## Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

Ratings:

Length: 364 pages5 hours

Stringent industrial requirements for sophisticated performance and circumstantial control of microdevices or nanotechnology manufacturing, and other types of machinery at multiple scales, require complex materials. The adjective ‘complex’ indicates that the substructure influences gross mechanical behaviour in a prominent way and interactions due to substructural changes are represented directly. Examples are liquid crystals, quasi-periodic alloys, polymeric bodies, spin glasses, magnetostrictive materials and ferroelectrics, suspensions, in particular liquids with gas bubbles, polarizable fluids, etc.

The mechanical behaviour of complex bodies described in this book gives rise to a wide variety of challenging problems from the macroscopic- to the nano-world. The chapters composing this book explore various aspects of these problems, giving rise to new areas of discussion together with specific solutions.

Contributors are Carlo Cercignani, Gianfranco Capriz, Pierre Degond, Antonio Fasano, Harley T. Johnson, Sukky Jun, Krishna Kannan, Wing Kam Liu, Alberto Mancini, Paolo Maria Mariano, Ingo Müller, Kumbakonan R. Rajagopal, Jan Jerzy Slawianowski.

*Material Substructures in Complex Bodies *will be a useful tool for researchers and students working on the basic mathematical and physical problems accruing from the mechanics of materials.

Publisher: Elsevier ScienceReleased: Jul 28, 2011ISBN: 9780080554716Format: book

You've reached the end of this preview. Sign up to read more!

Page 1 of 1

**Material Substructures in Complex Bodies: From Atomic Level to Continuum **

First edition

Gianfranco Capriz

*Università di Pisa *

Paolo Maria Mariano

*D.I.C., Università di Firenze *

ELSEVIER

Amsterdam • Boston • Heidelberg • London • New York • Oxford

Paris • San Diego • San Francisco • Singapore • Sydney • Tokyo

**Cover image **

**Title page **

**Copyright page **

**Contributors **

**Preface **

**Chapter One: Asymptotic Continuum Models for Plasmas and Disparate Mass Gaseous Binary Mixtures **

**Abstract **

**1.1 INTRODUCTION **

**1.2 THE KINETIC MODEL **

**1.3 MOMENT METHOD AND CONSERVATION LAWS FOR GAS MIXTURES: WHY IT CANNOT APPLY TO PLASMAS **

**1.4 THE PLASMA FLUID MODEL **

**1.5 SCALING HYPOTHESES **

**1.6 EXPANSION OF THE INTERSPECIES COLLISION OPERATORS **

**1.7 MOMENT METHOD AND CONSERVATION LAWS FOR PLASMAS **

**1.8 COMPUTATION OF THE FLUXES AND OF THE COLLISION TERMS **

**1.9 CONCLUSION **

**Appendix PROPERTIES OF ℒB (PROOFS) **

**ACKNOWLEDGEMENTS **

**Chapter Two: Microscopic Foundations of the Mechanics of Gases and Granular Materials **

**Abstract **

**2.1 INTRODUCTION **

**2.2 KINETIC THEORY OF SMOOTH SPHERES **

**2.3 COLLISION DYNAMICS OF ROUGH SPHERES **

**2.4 THE BOLTZMANN–ENSKOG EQUATION **

**2.5 THE MACROSCOPIC BALANCE EQUATIONS **

**2.6 CONCLUDING REMARKS **

**Chapter Three: Quantization of Affine Bodies: Theory and Applications in Mechanics of Structured Media **

**Abstract **

**3.1 INTRODUCTION **

**3.2 CLASSICAL PRELIMINARIES **

**3.3 GENERAL IDEAS OF QUANTIZATION **

**ACKNOWLEDGEMENTS **

**Chapter Four: Moving Least-Square Basis for Band-Structure Calculations of Natural and Artificial Crystals **

**Abstract **

**ACKNOWLEDGEMENTS **

**4.1 INTRODUCTION **

**4.2 MLS BASIS AND PERIODICITY **

**4.3 ATOMIC CRYSTALS AND SEMICONDUCTORS **

**4.4 PHOXONIC CRYSTALS **

**4.5 STRAIN-TUNABLE PHOTONIC BANDGAP MATERIALS **

**4.6 CONCLUDING REMARKS **

**Chapter Five: Modelling Ziegler–Natta Polymerization in High Pressure Reactors **

**Abstract **

**5.1 INTRODUCTION **

**MODELLING THE GROWTH OF THE AGGLOMERATE (MACROSCALE) **

**MODELLING THE GROWTH OF MICROSPHERES **

**BRIDGING THE TWO SCALES. THE COMPLETE MODEL **

**Chapter Six: Pseudofluids **

**Abstract **

**6.1 Preamble **

**6.2 MATERIAL ELEMENT **

**6.3 BASIC FIELDS **

**6.4 MEASURES OF DEFORMATION AND DISTORSION **

**6.5 STRAIN RATES AND DISTORSION RATES **

**6.6 INERTIA MEASURES **

**6.7 RELATIONS WITH THERMAL CONCEPTS **

**6.8 BALANCE EQUATIONS **

**6.9 BOUNDARY CONDITIONS: SAMPLE FLOWS **

**6.10 A LAGRANGIAN APPROACH **

**CLOSING REMARK AND ACKNOWLEDGEMENT **

**Chapter Seven: A Thermodynamical Framework Incorporating the Effect of the Thermal History on the Solidification of Molten Polymers **

**Abstract **

**7.1 INTRODUCTION **

**7.2 KINEMATICS **

**7.3 MODELING **

**7.4 SUMMARY AND CONCLUSIONS **

**ACKNOWLEDGEMENT **

**Chapter Eight: Effects of Stress on Formation and Properties of Semiconductor Nanostructures **

**Abstract **

**8.1 OVERVIEW **

**8.2 BACKGROUND **

**8.3 EFFECTS OF STRESS ON THE FORMATION OF SEMICONDUCTOR NANOSTRUCTURES **

**8.4 STRESS EFFECTS ON THE ELECTRONIC/OPTICAL PROPERTIES OF SEMICONDUCTOR NANOSTRUCTURES **

**8.5 CONCLUSIONS **

**Chapter Nine: Continua with Spin Structure **

**Abstract **

**9.1 INTRODUCTION **

**9.2 SPATIAL REPRESENTATION **

**9.3 REFERENTIAL DESCRIPTION: INVARIANCE WITH RESPECT TO RELABELING **

**9.4 COVARIANT EVOLUTION OF INTERSTITIAL POINT DEFECTS AND DISCLINATIONS **

**ACKNOWLEDGEMENTS **

**Index **

Elsevier

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK

84 Theobald’s Road, London WC1X 8RR, UK

First edition 2007

Copyright © 2007 Elsevier BV. All rights reserved

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+ 44) (0) 1865 843830; fax (+ 44) (0) 1865 853333; email: permis-sions@elsevier.com. Alternatively you can submit your request online by visiting the Elsevier web site at **http://elsevier.com/locate/permissions, and selecting **

*Obtaining permission to use Elsevier material *

Notice

No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made

**British Library Cataloguing in Publication Data **

Material substructures in complex bodies : from atomic level to continuum 1. Microstructure 2. Materials science 3. Atomic structure 4. Molecular structure 5. Nanostructured materials I. Capriz, G. (Gianfranco) II. Mariano, Paolo Maria, 1966– 620.1′1299

**Library of Congress Control Number**: **200693845 **

ISBN 13: 978-0-08-044535-9

ISBN 10: 0-08-044535-7

For information on all Elsevier publications visit our web site at books.elsevier.com

Typeset by Charon Tec Ltd (A Macmillan Company), Chennai, India.

**www.charontec.com **

Printed and bound in Great Britain

06 07 08 09 10 10 9 8 7 6 5 4 3 2 1

**Prof. Gianfranco Capriz, gianfranco.capriz@mac.com Dipartimento di Matematica Università di Pisa, Largo B. Pontecorvo 5, I-56127 Pisa, Italy. **

**Prof. Carlo Cercignani, carlo.cercignani@mate.polimi.it Dipartimento di Matematica, Politecnico di Milano, Piazza Leonardo daVinci 32, 20133 Milano, Italy, Phone: + 39 02 2399 4557; Fax: + 39 02 2399 4606;. **

**Prof. Pierre Degond, degond@mip.ups-tlse.fr Laboratoire MIP, Université Paul Sabatier, 31062 Toulouse Cedex 9, France, Phone: + 33 05 61 55 63 69; Fax: + 33 05 61 55 83 85. **

**Prof. Antonio Fasano, fasano@math.unifi.it Dipartimento di Matematica **

U. Dini, Università di Firenze, viale Morgani 67\A, I-50134 Firenze Italy, Phone: + 39 055 4237145, Fax: + 39 055 4222695.

**Dr. Harley T. Johnson, htj@uiuc.edu Department of Mechanical Science and Engineering, University of Illinois at Urbana-Champaign, 1206 W. Green Street, Urbana IL 61801-2906, USA, Phone: + 1 217-265-5468; Fax: + 1 217-244-6534. **

**Sukky Jun, juns@fiu.edu Department of Mechanical and Materials Engineering, Florida International University, 10555 West Flagler Street, EC 3463, Miami, FL 33174, USA, Phone:+1 305 348 1217; Fax:+1 305 348 1932. **

**K. Kannan** Department of Biomedical Engineering –Texas A&M University – 230A Engineering/ Physics Building OfficeWing – 3120 TAMU.

**Prof. Wing Kam Liu, w-liu@northwestern.edu Northwestern University, Department of Mechanical Engineering, 2145 Sheridan, Evanston, IL 60208-3111, USA, Phone: + 1 847-491-7094; Fax: + 1 847-491-3915. **

**Alberto Mancini** Dipartimento di Matematica U. Dini

– Università degli Studi di Firenze – Viale Morgagni 67/a 50134 Firenze, Italy.

**Prof. Paolo Maria Mariano, paolo.mariano@unifi.it D.I.C., Università di Firenze, via Santa Marta 3, I-50139 Firenze, Italy, Phone: + 39.055.4796470; Fax: + 39.055.4796320. **

**Prof. K.R. Rajagopal, krajagopal@tamu.edu Department of Mechanical Engineering, Texas A&M University, 230A Engineering/Physics Building Office Wing College Station,Texas 77843, USA, Phone: + 1 979-862-4552; Fax: + 1 979-845-3081. **

**Prof. Jan J. Sławianowski, jslawian@ippt.gov.pl Institute of Fundamental Technological Research, Polish Academy of Sciences, ul. Swietokrzyska 21, 00-049 Warzaw, Poland, Phone: + 48 (22) 8261281; Fax: + 48 (22) 8269815. **

**Gianfranco Capriz, ***Bridport (UK) *

**Paolo Maria Mariano, ***Firenze (Italy) *

Stringent industrial requirement of sophisticated performances and of circumstantial control for micro-devices and other types of machinery at multiple scales can be satisfied often only by resort to or allowance for complex materials. The adjective complex

beckons to the fact that the substructure influences gross mechanical behaviour in a prominent way. Interactions due to substructural changes are represented directly. Examples, just to list a few, are liquid crystals, quasi-periodic alloys, polymeric bodies, spin glasses, magnetostrictive materials and ferroelectrics, suspensions, in particular liquids with gas bubbles, polarizable fluids, etc. Hopefully, substructures can be exploited, even invented anew, to reach predetermined goals.To help in the process, theories must be developed so that severe challenging theoretical problems arise; often of fundamental nature. A precise grasp of the physical meaning of mathematical entities is critical for the correct, adequate proposal of models of behaviour and even of consequent computational analyses. A basic problem is of bridging scales even from atomic to macroscopic level, translating through continuum limit the prominent aspects of the subtle discrete substructural features. Their number and nature may be also enriched by specific circumstances. The collection of chapters composing this book aims to underline some aspects of these questions, proposing also new matter of discussion together with specific solutions.

In **Chapter 1, Pierre Degond derives hydrodynamic models of plasmas and disparate mass binary mixtures by evaluating the continuum limit of kinetic small-scale events represented by means of Fokker-Planck or Boltzmann equations. Macroscopic diffusion equations for density of particles and energy follow, coupled with a Euler-type equation for ions or heavy species. Inconsistencies in existing models are evidenced. **

In **Chapter 2, Carlo Cercignani continues the discussion on how kinetic schemes based on Boltzmann equation may offer microscopic foundations of continuous dynamical models. He examines how old and new techniques in the kinetic theories of dense gases may be useful for describing the fast flow of granular materials. **

Substructural kinetic effects may be not as prominent in some circumstances as quantum phenomena. In **Chapter 3, Jan Jerzy Slawianowski develops a quantization scheme for affine bodies, a special class of complex bodies where the natural morphological descriptor is a second-order tensor: in other words, each material element is considered as a system which may (microscopically) deform independently of the neighbouring fellows. **

Once reasonable models have been established, computational techniques are essential in finding explicit solutions in special cases.When phenomena at various scales are involved, non-trivial computational problems arise and may be tackled with different methods, depending on circumstances. In their **Chapter 4, Sukky Jun andWing Kam Liu discuss computational methods appropriate to analyse the formation of electronic band structures in periodic atomic lattices. The approach makes use of periodic meshless shape functions based on the moving least-square approximation. Wave equations are analysed in the reciprocal space determined by the standard Fourier basis. The analyses of semiconductors, photonic and phononic crystals are natural applications. Complex bodies are produced in non-simple industrial processes so that the process of formation of substructures deserves to be described per se. **

Amid possible industrial processes, in **Chapter 5, Antonio Fasano, Krishna Kannan, Alberto Mancini and Kumbakonam R. Rajagopal propose a new model for the Ziegler-Natta polymerization in a high-pressure reactor by considering, after fragmentation, a single agglomerate of catalytic particles, then analyzing the mechanics of growing nano-spheres.A non-linear hyperbolic system of governing equations arises. **

Other aspects of the mechanics of polymers are further discussed by *Krishna Kannan *and *Kumbakonam R*. *Rajagopal *in their **Chapter 7.The attention is focused on the solidification process of molten polymers where there is competition between the effects of substructural quenching and deformation of the melt:The former effect is an obstacle to the crystallization while the latter enhances it in a way in which memory effects have to be accounted for. Deformation and the corresponding macroscopic stress influence also the formation of nanostructures in semiconductors during their fabrication and the collective mechanical behaviour in applications. The modelling of these effects include atomistic, continuum and multiscale features. These topics are discussed in Chapter 8 by HarleyT. Johnson. **

Finally, our personal contributions are in **Chapters 6 and 9. Basic foundations of the mechanics of bodies in which substructural phenomena have kinetic nature are discussed in Chapter 6 (by G.C.) without resorting to the use of some version of Boltzmann equation. The interaction between gross deformation and spin structures are discussed in Chapter 9 (by P.M.M.) paying attention on the evolution of disclination lines and point defects. The covariance of the relevant evolution equations is proven. **

**Chapter One **

**Asymptotic Continuum Models for Plasmas and Disparate Mass Gaseous Binary Mixtures **

*Pierre Degond * degond@mip.ups-tlse.fr * MIP, UMR 5640 (CNRS-UPS-INSA), Université Paul Sabatier, 118, route de Narbonne, 31062 Toulouse Cedex, France *

We review the derivation of macroscopic models for plasmas and disparate mass binary mixtures from a large-scale limit of the underlying kinetic (Fokker–PlanckorBoltzmann) equations. The Knudsen number (ratio of the collision mean free-path to the size of the system) is supposed to be of the same order of magnitude as the square root of the mass ratio between the particles. The so-obtained macroscopic model consists of a system of two diffusion equations for the electron (or light species) density and energy (often referred to as the energy-transport model), coupled with the gas dynamics Euler system for the ions (or heavy species). This mathematical properties of this system are reviewed and its applicability to various physical contexts is outlined.

**Gas mixture **

**Disparate masses **

**Plasmas **

**Boltzmann equation **

**Fokker–Planck-Landau equation **

**Conservation equations **

**Local thermodynamical equilibrium **

**Energy-transport system **

**Hydrodynamic equations **

**Onsager relation **

**41A60 **

**35Q20 **

**76P05 **

**1.1 Introduction 2 **

**1.2 The Kinetic Model 6 **

**1.3 Moment Method and Conservation Laws for Gas Mixtures: Why it Cannot Apply to Plasmas 11 **

**1.3.1 Properties of the collision operators 11 **

**1.3.2 Moments and conservation laws 13 **

**1.3.3 Closure of the moment system: LTE 14 **

**1.3.4 Why the standard mixture model does not apply to plasmas 17 **

**1.4 The Plasma Fluid Model 10 **

**1.4.1 Energy-transport form of the system 18 **

**1.4.2 Hydrodynamic form of the system 21 **

**1.4.3 Discussion of the plasma fluid model and applications 22 **

**1.4.4 Approximate expression of the diffusion matrices 24 **

**1.5 Scaling Hypotheses 29 **

**1.6 Expansion of the Interspecies Collision Operators 34 **

**1.7 Moment Method and Conservation Laws for Plasmas 40 **

**1.7.1 Properties of the expanded collision operators 40 **

**1.7.2 Moments and conservation laws for the scaled kinetic model 42 **

**1.7.3 Closure of the plasma moment system 44 **

**1.8 Computation of the Fluxes and of the Collision Terms 49 **

**1.8.1 Preliminaries 49 **

**B 50 **

**1.8.3 Resolution of the perturbation equation (1.242) 51 **

**1.8.4 Computation of the fluxes 52 **

**1.8.5 Expression of the fluxes in terms of ne and Te 54 **

**1.8.6 Expression of the collision terms 56 **

**1.8.7 Back to physical variables 57 **

**1.9 Conclusion 57 **

This chapter reviews the derivation of macroscopic models for plasmas and disparate mass binary mixtures from a large-scale limit of the underlying kinetic (Fokker–Planck or Boltzmann) equations. The material is based on a series of work starting with Refs. [**1], [2] for the case of unmagnetized gases and plasmas and complemented by Ref. [3] for the magnetized cases. A related series of works is concerned with semiconductor models [4], [5]. **

The systematic derivation of macroscopic continuum models from microscopic models has been the cornerstone of modern statistical physics since the seminal works of Boltzmann and Maxwell. The mathematical approach based on asymptotic expansions has been proposed and implemented by Hilbert **6] and independently by Chapman [7] and Enskog [8] and allows to rigorously derive Euler and Navier–Stokes equations from the Boltzmann equation. This approach has been later pursued by Grad [9–11] and Cercignani [12]. A recent review on these topics and the latests mathematical developments can be found in Ref. [13] and in the works of Caflisch [14], Kawashima et al. [15] and more recently Bardos et al. [16]. In the same vein, the passage from microscopic models of radiative transfer, neutron transport or semiconductors toward diffusion models (respectively called the Rosseland approximation, the diffusion approximation or the drift-diffusion model) has been investigated in Refs. [17–21]. **

In the present work, we concentrate on plasmas or disparate mass binary mixtures, and stay at the level of the formal asymptotics, which is rich and complex enough to justify a study of its own. Our goal is to study the passage from the kinetic description of the plasma or mixture toward macroscopic equations. It should be noted that there exists no mathematically rigorous justification yet of the formal theory presented below.

The derivation of macroscopic models for plasmas and disparate mass binary mixtures possesses already quite a rich history. Macroscopic transport models for plasmas have been solidly established by Braginskii [**22] while the first numerical computation of the electron thermal diffusivity dates back to the work of Spitzer and Härm [23]. Disparate mass binary mixtures have been considered in Refs. [24–28]. **

In ordinary gas mixtures (when the particles have similar masses),the momenta and energies of each species relax one to each other on the same time scale as relaxation of the distribution functions to local thermodynamical equilibrium (LTE, this scale being characterized by the collision times and mean free-paths). As a consequence, the macroscopic velocity and temperature of the two species are equal (at leading order) and the resulting system contains only four equations: two conservation equations for the densities of each species, but only one momentum equation and one energy equation for the global mixture. If Navier– Stokes corrections are included, specific interspecies diffusion terms are added to the momentum and energy equations to account for the velocity and temperature differences between the two species which are first order in the Knudsen number [**29] (we recall that the Knudsen number is the ratio of the collision mean free-path to the size of the system). **

In Ref. [**22], Braginskii naturally considers two sets of coupled hydrodynamic equations (of Navier–Stokes type) for the electrons and the ions. He therefore assumes that the velocity and temperature differences between the species are zeroth order in the Knudsen number, which implicitly means that momentum and energy relaxations of the species one to each other occurs on a longer time scale than relaxation of the distribution functions to LTE. This obviously requires some additional mechanisms compared with an ordinary mixture. This mechanism is related to the small mass ratio between the electrons and the ions, which results in a scale separation between the relaxations to thermal equilibria of each species, as well as momentum and energy relaxation of the species to each other. **

In the present chapter, we propose a consistent asymptotic analysis of the set of the plasma kinetic equations which takes into account that both the Knudsen number and the mass ratio are simultaneously small. The relevant scaling is indeed that the Knudsen number is of the same order as the square root of the mass ratio.

When the Knudsen number goes to zero in this way, one does not get a system of hydrodynamic equations for the each species, like in Ref. [**22], but a system of diffusion equations for the electron density and energy, coupled with a hydrodynamic equations for the ions. The diffusion system for the electrons (sometimes referred to as the energy-transport system in the literature) can be viewed as a compressible Navier–Stokes system in which all terms proportional to the electron inertia are neglected (i.e. the inertia and viscosity terms in the momentum balance equation, the drift component of the total energy in the energy balance equation, etc.). In other words, the coupled Navier–Stokes system in Ref. [22] retains terms of second order in the Knudsen number, while of course, some other terms of the same order (like second order (Burnett) corrections in the ion balance equations) are dropped out. To some extent, Braginskii’s model (which has moreor less become the conventional plasma models) lacks consistency in this respect. The present work can be viewed as an attempt to correct this inconsistency of Braginskii’s model. **

Braginskii’s model [**22] with Sptizer–Härm [23] thermal conductivity suffers from other deficiencies that shall not be addressed in the present chapter, such as the inadequacy of Sptizer–Härm formula for hot-electron transport (a review about this question can be found in Ref. [30]), the anomalously large value of certain transport coefficients due to plasma turbulence [31], the need for closures in the regimes where collisions are not dominant [32] and more recently [33], etc. In the present work, we do not question the LTE assumption which is the conventional strategy for deriving macroscopic model. Therefore, we cannot address the above-mentioned points, which are obviously situations where the LTE assumption is not valid. The theory of extended thermodynamics [34], recently revisited by Levermore [35] has been designed to address such situations. We shall not develop this viewpoint and just mention the recent work of Anile et al. [36,37] who have practically implemented this theory in the semiconductor context. **

In disparate mass binary mixtures, the derivation of macroscopic models from kinetic ones has been studied by a large number of authors [**24–28] who have proposed a large panel of different models. Our approach bears similarities with that of Petit and Darrozes [28]. **

Although based on earlier articles [**1–3], this presentation is different in several aspects. In the present work, we tried to make the text accessible to a physics audience (while the mathematical formalism of the previous chapters makes them difficult for non-mathematicians). First, we base our discussion on the physical conservation laws (mass, momentum and energy balance equations, entropy inequality), instead of using exclusively the Hilbert method like in the previous works. Second, we express the energy-transport fluxes in terms of the gradients of the entropic variables (chemical potential and reciprocal of the temperature essentially), which makes Onsager’s reciprocity relation particularly simple. The other major additions concern discussions of the physical motivation and implication of the model and the derivation of explicit formulae for the diffusion coefficients and relaxation rates in some simple cases (numerical computations are required for the most general case). **

These models have been implemented in various application cases such as the modeling of a plasma opening switch [**38,39], and the description of the primary discharge development in the process of arcing on satellite solar cells [40] (see also the review [41]). In the second case, surface collision mechanisms were added, following [42].The energy-transport model appears in the solid-state physics literature as early as Refs. [43–45]. It is routinely used in semiconductor modeling [46–50]. The mathematical analysis of the energy-transport model has been conducted in Refs. [51–53]. As previously stated, its derivation from kinetic models has been formally established in Refs. [1–3] in the case of plasmas, and in Refs. [4,5] for the semiconductor case. Elements towards a rigorous convergence proofs are collected in Ref. [54] for the semiconductor case. To our knowledge, there is no mathematically rigorous analysis of the plasma case. Another approach, which consists in deriving the energy-transport model from a relaxation limit of the hydrodynamic equations, is reported in Ref. [55].The energy-transport model can also be derived from a relaxation limit of the so-called spherical harmonics expansion (SHE) model, also called the Fokker–Planck (FP) equation [4,56]. For other references about the SHE/FP model and its numerical simulation, we refer the reader to Refs. [57–65], [84], [85]. A review of the relation between SHE/FP models and energy-transport models for semiconductors can be found in Ref. [66]. **

For general references concerning rarefied gas dynamics, we refer the reader to [**12], [13], [67]; for plasmas, to [68–70]; for semiconductors, to [71], [72]. **

Extensions of the present work have been done in several directions. A derivation of an energy-transport model for relativistic electrons can be found in Ref. [**39].The case where the ionization levels of the ions is large has been investigated in Ref.[3]:it leads to an FP/SHE model for the electrons, coupled with a hydro-dynamic model for the ions. Impact ionization for the case of semiconductors has been taken into account in Refs. [73–75], and leads to non-conventional energy-transport models. A somewhat related study for plasmas has been done in Ref. [76]. Relaxation limits of the energy-transport model under high-field conditions toward non-equilibrium drift-diffusion models have been investigated in Ref. [77]. **

The outline of the chapter is the following: we start with the kinetic level in **Section 1.2 and investigate the moment approach and LTE closure for conventional mixtures in Section 1.3.We show why the case of plasmas or disparate mass mixtures requires a special treatment. Then, we state our plasma fluid model in Section 1.4. In this section, we discuss the physical implications of the model and provide analytical formulae for the diffusion and relaxation coefficients for some specific examples of collision cross-sections. We then develop the arguments leading to the plasma fluid model in the forthcoming sections. **

**Section 1.5 is devoted to the scaling assumptions (which are the key point of our approach) and the subsequent dimensionless form of the equations. The most important dimensionless parameter in this analysis is the square root of the mass ratio ε, which is supposed equal to the Knudsen number (ratio of the collision mean free-path to the macroscopic scale). Section 1.6 gives the expansion of the interspecies collision operators in terms of ε and lists some useful properties of these operators. Then, the moment approach and the LTE closure are reprocessed in the framework of this new scaling. In doing so, the evaluation of the electron fluxes requires the inversion of an auxiliary equation involving the linearized collision operators. This analysis is performed in Section 1.8 and, after scaling the results back to physical variables, leads to the model stated in Section 1.4. The analysis of the operator involved in the auxiliary equation is deferred to Appendix A. **

We investigate a fully ionized plasma model consisting of electrons and one species of ions, and no neutrals. We are interested in the macroscopic effects induced to the mixture by the collisions among the particles, being of the same species or not. In this respect, the specificity of plasmas is that the considered species have very different masses. On the other hand, all the conclusions of this study will be relevant to disparate mass binary gases as well, the nature of the interaction (of Coulombic type in the case of plasmas, and of short-range type for neutral molecules) being of little importance in the conclusions of this study. Therefore, we do not restrict ourselves to the Landau collision operator (relevant for charged particles) but allow ourselves to also consider Boltzmann type operators.

We shall use the index e

for electrons in the plasma case, and for the light species in the gas-dynamic case and the index i

for ions or the heavy species.

We start from kinetic models of Boltzmann or Fokker–Planck type for the two species of particles. *Let fe*(*x,v, t*) denote the electron (or light species) distribution function, which depends on the space coordinate *x *³, the velocity coordinate *v *³ and the time *t > *0. The pair (*x*, *v*) spans the phase-space and the quantity *fe*(*x, v, t*) d*x *d*v *represents the electron number density at a given point (*x*, *v*) of phase space and at time *t. *Similarly, we denote by *f*i(*x*, *v, t*) the ion (or heavy species) distribution function. The functions *f*e and *f *i evolve in time according to the so-called Boltzmann (or Fokker–Planck) equations:

where *m*e and *m*i denote the electron and ion masses, *Q*ee( *f*e, *f*e), *Q*ii( *f*i, *f*i) are the electron–electron and ion–ion collision operators, *Q*ei( *f*e, *f*i), *Q*ie( *f*i, *f*e) are the electron–ion collision operators, respectively, acting on the electron and ion distribution functions and finally *F*e, *F*i are force terms acting on the electrons and ions.

These equations express that the rates of change of *f*e or *f*i while following the particles in their classical mechanical motion (i.e. the left-hand sides) are due to the pair collisions between the particles (expressed by the collision operators on the right-hand sides). The derivation and justification of the Boltzmann equation is beyond the scope of the present work (see e.g. Ref. [**12], [13], [67]) . For rarefied gas mixtures, an example of force acting on the molecules is the gravity and the collision operators are of Boltzmann type (see below). **

In the case of a plasma, the main force acting on the charged particles is the Lorentz force:

where *q*e and *q*i denote the absolute value of the electron and ion charges, *E *= *E*(*x, t*) is the electric field and *B *= *B*(*x, t*)*, *the magnetic field. A specificity of plasmas is that the two terms of the Lorentz force may have different orders of magnitude. This kind of splitting does not occur (at least to our knowledge) in gas mixtures.

In plasmas, the collision operators are of Fokker–Planck–Landau type (see below), and we shall distinguish them from Boltzmann type operators by using the superscript P for the Fokker–Planck–Landau operator, and B for the Boltzmann operator. Therefore, the case of rarefied gas mixtures is described by equations **(1.1) and (1.2) where Q= QB while, in the case of plasmas these equations take the form: **

Throughout the text, all statements valid for both *Q*B and *Q*P will be stated using the generic notation *Q*.

In plasmas, the collisions are modeled by the Fokker–Planck–Landau operator. This operator is written, for collisions of the species *α *against the species *β*, *α *and *β *being equal to either e

or i

:

where *μαβ *= *mαmβ/*(*mα *+ *mβ*) is the reduced mass of the pair of particles, *S*(*v*) = Id — (*v *⊗ *v*)*/|v|*² is the 3×3 projection matrix onto the plane orthogonal to *v *and *σαβ*p is the scattering cross-section. As usual in kinetic theory, we have introduced the short-cuts *f ≡ f*(*v*)*,f*1 *≡ f *(*v*1) and for instance (∇*vfβ*)1 ≡ (∇*vfβ*)(*v*1)*. *It is customary to make *σαβ*p depend on the relative velocity of the colliding particles *|v − v*1|. In fact, for our purpose, we shall prefer to make it depend on the relative energy *μαβ|v − v*1*|*² i.e. *σαβ*P = *σαβ*P(*μαβ*|*v* − *v*1|²) . We have *σαβ*p = *σαβ*p*. *

In rarefied gas mixtures, the collision operator is the Boltzmann operator, given by:

where again *fα* ≡ *fα*(*v*), (*fβ*)1 ≡ *fβ*(*v*1), *fα*′ ≡ *fα*(*v*′), (*fβ*)1′ ≡ *fβ*(*vα*′)*. *The pair (*v*′, *vα*′) is related with (*v, v*1) and *co *through the collision transform

where *v*G is the center-of-mass velocity

Introducing the unit vector

*v *and *v*1 can be written

We also note that

Therefore, *e *and *ω, *respectively, define the direction of the relative velocities before and after the collision in the center-of-mass frame. *ω *= {*ω *³, |*ω*| = 1}.

Relation **(1.7) can be deduced from momentum and energy conservation during the collision y( v, v1) ↔ (v′, v1′), i.e.: **

Indeed, momentum conservation implies that the center-of-mass velocity *vG *is conserved during the collision. Then, from energy conservation, we deduce that the magnitude of the relative velocities *|v *− *v*1*| *and |*v*′ − *v*1′| are unchanged. Only the direction of the relative velocity changes from *e *to a unit vector *ω *randomly distributed on the unit sphere. Formulas **(1.7) and (1.8) then just express the reconstruction of the velocities from the center-of-mass velocity and the relative velocities. **

The scattering angle *θ *= cos− 1 (*ω *· *e*) is the angle between the directions of the relative velocities. *θ *ranges in [0, *π*]*. *The plane defined by the two vectors (*e*, *ω*) is the collision plane. All planes containing the relative velocity *v − v*1 are possible collision planes. All these planes can be located with respect to a reference plane by using a single angle *φ *∈ [0, 2π]*. *The pair (*θ, φ*) defines a spherical coordinate system for the vector *ω *and we have *dω *= *dθ *sin *θ dφ. *

The energy in the center-of-mass frame (or relative energy) is (up to the factor 2):

Like in the case of the Fokker–Planck–Landau operator, we shall make the differential scattering cross-section *σαβ*B depend on the relative energy rather than on the relative velocity |*v − v*1*| *itself. *σαβ*B also depends on the scattering angle cos*θ*, i.e. in general,

The *θ *dependence of *σαβ*B describes the relative probability that a collision with a definite scattering angle occurs. It depends on the nature of the collision mechanism. Of course, we have *σαβ*B = *σβα*B

For hard sphere collisions, i.e. if the molecules can be modeled like billiard balls of radii *rα *and *rβ, *respectively, elastically bouncing on each other, we have

It is independent of |*v − v*1| and *θ. *For rarefied gas dynamics, a better model is obtained by letting *rα *+ *rβ *depend on |*v − v*1| in a specific fashion. This is called the variable hard sphere model (VHS). For electron-neutral or ion-neutral collisions, isotropic scattering cross sections are often considered. This also gives rise to scattering cross sections which only depend on the energy.

We now review the relation between the Boltzmann and the Fokker–Planck operators and in particular, between the scattering

Close Dialog## Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

Loading