1: % This is the response to REFEREE ------------end of LEATEX FILE---
2: % To: prltex@aps.org
3: %
4: % Dear Sirs,
5: %
6: % Enclosed please find a REVTEX file of a manuscript entitled
7: %New features of dislocation structures arising
8: %from lattice discreteness.
9: % Oleg N. Mryasov, Yu. N. Gornostyrev and A.J. Freeman
10: % which we are submitting for publication in Physical Review Letters .
11: % The figures to the manuscript are submitted in separate messages
12: % following this one with subject lines:
13: % fig1ab.ps
14: % fig1cd.ps
15: % fig2a.ps
16: % fig2b.ps
17: % fig2c.ps
18: % !!!!!!!!!!!!!!!!!!!!!!!
19: % Dr.~Oleg ~N.~Mryasov \\
20: %Deptartment of Materials Science \\
21: %University of California \\
22: %577 Evans Hall, R.508-14 \\
23: %Berkeley, CA 94720-1760 \\
24: %fax: 510-643-6629, ph: 510-642-8484 \\
25: %e-mail : onm@smasher.mse.berkely.edu \\
26: %------------------------------------------------cut here -------------------
27: \documentstyle[preprint,aps,epsfig]{revtex}
28: %\documentstyle[12pt,prl,aps,epsfig]{revtex}
29: %\documentstyle[prl,aps,multicol,epsfig]{revtex}
30: \begin{document}
31: \newcommand{\mex}[1]{$ \langle #1 \rangle $ }
32: %%%0%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
33: % for non preprint uncomment the following and ...
34: %\twocolumn[\hsize\textwidth\columnwidth\hsize\csname
35: %@twocolumnfalse\endcsname
36: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
37: %\draft
38:
39: \title{
40: New features of dislocation structures arising
41: from lattice discreteness.
42: }
43: \author{
44: Oleg N. Mryasov$^{1,2}$, Yu. N. Gornostyrev$^{2,3}$ and A.J. Freeman$^{2}$
45: }
46: \address{
47: \(^1\) Department of Materials Science, University of California,
48: Berkeley, CA 94720-1760. \\
49: \(^2\) Department
50: of Physics and Astronomy, Northwestern University Evanston, IL
51: 60208-3112. \\
52: \(^3\) Institute of Metal Physics, Ekaterinburg, Russia.
53: }
54: \maketitle
55:
56:
57:
58:
59:
60:
61:
62: \begin{abstract}
63: New aspects of a relation between lattice and dislocation structures are
64: examined
65: within a physically transparent theoretical scheme.
66: Predicted features originating from the lattice discreteness include:
67: (i) multiple core dislocation structures and
68: (ii) their dependence on the position of the dislocation axis.
69: These effects, which in principle can be observed directly and may also
70: manifest themselves in dislocation motion or/and transformation (cross-slip)
71: characteristics, are very general and present in any crystal
72: in which they may be more or less pronounced depending on the material.
73:
74: \end{abstract}
75:
76:
77: \pacs{
78: 61.72.Lk, % Linear Defects: dislocations and disclinations
79: 61.72.Bb, % Theory and models of crystal defects
80: 61.72.Nn, % Stacking faults and other planar faults
81: 61.82.Bg, % Metals and alloys
82: 62.20.Fe % Deformation and plasticity
83: }
84: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
85: % ... and also the following
86: %\vskip2pc]
87: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
88:
89: It is widely accepted that such defects as dislocations
90: significantly influence a number of properties in real materials.
91: Thus, understanding the relationship between lattice and dislocation
92: structures
93: is one of the fundamental problems of materials physics.
94: Despite recent developments of powerful atomistic simulation techniques,
95: up to now
96: our understanding of
97: the relation between lattice
98: and dislocation structures
99: is based on
100: results obtained within the framework of the
101: Peierls-Nabarro (PN) model\cite{Peierls,HirthLote}.
102: This model has provided both language for interpretation
103: experimental/theoretical results and simple relations between
104: dislocation properties and lattice discreteness characteristics
105: (periodicity, symmetry, etc.)
106: \cite{Peierls,HirthLote,nabb_50y,duePRL}.
107: %duePRB,
108: %ourPN1,prb2}.
109: This remarkable breakthrough in understanding how dislocation properties
110: are related with lattice characteristics
111: became possible due to two features of this model:
112: (i) its high tractability and
113: (ii) its combined different length scale descriptions.
114:
115: The combined "continuum/atomistic" descriptions in the PN model
116: follow clearly from the structure of the energy functionals \cite{schoeck:94}, $E_{tot}$, of the
117: dislocation displacement distribution $\vec{u}(x)$
118: %\cite{note1}
119: (here $x$ is a distance from the dislocation axis in the slip plane)
120: \begin{eqnarray}
121: E_{tot}(\vec{u}(x))= E_{el} (\vec{u}(x)) + E_{mis} (\vec{u}(x))
122: \label{funct}
123: \end{eqnarray}
124: with a linear elastic ($E_{el}$ ) and a non-linear atomistic
125: misfit energy term ($E_{mis}$)
126: \begin{eqnarray}
127: E _{mis}= h \sum \limits _n \Phi (\vec u (nh-l)) ,
128: \label{eq:misf}
129: \end{eqnarray}
130: where $\Phi (\vec u (x))$ is a periodic energy profile
131: which is often approximated by
132: the so-called generalized
133: stacking fault energy (GSF) or $\gamma$-surface \cite{vitekGSF}.
134: Indeed, the $E_{mis}$ term represents the most apparent and important
135: lattice properties
136: - discreteness/periodicity/symmetry - and allows one to investigate
137: within the PN model the relation between lattice and dislocation properties.
138: $E_{mis}$
139: %energy Eq.\ref{eq:misf}
140: can be expanded in a Fourier series as
141: \begin{eqnarray}
142: E _{mis} = E_{mis}^0 + h \sum \limits _{s=1}^{s=\infty} J_s \cos \frac
143: {2\pi sl}{h}
144: \label{eq:misf_ft}
145: \end{eqnarray}
146: where $E_{mis}^0$ $= \int \limits _{-\infty}^{\infty}\Phi (\vec u (x))dx$
147: is independent of the position of the dislocation axis $l$
148: and terms which are oscillatory with $h$ being a repeat distance normal
149: to the dislocation line
150: and
151: $n$ an integer number that counts atomic rows in the same direction.
152:
153: Now, a minimization of ${E_{tot}(\vec{u}(x))}$ allows one
154: to find the equilibrium dislocation structure, $\vec{u}(x)$.
155: In order to perform this minimization in analytic form,
156: a critical approximation $E _{mis} = E_{mis}^{0}$ has been made
157: \cite{Peierls}.
158: This "continuum" approximation in representing the misfit energy - which
159: is supposed to describe lattice discreteness - results in an obvious
160: inconsistency
161: which has been a subject of debate for years,
162: mainly in the context of Peierls stress determinations
163: \cite{nabb_50y,bulatov_l,ohsawa1}.
164: Thus, an interesting and fundamental issue arises - if
165: consistently represented in the atomistic interaction energy, how will
166: lattice discreteness be manifested
167: in the structure of dislocations ?
168:
169: Several recent attempts to overcome this inconsistency
170: resulted in
171: purely numerical procedures \cite{bulatov_l,ohsawa1}. Thus, one of the most
172: advantageous features of the PN model -
173: high tractability and transparency - has been sacrificed.
174: These authors also focused on the Peierls stress determination and
175: demonstrated that
176: indeed the discrete representation of the misfit energy
177: brings theoretical estimates much closer to experimental results.
178:
179: In this Letter, using
180: a physically transparent solution of the PN model with
181: a consistent discrete representation for
182: the misfit energy, we
183: examine how lattice discreteness
184: may influence dislocation structure.
185: This allows us to predict
186: new generic features
187: of the dislocation structure that are independent of the PN model assumptions,
188: and driven by lattice discreteness
189: such as multiple core structures
190: and their dependence on the position of the dislocation axis.
191:
192: To determine dislocation structure,
193: we perform a minimization of the total energy functional,
194: Eq. \ref{funct}, with a discrete representation of the
195: misfit energy, Eq.\ref{eq:misf}, using trial functions, $\vec{u}(x)$, defined
196: from the Laurent expansion \cite{Lejcek,prb2} of their derivatives
197: $\rho _{\beta}(x)$
198: \begin{eqnarray}
199: \rho _{\beta}(x)=\frac{du_{\beta}(x)}{dx} = Re \sum_{k=1}^N
200: \sum_{n=1}^{p_k} \frac{A_{nk}^{\beta}}
201: {(x-z_k^{\beta})^n} ,
202: \label{RhoEqn}
203: \end{eqnarray}
204: where
205: N is the maximal number, $p_{k}$ is the maximal
206: order of the poles $z_{k}^{\beta}$ and
207: $A_{nk}^{\beta}$ are expansion coefficients.
208: It is important to note that, by definition, these trial functions
209: provide a minimum of $E_{tot}$
210: for an arbitrary $\Phi(u)$ potential (not only sinusoidal,
211: as trial functions used in \cite{schoeck:94} to parameterize
212: total energy functional in the convenient form)
213: in case of the "continuum" approximation,
214: $E _{mis} = E_{mis}^{0}$ \cite{Lejcek,prb2}.
215: This choice of trial functions not only provides good accuracy and
216: stability of the minimization procedure \cite{prb2} but also allows
217: one to express $E_{tot}$ through
218: parameters describing the dislocation structure.
219: Indeed, the poles $z_k^{\beta}=x_k^{\beta}+i\omega_k^{\beta}$
220: have a clear meaning:$x_k^{\beta}$ gives the position and
221: $\omega_k^{\beta}$ gives the width of the
222: partials for the screw ($\beta$=1) and edge ($\beta$=2) components of the
223: displacement in the partial cores. For example, for the ordinary
224: dislocations
225: dissociated into two Shockley partials,
226: $x_k^{\beta} = l \pm d^{\beta}/2$, where d is the partials separation and $l$
227: gives the position of the whole ordinary dislocation center.
228: In the general case with these trial functions,
229: %both $E_{el}$ and $E_{mis}$ terms
230: $E_{tot}$
231: can be presented
232: as a numerical function of geometrical parameters,
233: in particular, for an ordinary dislocation
234: %$E_{tot}$ can be represented
235: as a function of the set of parameters ($\{{\bf g
236: }\}=\{d, \omega, l\}$) describing the dislocation's structure ($d, \omega$)
237: and its position in the lattice ($l$).
238:
239: As examples, we consider ordinary dislocations for fcc metal, Ir,
240: and an ordered alloy,
241: CuAu, with L1$_{0}$ structure. In these materials, this type of dislocation
242: normally splits into two Shockley partials \cite{HirthLote,prb2}
243: and represents a very typical example of dislocation structures.
244: To illustrate graphically minimization of $E_{tot}$,
245: let us introduce elastic ($F_{el}(d)$) and misfit ($F_{mis}(d)$)
246: generalized forces which are defined
247: as, \hspace{0.2cm}
248: %F_{mis}(d)= \frac{\partial E_{mis}(\{\bf g \})}{\partial d} ,
249: %\label{genf_def}
250: %\end{eqnarray}
251: %derivatives
252: $F_{el}(d)= - \partial E_{el}(d)/\partial d$
253: (the sign is chosen for convenience) and
254: $F_{mis}(d)= \partial E_{mis}(d)/\partial d$
255: (here and further we drop the $\beta$ index since $d^{1} \approx d^{2}$).
256: In this definition, for a given $d$
257: other geometrical parameters from
258: the complete set $\{\bf g\}$ are taken
259: to be such that they minimize $E_{tot}$.
260: Obviously, in this case the intersection of $F_{el}(d)$ and $F_{mis}(d)$
261: gives a partial
262: separation $d$ which corresponds to the minimum of Eq.\ref{funct},
263: provided that
264: the second derivatives are positive.
265:
266: The generalized forces calculated according to this definition
267: using ab-initio $\gamma$-surfaces (see \cite{prb2} for
268: details)
269: in the case of the screw orientation
270: of the unit dislocation for Ir and CuAu are presented in
271: Fig.\ref{fig:forces} (a,c).
272: For comparison, we also determine
273: generalized forces for dislocations
274: with simple model density distribution displacements composed of two
275: delta functions,
276: $\rho (x) = b_{1} \delta(x + d/2) + b_{2}\delta(x - d/2)$
277: %\cite{note2}
278: (see Fig. \ref{fig:forces} (b, d)).
279: In this case, we have a step function shaped dislocation for which
280: $E_{tot}$
281: has a very simple form,
282: \begin{eqnarray}
283: E_{tot}=H \cdot ln(\frac{1}{d}) + \gamma_{isf} \cdot d
284: \label{sing_funct}
285: \end{eqnarray}
286: and corresponding generalized forces
287: $F_{el}^{step}= H /d $ and $F_{mis}^{step} = \gamma_{isf} $, where
288: $ \gamma_{isf} $ is the intrinsic stacking fault energy and H is
289: a so-called prelogarithmic factor (see for example \cite{schoeck:94}).
290: Interestingly enough, for this model type of dislocation
291: the well-known simple relation between the equilibrium partials
292: separation and
293: stacking fault energy, $d=H/\gamma_{isf}$ \cite{HirthLote}, can be easily
294: recovered from
295: the functional dependence in Eq. \ref{sing_funct}.
296: As can be seen in Fig.\ref{funct},
297: in the limit of large separation distances ($d>>\omega$),
298: the PN model generalized forces defined within the "continuum" approximation
299: approach
300: %asymptotically
301: those for the step function shaped dislocation.
302:
303: We now focus on how the oscillatory part of the misfit energy, usually
304: neglected in the
305: PN model analysis, affects dislocation structure.
306: Remarkably enough, there are not only oscillations with
307: $l$ which can be expected from the Fourier expansion of the misfit energy,
308: Eq. \ref{eq:misf_ft}, but also oscillations with
309: partials separation $d$ for fixed $l$.
310: As can be seen in Fig. \ref{fig:forces},
311: for Ir, despite the rather small amplitude of these oscillations,
312: the effect of the misfit energy discrete representation is quite visible since
313: the intersection of the generalized forces happens to be in the area
314: where $d$ dependence of the F$_{el}$ is rather weak and so
315: solutions are affected most by the F$_{mis}$ oscillations.
316: Finally, for CuAu the effect is dramatic since the amplitude
317: of the oscillations is very large.
318: It is significant that the discrete representation of the misfit energy not
319: only
320: changes quantitatively parameters of the dislocation structure (as in the
321: case of Ir)
322: but may also result in qualitatively new effects (as in the case of CuAu).
323: As is evident from Fig.\ref{fig:forces}, one such general effect, which is
324: independent of the PN model approximations,
325: %\cite{note_pn_limit}
326: is the appearance of
327: multiple stable dislocation core configurations.
328: Indeed, conclusions we draw in this study about the possibility of multiple
329: core
330: configurations are based on generic features of the $F_{el}(d)$ and
331: $F_{mis}(d)$
332: dependencies (see Fig.\ref{fig:forces}) which follow from the general physics
333: of the linear (elastic part) and non-linear (misfit part) lattice response
334: \cite{note_pn_limit}.
335: % and so reflect essential mechanisms determining dislocation structure.
336:
337: Moreover, within the proposed scheme it is possible to
338: derive\cite{note_rig_derr}
339: the following convenient and physically transparent form for the energy
340: functional, Eq.\ref{funct},
341: \begin{eqnarray}
342: E_{tot}= E_{tot}^{0}(d,\omega) +
343: A(d,\omega) \cos \frac {2\pi l}{h} \cos \frac {\pi d}{h}
344: \label{tot_nform}
345: \end{eqnarray}
346: Here the $l$ independent first term $E_{tot}^{0} = E_{el} + E_{mis}^0 $
347: is the energy in the ``continuum'' approximation and the second term has an
348: explicit
349: oscillatory dependence both on $l$ and $d$.
350:
351: This form reveals that
352: the appearence of
353: %the dislocation structure (including
354: the above features associated with lattice discreteness
355: are dependent on the relative contribution of the energies represented by the $E_{tot}^{0}$
356: and the oscillatory terms.
357: In turn, the character of $E_{tot}^{0}(d)$ dependence and corresponding
358: generalized forces
359: %is determined
360: %by the competition of the
361: %partials
362: %attraction ($E_{mis}^0$) which
363: is driven by the competition of the partials attraction described by $E_{mis}^0$
364: ( term
365: which is dependent upon
366: the $\gamma$-surface energetic characteristics and
367: for large $d$, it can be well approximated by Eq. \ref {sing_funct})
368: %$\gamma_{isf}$)
369: and the
370: elastic repulsion ($E_{el}$ which is
371: dependent on the elastic constants and in the limit $d>>\omega$, has
372: a simple dependence, see Fig.\ref{fig:forces}).
373: The influence of the oscillatory term is predetermined by its amplitude
374: $A(d,\omega)$ which according to our analysis
375: is strongly dependent on
376: %the geometrical and energetic
377: characteristics of the $\gamma$-surface \cite{note_amp}.
378:
379: It is important that Eq. \ref{tot_nform} describes dislocation energetics
380: for a wide range of $d$ and $\omega$ \cite{note_schoeck_func}
381: and correspondingly a complex interdependence of all geometrical
382: parameters ($d$, $\omega$ and $l$).
383: Features of dislocation structure originating in this interdependence of
384: geometrical parameters
385: and their impact on dislocation energetics can be seen from the calculated
386: $d$ and $l$ dependencies of the $E_{tot}$ in Fig.\ref{fig:compl}.
387: Indeed, proof that there are can be more than one stable dislocation core
388: configuration (CuAu)
389: can be seen in Fig.\ref{fig:compl}(a). Next, the
390: dependence of the partials separation on the position of the dislocation
391: axis in the lattice is clearly seen in Fig.\ref{fig:compl}(b).
392: A comparison of the $d(l)$ dependencies determined within the
393: ``continuum'' approximation and with a discrete representation for $E_{mis}$
394: makes it evident that lattice discreteness is the origin
395: of the variation in equilibrium dislocation structure depending on
396: the position of the dislocation axis (core ``relaxation'').
397: This variation may result in
398: changes in the number of stable core configurations
399: and abrupt transitions between them
400: (as for CuAu, see Fig.\ref{fig:compl}(a,b)).
401:
402:
403: These predicted features may have a profound impact
404: on dislocation energetics.
405: As can be seen in Fig.\ref{fig:compl}(c), the dependence
406: of the dislocation structure on the position of the dislocation axis
407: may not only lower significantly the Peierls barrier (as also have been
408: found in \cite{bulatov_l,ohsawa1})
409: but may even modify the shape of the Peierls potential.
410: It is remarkable that core ``relaxation'' along with
411: the existence of the multiple core configurations (the case of CuAu)
412: adds a new feature to the Peierls potential - an additional minimum -
413: which according to the model analysis
414: may result in characteristic changes of the
415: temperature dependence of the yield stress\cite{takeuchi_doub}.
416: As can be seen clearly in Fig.\ref{fig:compl}, the unusual
417: shape of the Peierls potential in CuAu
418: originates in the abrupt transitions between
419: two stable core configurations, ``1'' and ``2''.
420:
421:
422: We find that in addition to the known dislocation
423: structure features,
424: lattice discreteness is
425: the origin of (i) multiple core configurations and (ii) their dependence
426: on the position of the dislocation axis.
427: Combination of these effects
428: may result in rather complex variations of the dislocation structure
429: over the crystal
430: including
431: changes in the number of stable core configurations and transitions between them.
432: As a result, one may have to consider a distribution of the core
433: configurations
434: in a crystal under ambient conditions rather than one characteristic
435: core structure which determines dislocation motion or cross-slip
436: properties.
437: %All this influences energetic characteristics (e.g., the Peierls barrier).
438: %which may in turn affect significantly dislocation motion and
439: %related characteristics of mechanical behavior.
440:
441: As follows from our analysis, predicted features of dislocation structure
442: originating
443: in lattice discreteness are always present in crystals, but as we demonstrate
444: with Ir and CuAu as examples, they appear more or less pronounced
445: depending on characteristics of the given material.
446: These fundamental characteristics can be identified within the proposed
447: theoretical analysis primarily due to its tractability and physical
448: transparency \cite{note_amp}.
449: While these features, namely multiple core configurations,
450: can be directly verified, in principle, in high resolution electron
451: microscopy experiments, they
452: may also reveal themselves indirectly in low temperature internal friction
453: experiments \cite{intf:exp} and in mechanical properties which depend
454: on elementary processes
455: that are sensitive to the dislocation structure. Among such processes, we
456: would emphasize
457: cross-slip, where dislocation core structure and its changes under local
458: stress may play an important role.
459:
460: Work at Northwestern University was supported by the Air Force Office of
461: Scientific Research
462: (Grant No. F49620-95-1-0189)
463: and at UCB by the Office of Basic Energy Science, Division of
464: Materials Science, of the U.S. Department of Energy under contract No.
465: DE-AC04-94AL85000.
466:
467:
468: %\bibliography{pn_let,my-status,intr-inmet,flmto}
469: %\bibliographystyle{prsty}
470:
471: \begin{thebibliography}{10}
472: %\bibitem[*]{*}
473: %Present Address: Department of Materials Science,
474: %University of California,577 Evans Hall, Berkeley, CA 94720;
475: %e-mail:onm@barguzin.mse.berkeley.edu
476: \bibitem{Peierls}
477: R. Peierls, Proc. Phys. Soc. London {\bf 52}, 34 (1940);
478: %
479: %\bibitem{Nabarro}
480: F.~R.~N. Nabarro, Adv. Phys. {\bf 1}, 271 (1952).
481:
482: \bibitem{HirthLote}
483: J.Hirth and J.Lote, {\em Theory of dislocation.}, {\em McGraw-Hill Series in
484: Material Science and Engineering} (McGraw-Hill Book Company, New York, 1968).
485:
486: \bibitem{nabb_50y}
487: F.~R.~N. Nabarro, Mat., Sci., Eng., A. {\bf 234}, 67 (1997).
488:
489: % Nabarro FRN
490: % Theoretical and experimental estimates of the Peierls stress
491: % PHILOS MAG A 75: (3) 703-711 MAR 1997
492:
493: \bibitem{duePRL}
494: B. Jo\'se and M.~S. Duesbery, Phys.Rev.Lett {\bf 78}, 266 (1997).
495:
496: \bibitem{schoeck:94}
497: G. Schoeck, Philos. Mag. A {\bf 69}, 1085 (1994).
498:
499:
500: %\bibitem{duePRB}
501: %Q.~R. B.~Jo\'se and M.~S. Duesbery, Phys.Rev.B. {\bf 50}, 5890 (1994).
502:
503: %\bibitem{ourPN1}
504: %N.~I. Medvedeva, O.~N. Mryasov, Y.~N. Gornostyrev, D.~L. Novikov, and A.~J.
505: % Freeman, Phys. Rev. B. {\bf 54}, 13506 (1996).
506:
507: \bibitem{vitekGSF}
508: V. Vitek, Cryst.Latt.Def. {\bf 5}, 1 (1974).
509:
510: \bibitem{bulatov_l}
511: V.~V. Bulatov and E. Kaxiras, Phys.Rev.Lett. {\bf 78}, 4221 (1997).
512:
513: \bibitem{ohsawa1}
514: K. Ohsawa, E. Kuramoto, and T. Suzuki, Mat. Sci. Eng. {\bf A234}, 302 (1997).
515:
516:
517: \bibitem{Lejcek}
518: D. Lejcek, Czech. J. Phys. {\bf B26}, 294 (1976).
519:
520: \bibitem{prb2}
521: O.~N. Mryasov, Y.~N. Gornostyrev, and A.~J. Freeman, Phys. Rev. B {\bf 58}, 11
522: 927 (1998).
523:
524:
525: \bibitem{note_pn_limit}
526: For drawn here conclusions about features of dislocation structures
527: originating in lattice discreteness
528: only the character of the $F_{el} (d)$ and $F_{mis}(d)$ is important.
529: The oscillatory $F_{mis}(d)$ dependence originates in lattice discreteness/periodicity.
530: Characteristic $F_{el} (d)$ dependence originates from the
531: long wave length limit behavior of the
532: linear lattice response Green function.
533:
534: \bibitem{note_rig_derr}
535: This form can be derived if (i) summation in Eq. \ref{eq:misf_ft} is
536: limited to the s=1 oscillatory term
537: (a well justified approximation, since
538: this series is very rapidly converging ($J_{2}/J_{1} \ll 1 $));
539: (ii) a dislocation density distribution function
540: represented as a sum of the partial dislocation densities,
541: $\rho(x) = \rho^{1}(x+d) + \rho^{2}(x-d)$; and (iii)
542: parity properties - positive for $\rho(x) $ and negative for the
543: $\partial \Phi ( u (x))/ \partial u$ functions are taken into account.
544:
545: \bibitem{note_amp}
546: The estimate (based on the saddle point method)
547: allows to find that
548: $A(d,\omega) \sim \exp (- \frac {2 \pi \omega }{h}) b_{k}H_{k}$, where $H_{k}$
549: are the main values of the
550: $\gamma$-surface
551: curvature tensor at the unstable stacking fault point (H).
552: We also find that $\omega$ is the inverse proportional to the $H$, thus
553: $A(d,\omega$) is
554: strongly dependent on the geometrical ($H$) and energetic
555: ($\gamma_{isf}$) characteristics
556: of the $\gamma$-surface.
557: Details of the analysis of the relation between dislocation structure
558: and $\gamma$-surface geometry/energetics will be published elsewhere.
559:
560: \bibitem{note_schoeck_func}
561: In the limit $d>>\omega$, $A(d,\omega) \approx A$ and derived in this paper
562: expression Eq.\ref{tot_nform} can be transformed to the one which similar
563: to suggested by Schoeck \cite{schoeck_oss94}.
564:
565: \bibitem{schoeck_oss94}
566: G. Schoeck and W. Puschl, Mat. Sci. and Eng. A {\bf 189}, 61 (1994).
567:
568: \bibitem{takeuchi_doub}
569: T. Suzuki, S. Takeuchi and H. Yoshinaga,
570: {\em Dislocation dynamics and plasticity.}, {\em
571: Springer series in materials science ; v. 12. }
572: (Springer-Verlag, Berlin, New York, 1991).
573:
574: \bibitem{intf:exp}
575: I.~G. Ritchie and G. Fantozzi, in {\em Dislocations in Solids.}, edited by
576: F.~R.~N. Nabarro (Elsevier Science Publishers, New York, 1992), p.\ 57.
577:
578: \bibitem{friedel:coll}
579: J. Friedel, in {\em Dislocations in Solids.}, edited by F.~R.~N. Nabarro
580: (Noth-Hoolland Publishing Company, Amsterdam, 1979), p.\ 57.
581:
582: \end{thebibliography}
583:
584:
585:
586:
587:
588: \begin{figure}[t]
589: \caption{
590: Dependence of the generalized elastic and misfit forces (in $J/m^{2}$) on
591: partials separation $d$
592: (in lattice constant units) for
593: Ir
594: and CuAu
595: calculated within the PN model (left panel (a), (c)) and
596: for simple model step function shaped ordinary dislocation
597: (right panels (b), (d)).
598: The misfit forces corresponding to the consistent discrete representation
599: of the misfit energy in the PN model are presented by solid lines and those
600: calculated within
601: the "continuum" approximation
602: by dotted-dashed lines.
603: }
604: \label{fig:forces}
605: \end{figure}
606:
607:
608: \begin{figure}[t]
609: \caption{
610: Dislocation energy (in $J/m$) as (a) a function of the partials separation
611: $d$,
612: (b) a corresponding dependence of
613: $d$ on $l$ (for more than one stable core configuration,
614: solutions which are close in energy are numbered ``1''
615: and ``2'') and (c) a corresponding dependence of
616: the total energy on the position of the ordinary dislocation
617: center given by $l$ in units of a repeat distance in the direction
618: normal to the dislocation line ($h$),
619: calculated for Ir (left panels) and CuAu (right panels);
620: the ones determined within the "continuum" approximation (see text) are
621: presented by the dotted-dashed lines.
622: }
623: \label{fig:compl}
624: \end{figure}
625: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
626:
627:
628:
629: \end{document}
630:
631:
632: