of a statistically
equilibrium response) and consider the possibility that the conventional CB rule is operative
only in a subset of the loading parameters while in the complementary subset the CB strategy
should be appropriately generalized or even completely ruled out.
The main reason for the failure of the classical Cauchy-Born rule is the geometrical frus-
tration caused by incompatible optimality demands imposed by (generically non-convex and
long range) potentials fnwith positive integer nand the discreteness of the lattice. More
specifically, while non-convexity entails the formation of microstructures, incompatibility be-
tween interactions operating at different scales can produce nontrivial mixing effects which
are exacerbated in the case of incommensuration between the optimal microstructures and the
scale of the underlying lattice. Unraveling the intricacies of the underlying interplay between
non-convexity, non-locality and discreteness, represents the main goal of this study.
If the classical Cauchy-Born rule fails, the natural task is to search for a nontrivial gen-
eralization of the Cauchy-Born rule. In this perspective, we pose the problem of finding the
conditions for which the minimal arrays in (1.1) have ‘global’ features in the sense that solv-
ing a ‘local’ problem on a finite domain opens the way towards describing the limit in (1.2).
More specifically, the question is whether the limiting energy fhom(z) can be approximately
computed by solving a finite set of ‘cell’ problems modeled on (1.1) and potentially producing
non-affine optimal configurations. The validity of the so-interpreted generalized Cauchy-Born
(GCB) rule would then require that even if the implied ‘local’ problems could be solved only
on some subsets of parameters, the knowledge of the corresponding solutions would ensure the
recovery of the macroscopic (homogenized) energy in the whole range of loading parameters.
Note that in local problems like (1.1) the presence of interactions fnwith n∈ {1, . . . , k}
requires kboundary conditions on each side. By fixing parametrically only the average strain
zin (1.2) we effectively assume that the remaining boundary conditions are natural. This
simplifying assumption may stay on the way of acquiring, for the given ‘local’ problem, the
corresponding ‘global’ features. That is why we will understand the ‘local’ GCB problem
as having the right boundary conditions to ensure the recovery of the macroscopic energy
fhom(z). The simplest case is when the value of fhom(z) can be achieved on arrays such that
i7→ uizi is periodic with a given period, but in general one should be allowed to adjust
boundary conditions accordingly while keeping in mind that these changes should not affect
the minimizers in an asymptotic sense.
We now illustrate the main difficulties on the way of generalizing the classical CB rule
with some known cases. We start with the simplest example where the conventional CB rule
works trivially. It is the case of convex nearest-neighbor (NN) interactions; i.e., when fn= 0
for all n2, and f1=fis a strictly convex function. In this case, the unique minimizer
of the problem in (1.2) is the affine interpolation uk
j=zj. It is independent of kand hence
‘global’: in this case the classical Cauchy-Born rule is applicable in its simplest form, and
3
fhom(z) = f(z).
If we make the above example only a little more complex considering also convex next-
to-nearest-neighbour (NNN) interactions; i.e., fn= 0 for all n3, with f1and f2convex
functions, we loose this exact characterization of the minimal arrays. However, the discrepancy
between uk
jand zj decays fast away from the endpoints j= 0 and j=kof the array.
A slight adjustment of the boundary-value problems, say by imposing additional boundary
conditions u1=zand uk1=z(k1) (which do not influence the asymptotic value of
the minima in (1.2)) reestablishes the affine interpolations uk
j=zj as minimizers, so that
fhom(z) = 2(f1(z) + f2(2z)). In this case the classical Cauchy-Born rule is applicable, given
that we modify boundary conditions in the ‘cell’ problem. Note that this analysis extends to
any sufficiently fast decaying set of convex potentials fn, giving fhom(z) = 2 P
n=1 fn(nz).
Even if we abandon the convex setting, we may still easily describe the behavior of min-
imum problems in (1.2) in the case of nearest-neighbor interaction, with f1=f. It can be
shown that fhom in (1.2) is given by the convexification f∗∗ of the NN potential [25]. However
the classical Cauchy-Born rule in this case has to be properly generalized. Suppose, for in-
stance, that the potential fhas a double-well form. In this case the relaxation points towards
configurations containing mixtures of the two energy wells. Since in this setting there are no
obstacles to simple mixing, the relaxation strategy providing fhom is straightforward. Indeed,
for each zthere exist z1,z2,θ[0,1] such that f∗∗(z) = θf(z1) + (1 θ)f(z2). Hence, we
can construct a function uz:ZRwith uz
iuz
i1∈ {z1, z2},uz
0= 0 and |uz
iiz| ≤ C.
Such uzmay be chosen periodic, if θis rational, or quasiperiodic (loosely speaking, as the
trace on Zof a periodic function with an irrational period) otherwise. In both cases we obtain
‘local’ minimizers with ‘global’ properties which allows one to talk about the applicability of
the GCB rule.
The situation is more complex in the case when non-convexity is combined with frustrated
(incompatible) interactions. To show this effect in the simplest setting it is sufficient to
account for nearest-neighbor and next-to-nearest-neighbor interactions only and we make the
simplest nontrivial choice by assuming that f1is a ‘double-well’ potential and that f2is a
convex potential. In this case the homogenized potential fhom is also known explicitly [24, 77].
Its domain can be subdivided in three zones: two zones of ‘convexity’ where minimizers are
trivial (as for convex potentials) and a zone where (approximate) minimizers in (1.2) are
two-periodic functions with uz
iuz
i1∈ {z1, z2}and z1+z2=z(in a sense, a constrained
non-convex case as above). Hence, in these three zones we have minimizers with a ‘global’
form because the macroscopic energy can be obtained by solving elementary ‘cell’ problems.
One can say that in the two zones of ‘convexity’ the classical CB rule is applicable. In
the ‘two-periodic’ third zone we see that the homogeneity of the minimizers is lost but an
appropriately augmented GCB rule still holds. For the remaining values of zno ‘local’ GCB
rule is applicable since in those cases the unique (up to reflections) minimizer is a ‘two-phase’
configuration with affine and two-periodic minimizers coexisting while being separated by a
single ‘interface’ [22]. The frustration (incompatibility) manifests itself in this case through
the impossibility of the penalty-free accommodation of next-to-nearest interactions across such
4
an internal boundary layer. As a consequence, as kdiverges, such minimizers tend to an affine
interpolation between the ‘convex’ and ‘oscillating’ zones which delivers the correct value of
fhom(z) without being a solution of any finite ‘cell’ problem. Effectively, the ‘representative
cell’ in this case has an infinite size and therefore no GCB-type ‘local’ description of the
macroscopic state is available. A somewhat similar situation is encountered in continuum
homogenization of both random [62] and strongly nonlinear [23, 72] elastic composites.
In what follows, we interpret the loss of ‘locality’ in homogenization problems, which was
illustrated above on the simplest example, as a failure of the GCB rule. To shed some light
on the mechanism of this phenomenon, we consider below a class of analytically transparent
discrete problems combining nonconvexity with geometrical frustration.
More specifically, given the complexity of a general asymptotic analysis for even one-
dimensional problems of this type, we limit our attention to a class of discrete functionals of
type (1.1) with f1(z) = 1
2f(z) + m1z2, where the function f(z) is non-convex, and quadratic
fn(z) = fn(z) = mnz2for n2. The coefficients mnwhich introduce nonlocality and frus-
tration, are assumed to be non negative and sufficiently integrable. In other words, we suppose
that the non-convexity is ‘localized’ in the nearest-neighbor interactions, while all other inter-
actions are quadratic. The positivity of the infinite sequence m={mn:n1}is chosen to
ensure that the implied quadratic ‘penalty’ is a measure of the distance of the configuration
uifrom the affine configuration Lz(i) = zi and can be then seen as a non-local version of the
gradient of uLz. One can also say that such penalization brings anti-ferromagnetic inter-
actions; an alternative, ferromagnetic-type quadratic penalty, was considered, for instance, in
[80].
The advantage of this choice of fnis that the ensuing problem can exhibit both ‘local’
(GCB) and ‘global’ behavior depending on the structure of the sequence of scalar parameters
mn. Therefore our goal will be to use the chosen class of functionals to characterize the
difference between CB, GCB and non-GCB problems in terms of such sequences. We show
that in this naturally limited but still sufficiently rich framework one can precisely specify
the factors preventing the GCB-type description of the macroscopic energy and pointing
instead towards the non-GCB nature of the minimizers. Moreover, the considered example
allow us to abstract some general technical tools which can facilitate the detection and the
characterization of the non-GCB asymptotic behavior in more general minimization problems.
We reiterate that even in the absence of an adequate ‘cell’ problem, the ensuing value of
fhom(z) is fully determined by the homogenization formula which in our case takes the form
fhom(z) = b
Qmf(z) where
b
Qmf(z) = lim
k+
1
kminnk
X
i=1
f(uiui1) +
k
X
i,j=0
mij(uiuj)2:u0= 0, uk=kzo.(1.3)
The nontrivial part of the mapping b
Qmf, accentuating the nonlinearity of the problem, is
carried by the operator Qmf(z) = b
Qmf(z)2Pn1mnn2z2.Thus, if fis convex, this
mapping, to which we refer as the m-transform of f, is the identity; actually, the same
5
摘要:

BeyondtheclassicalCauchy-BornruleAndreaBraidesSISSA,viaBonomea265,Trieste,ItalyAndreaCausinandMargheritaSolciDADU,UniversitadiSassaripiazzaDuomo6,07041Alghero(SS),ItalyLevTruskinovskyPMMH,CNRS-UMR7636PSL-ESPCI,10RueVauquelin,75005Paris,FranceAbstractPhysicallymotivatedvariationalproblemsinvolvingno...

展开>> 收起<<
Beyond the classical Cauchy-Born rule Andrea Braides SISSA via Bonomea 265 Trieste Italy.pdf

共117页,预览5页

还剩页未读, 继续阅读

声明:本站为文档C2C交易模式,即用户上传的文档直接被用户下载,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私,请立即通知玖贝云文库,我们立即给予删除!
分类:图书资源 价格:10玖币 属性:117 页 大小:2.11MB 格式:PDF 时间:2025-04-27

开通VIP享超值会员特权

  • 多端同步记录
  • 高速下载文档
  • 免费文档工具
  • 分享文档赚钱
  • 每日登录抽奖
  • 优质衍生服务
/ 117
客服
关注