• Aucun résultat trouvé

Electron microscopy diffraction contrast in quasicrystals: some rules

N/A
N/A
Protected

Academic year: 2021

Partager "Electron microscopy diffraction contrast in quasicrystals: some rules"

Copied!
8
0
0

Texte intégral

(1)

HAL Id: jpa-00246517

https://hal.archives-ouvertes.fr/jpa-00246517

Submitted on 1 Jan 1992

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Electron microscopy diffraction contrast in quasicrystals:

some rules

Maurice Kléman, Wolfgang Staiger, Dapeng Yu

To cite this version:

Maurice Kléman, Wolfgang Staiger, Dapeng Yu. Electron microscopy diffraction contrast in quasicrys- tals: some rules. Journal de Physique I, EDP Sciences, 1992, 2 (5), pp.537-543. �10.1051/jp1:1992165�.

�jpa-00246517�

(2)

Classification

Physics

Abstracts

61.42 61.70J

Short Communication

Electron ndcroscopy diffraction contrast in quasicrystals:

some

rules

Maurice

K14man, Wolfgang Staiger

and

Dapeng

Yu

Universit4 de

Paris-Sud,

Centre

d'orsay,

Laboratoire de

Physique

des

Solides,

B£timent s10, 9140s

Orsay

Cedex, France

(Received

10 October 1991, revised 27

January1992, accepted

19

February 1992)

Abstract A number of authors have observed dislocations in icosahedral

quasicrystals

and advanced

some rules to the effect of

measuring

their

Burgers

vectors, on the basis of the two-beam method in the kinematical

approximation.

We first

shortly

review their

results,

which state that the determination of the direction of the

Burgers

vector b

=

bjj+b

i,

(bjj

in ll~jj, bi in lI~

i),

which has six components, can be

suitably simplified

if the five necessary extinctions fall into two classes, viz. four so-called

'strong

extinction conditions'

(SEC) G( bjj

= G

[

b i=o, and one so-called'weak extinction condition'

(WEC) G(j bjj+G[ bi

= 0. We then state the

relationships

between the

bjj

and the bi components which result from these conditions.

We show that the modulus and

sign

can be determined without

ambiguity by examining

two incommensurate

non-extinguishing

diffractions

(G"

b

= n

#

o, n

integer).

Any

'perfect'

dislocation

obeys

at least

approximately

the rules which follow from this

an4lysis.

Dislocations which fail to

obey

them can be

'imperfect'

dislocations

(we

define

them),

whose existence should not be discarded.

1 Introduction.

An

important question concerning quasi-periodic crystals

with icosahedral

symmetry,

since their

discovery

in Al-Mn

alloy [I]

is how far and

why

do

they

differ from the

perfect quasicrys-

talline order. The nature of the

imperfections,

their

origin

and their role in

plastic

deformation

are far from

being

understood. Dislocations and

phason

defects

(a special type

of defects in

aperiodic crystals)

have been

recognized

as

specific

defects

by

Socolar et al. [2] on the basis of

a

density

wave

description

or a unit cell

picture.

Dislocations have been first classified

by

the

topological

method

[3, 4].

A

generalized

Volterra process has also been

employed [5, 6], starting

from a dislocation built in a

hypercubic crystal lZ~ belonging

to a 6-dimensional space

IE~,

and

projecting

it in the

physical

space

IPjj.

It is this method we use in the

sequel:

the

Burgers

(3)

538 JOURNAL DE

PHYSIQUE

I N°5

vector in

IE~,

viz. b

=

bjj

+

bi,

has

components

in both

physical (bjj

e

IPjj)

and

complemen-

tary

(bi

e

IPi)

spaces, hence 6

parameters

are needed to index it. As in usual

crystalline

materials,

the

bjj components

are a measure of

plastic deformation;

the

bi components

are

a measure of the

matching

rule violations

(the phason defects)

which

necessarily

accompany

the dislocation line.

Experimentally,

dislocations have been observed in HREM

images,

either

directly

as in icosahedral

quasicrystals

of Al-Mn-Si

by Hiraga

et al.

[7],

or

by

means of

image processing

as in icosahedral Al-Mn

quasicrystal (Wang

et al.

[8]).

Dislocations in icosahedral

Al-Cu-Fe have been observed in conventional electron

microscopy

and

analyzed by

Devaud-

Rzepski

et al. [9] and

Zhang

et al.

[lo]

the component of the

Burgers

vector in

physical

space

was found to be

parallel

to a 2-fold direction. The contrast

theory

of a

quasicrystal

dislocation with electron

microscopy

has

already

been

approached [9, 11, 12].

Dislocationlike

images

and

grain

boundaries have also been observed

by

Chen et al.

[13]

and Yu et al.

[14]

in Al-Li-Cu

alloys.

In a usual

crystal,

in the kinematical twc-beam mode of

observation,

a dislocation of

Burgers

vector b

gets practically

out of contrast for diffraction vectors G" which

obey

the

relationship

G° b

= 0

[15]. Geometrically,

this

relationship

means that the relevant

diffracting

vectors all

belong

to the same

plane and,

as a matter of

fact,

all the

diffracting

vectors

belonging

to this

plane

are relevant. In

fact,

extinction is

strictly

observed when the dislocation is

parallel

to the film

surface,

is of screw character and the

elasticity

is

isotropic [16].

Small

departures

to these conditions

yield

a residual contrast which in many cases is not

perceived (in particular

mixed dislocations

parallel

to the

film).

As a consequence the

Burgers

vector can be

unambiguously

determined in direction

(but

not in

length),

when two extinctions with

linearly independent

diffraction vectors G° and

GP

are known. The

length

and

sign

themselves can be determined

by considering

other

reflections,

of the

type

b = n, where n

~

0 is an

integer.

The condition of

isotropic elasticity

is in

principle

satisfied in

quasicrystals,

whose icosahe- dral

symmetry

is

larger

than the

symmetry

of cubic

crystals,

and it has been shown in other

respects [11, 12]

that the

generalization

of the condition of extinction G.b

= 0 to a

quasicrystal

is

simply:

Gjj bjj

+

Gi bi

= 0.

(1)

Therefore one needs five

independent

extinctions in order to

get

the direction in

E~

of the

Burgers

vector. Here

Gjj

and

Gi

are

projections

in

IPjj

and

IPi

of the total diffraction vector

G =

Gjj

+ G

i, with

Gjj Gi

= 0.

Only Gjj

is

experimentally known,

but because of the

incommensurability, Gi

can be deduced

unambiguously

from the

knowledge

of

Gjj.

Since six

vectors cannot be

independent,

there is no way to obtain the modulus and the

sign

of the

Burgers

vector

by

the

only

use of

extinguishing

diffractions. We revisit the

geometry

of the

extinguishing

diffraction vectors in the line of reference

[12],

and then turn our attention to the diffraction vectors

Gjj bjj

+

Gi bi

= n

~ 0, showing

that

they

can be used to measure the modulus and the

sign

of the

Burgers vector,

as in the usual 3D case.

Finally

we

analyse

what

happens

when the extinctions do not

obey

the rules which are established.

2,

strong

and weak extinction conditions.

This section

presents

the results of reference

[12]

in a

slightly

different

geometrical

manner, and

brings

some new results to the

question

of the determination of the direction of the

Burgers

vector of a

dislocation,

in

particular equations (5, 6, 11, 12)

and

(13)

and the final comments.

(4)

An extinction G° is of the

strong type (SEC)

if it

obeys

the two conditions:

G( .bjj

= 0

(2)

G

[ bi

= 0.

(3)

If it h so, it can be

proved

that all the diffractions

lG(

in the same row

yield

extinction. I is either an

integer (in

this case the extinction

AG(

is of

evidence)

or it is an irrational of the

type (p

+

Tq)

where p and q are

integers

and T =

(1+ v$)/2

is the

golden

ratio

(~). Assunfing

now that these SEC are true for any diffraction vector

belonging

to two different rows, then

bjj

is known in direction

bll

" fill

(~i

~

~~

~~~~

Furthermore,

since

G( (resp. G()

is the

projection

in IPjj of a lattice vector

G( (resp.

G

(),

its lift G

[ (resp. G()

in

IPi

is determined. Hence

bi

is also known in direction:

bi

=

Pi (G [

A

G( )1 (4b)

fljj and

pi

are two constants which are

related,

since

bjj

and

bi

are

components

of the same

vector b

(see below).

Since the direction which carries

bjj (resp. bi)

is the

projection

in

IPjj

(resp. IPi)

of a 4-dimensional

plane

in

IE~,

b

belongs

to the intersection of two 4-dimensional

planes

in

IE~,

I-e- to a 2-dimensional

plane.

Any exiting

diffraction vector of the set

Gjj

e

(woG(

+

wpG(

w~ = pa + Tqo, wp =

pp + Tqp; p~, q~, pp, qp,

integers)

would also

yield

an extinction of the same dislocation with

Burgers

vector

b;

we call

Ifl(P (resp.

Ill

[P)

the

2-plane

in IPjj

(resp.

in

IPi)

which contains

Gjj (resp. Gi). bjj (resp. bi)

is

perpendicular

to the

plane D(P (resp.

D

[P).

The

explicit

relation between fljj and

pi

is obtained as follows. Let us

introduce,

after Cahn et al.

[17],

the notation b

=

q(h/h', k/k', I/I')

for a 6-dimensional vector b in

lZ~, meaning

that the

components

of b in IPjj are

bjj

= ~ < h +

h'T,

k +

k'T,

I + l'T > and its

components

in

IPi

are

bi

=

11 < h'-

hT,k'- kT,

I'- lT >, these

components being respectively along

a set of three

orthogonal

2-fold axes e(j

(I

=

1, 2, 3)

of the icosahedral

phase

in IPjj and

along

the lifts

e[

in

IPi

Here ~~ =

l/2(2

+

T)

and the

hyperlattice parameter

is taken

equal

to

unity (a

vector G in

reciprocal

space writes

similarly

G

=

~~~ (t/t', u/u', v/v');

Gjj

=

~~~

< t +

t'T,

u +

u'T,

v + v'T >;

etc.). Letting

fljj = p + Tq,

pi

= T + Ts,

(5)

where p, q, r, and s are

integers (we drop

here for convenience the

prefactors

~ or

~~~),

and

writing

that

bjj

and

bi

are

components

of the same vector in

lZ~,

we

get:

r " P +

2q,

s =

-(p

+

q). (6)

(~ A diffraction vector G

= Zn;ei in 6-dimensional

reciprocal

space IE~ defines a row of diffraction

vectors

AG°,

where is a rational. If G° is the diffraction vector of this set such that the

n;'s

are

relative

primes,

all the set obtains with

integer

A's. A row of

diffracting

vectors in

ll~jj obtains as the

Projection

of at least two rows in

IE~;

G° and

G°'

differ

by

a factor of r.

(5)

540 JOURNAL DE

PHYSIQUE

I N°5

For

example,

with

fljj =

I,

we have

pi

"

-T~~

More

generally,

with fljj = T" =

In-

i +

Tin,

(the

Fibonacci numbers are:

lo

=

0, Ii

"

1, f2

=

1, f3

=

2, etc..., f-n

=

(-1)"~~ In, In

=

In-i

+

fn-2),

we

get Pi

"

(-I)"~~T~"~l,

for n

integer positive

or

negative,

viz.

fli/fly(

"

(_i)n-l~-2n+1

can be deduced

from

a

complete

set

of

SEC.

p

and q

are

arbitrary ntegers. The

one "weak extinction contrast"

diffraction

vector G~ =

G(j

+

G[ obeying

the relation:

'~~ ~ ~

~(l '~ll

"

-T~~ Gi bi (G~

=

~~~ (ii /t(,

vi

/u[,

vi

/v();

b

=

~(h/h', k/k', I/I) m"

=

t(h

+

h'ti

+

t[h'+ u(k +k'u;

+

u(k'

+

vii

+

l'vi

+

v(I')

lifts

partially

this arbitrariness.

Equation (7) expresser

the fact that all the

G(j's (resp.

all the

Gl's)

which

obey

WEC rules

can be

partitioned

into sets which have the same

projection m'

on

bjj

and on

bi Therefore,

if

one knows several rows in

IPjj (and consequently

en

IPi) along

which WEC rules are

satisfied,

there is in IPjj

(resp.

in

IPi)

a

family

of

parallel 2-planes

lTjj

(resp. Ifli)

which cut those rows at the extremities of

diffracting

vectors

G(j's (resp. Gl's). bjj (resp. bi)

is

perpendicular

to this

family

of

planes

Ifljj

(resp. Ifli).

See

figure

2 in reference

[12].

This result is no

surprise,

because we

already

know that the

2-plane Ifl(P

in IPjj

(resp.

the

2-plane Ifl[P

in

IPi)

is

perpendicular

to

bjj (resp. bi).

Therefore the

family

of

planes

Ifljj

(resp.

Ifli)

is

parallel

to

Ifl(P (resp.

Ill

[P).

We have anyway

got something

new

by finding

one WEC diffraction vector

G(j (it

is

enough

to know

only one):

the

integers

p and q introduced above

can be

expressed

but to some

integer

constant.

Let us introduce the notation G° =

~~~(t~/t[, uo/u[, va/v[)

and

GP

=

~~~(tp/tj,

up

/uj, vp/vj),

we

have, according

to

equations (4)

and

(6):

~~~bjj

=

pA

+

qC

+

T(qA

+

(p

+

q)C) (8)

~~~bi

=

qA

+

(p

+

q)C T(pA

+

qC) (9)

where A and C are two 3-vectors whose

components

are written:

AI

" uovp upva +

u[vj ujv[ A2

=

A3

=

Ci

=

u[vp viva

+

u«vj upv[

+

u[vj ujv[ C2

=

C3

=

Introducing similarly

the vector notation

t,

=

(t;,

u;,

v;)

,

t(

=

(t(, u(, vi

,

we write the scalar

product:

G(j bjj

+

G[ bi

=

(2

+

T) (t, (pA

+

qC)

+

t( (qA

+

(p

+

q)C))

= 0

(10)

We have therefore

p =

ii (-(t,

+

t()

C

t( A) (11)

q =

ii (t(

C + t;

A) (12)

where

ii

is some

integer.

But note that

i;G~

is also a vector

satisfying

a WEC rule.

Renaming it,

we

get,

after some easy

algebra:

~~~~~' '~"

~

~~~~~~ '~~

= ~~ P~ Pq

~i~~

(6)

If q~

p~

pq takes the smallest

possible

non-2ero

values,

I-e- q~

p~

pq =

(-1)"+~ (n arbitrary integer),

we have p

=

In-ii

q

=

in.

Therefore p and q are

proportional

to two successive Fibonacci numbers.

Note that all the vectors

bjj (resp. bi)

which

satisfy

the same SEC and WEC are

necessarily

commensurate.

If,

for

example, bjj

is a

possible choice,

so that

according

to

equation (7)

we

have

T~~G(j bjj

=

m',

it cannot be so for

b(j

=

(p'+ Tq')bji,

because

T~~G(j

b~j #

lll~(/

+

Tg')

#

lll~p'

+

Tlll~q' (14)

does not

obey equation (7).

In

fact, G(j

would not

extinguish

a dislocation of

Burgers

vector

~~l'

3. Other dilEraction vectors: modulus and

sign

of the dislocations.

Let G =

~~~(t It', u/u',

v

Iv')

be a diffraction vector which does not

extinguish

the dislocation b =

~(h/h', k/k', I/I').

We have:

Gll '~ll "In

+

(n ill)Tl

Gi bi

=

In

+

mT)

G b

= n.

(15)

where n and m are

integers

which

read,

in the notation we have

already

used:

n=t.h+t"h', m=t.h-(t.h'+t"h) (16)

In

principle,

as we shall

show,

the value and

sign

of n can be determined

experimentally by

the same method as in a usual

crystal.

The

important

issue is to determine

Gjj bjj

and

Gi bi separately.

Let us show that the consideration of another diffraction vector

G',

whose

component

in IPjj is

parallel

to G and whose~modulus

G(j)

is incommensurate with

(Gjj(

is

sufficient to measure m, from which the two scalar

products Gjj bjj

and

Gi bi

of

equation (15)

can be deduced. Let indeed

G(j

= (tY +

flT) Gjj

be another diffraction vector.

G(j

suffices to determine

G[

and we have indeed:

G[

=

(a'

+

fl'T)Gi (17)

where tY' = tY +

fl

and

fl'

=

-fl

are

integers.

It is now easy to show that the values of n' and m' relative to G'=

G(j

+

G[,

v12.:

read now:

m'

= am

fin, n'

= an +

fl(n m). (19)

Since tY and

fl

are

known,

the value of

n',

which is measured as the rank of an extinction

contour

[16], yields

m. Observe that m can

only

be reached in an observation in which

fl ~

0.

(7)

542 JOURNAL DE

PHYSIQUE

I N°5

When

fl

=

0,

G'is

parallel

to

G,

while it is not when

fl ~ 0, although G(j

and

G[

are

respectively parallel

to

Gjj

and

Gi (incommensurability

of G and

G').

The

integral

values n,

n', etc.,

relative to the vectors

G, G', etc.,

are determined exper-

imentally by

the same method as in usual

crystals,

which consists in the observation of the

extinction contours at their intersection with

dislocations,

arid the measure of the

displacement

of the dislocation

image

when the diffraction vector varies. This

displacement

is

directly

re- lated to n. In

effect,

as shown in many instances

(see

for

example [12]),

the

phase

which enters the

amplitude

diffracted in a dark field

experiment

with a diffraction vector

Gjj

does include

the total

phase

shift 2xG b

= 2xn due to the 6D

dislocation,

and not the

partial phase

shift

Gjj bjj

due to the

diffracting

vector in

physical

space. This remarkable

property

enables us to extend to

quasicrystals,

mutatis

mutandis,

the whole

theory

of contrast of usual

crystals,

with this

unique (but fundamental)

difference that now each relevant diffraction G has to be

supplemented by

a diffraction G'

parallel

to G and incommensurate with it.

4. Other extinction conditions?

Coming

back to the use of

extinctions,

let us now

investigate

the

possiblity

of

having only

one row of diffraction vectors G"

=

G(+G [

which

obey

a SEC rule. What about the other

extinguishing

diffractions

G~?

We have

bjj

=

G(

A

ujj bi

" G

[

A vi where

ujj (resp. vi)

is some unknown 3-dimensional vector in

Vii (resp. IPi)

which has to be determined

by

the other relevant diffractions

G~

which

yield

relations of the

type G(j bjj

+

G[ b1

= 0. Two

cases arise:

a)

in the most usual case, one

expects

b to be a lattice vector of the

hypercubic

lattice

(a

sc-called

perfect dislocation);

then

necessarily ujj (resp. vi)

is a lattice vector

too,

and ui

(resp.

vii is

unambiguously

defined.

Consequently

one

has, by

a calculation of the same

type

as

above,

~" ~

~~~~i

~

~"' ~i

=

-~~~Gl

A Ui

(20)

These

expressions yield ujj

= T~

vii

and vi =

-T~~

ui

(we neglect

all

components parallel

to

G(

and G

[).

The vectors u and v

(which incidentally

are not

parallel

in

IE~) belong

to the

same row of

possible

lattice

reciprocal vectors,

but to a different row than G°. Therefore we

are in the same situation as above and the SEC rule is

obeyed by

a set of diffraction vectors

Gjj

+

(w«G(

+

wujj

w~ = p~ + Tqa, w = p + Tq; p~, qa, p, q

rationals) belonging

to the

same

plane.

b)

if b is not a lattice vector in

IE~,

then the SEC rule does not

apply

to all the diffraction vectors.

Reciprocally,

if the SEC rule does

apply

to

only

one diffraction row

(and

not

two),

the

corresponding Burgers

vector is not a lattice vector in

lE~.

This is worth some more

comments;

we restrict our discussion to a few. A first

possibility

in

E~

is of the usual

type,

the

corresponding

dislocation

being

what is

commonly

called a

partial dislocation,

for

example

if the

hypercubic

lattice is not

primitive (it

has been shown for

example

that the I-Al-Fe-Cu

phase

is face-centered

[9]);

another

possibility,

which would be characteristic of a

quasicrystal

and cannot be

excluded, especially

if the

"phason" part bi

and the

"phonon" part bjj

of the dislocation have

quite

different

energies,

or if the energy of interaction between these two

parts

is

excessively large,

would be to have a total

Burgers

vector where the

"phonon" part bjj

and the

"phason" part bi

are not

topologically

related

[18].

We shall call such an

object

an

imperfect dislocation, reserving

the term of

partial

dislocation in IE~ to the usual

type

of

partial.

But even in such a case can the

SEC

conditions be

obeyed by

diffraction vectors

lying

in a

plane,

for

example

if

bi

+ 0.

(8)

Pure

WEC,

I-e- without any SEC

diffracting vector,

is the

signature

of a dislocation with no

topological relationship

between

bjj

and

bi

This is the

general

case of an

imperfect dislocation,

for any

perfect dislocation,

whose

Burgers

vector in IPjj is of the

type bjj

=

AG(

A

G(,

should

produce

SEC. But note that the

analysis leading

to

equations (11)

and

(12)

is no

longer valid, and, consequently,

the ends of the WEC diffraction vectors in IPjj cannot be

partitioned

in a set of

2-planes perpendicular

to

bjj.

Acknowledgements.

We wish to thank Proh

Gratias,

Urban and

Wollgarten,

and Dr.

Zhang,

for

letting

us have

a

preprint

of their work

(Ref. [12]) prior

to

publication,

and one of the referees for

inviting

us

to

investigate

the non-extinction case.

References

Ill

SHECHTMAN D., BLECH

I.,

GRATIAS D. and CAHN

J-W-, Phys.

Rev. Lett. 53

(1984)

1951.

[2]SOCOLAR J-E-S-,

LUBENSKY T.C. and STEINHARDT

P.J(, Phys.

Rev. B 34

(1986)

3345.

[3] KLfIMAN M., GEFEN Y. and

PAVLOVITCH, Europhys.

Lett. 1

(1986)

61.

[4] BOHSUNG J. and TREBIN

H.-R., Phys.

Rev. Lett. 58

(1987)

2277.

[s]

KLfiMAN M., Proc. of the

ILL/ODDEST

Workshop on quasicrystalline materials, Ch. Janet and J-M- Dubois Eds.

(World Scientific, Singapore, 1988)

p. 318.

[6] KLfIMAN M. and SOMMERS C., Acta Metall. Mater. 39

(1991)

287.

[7] HIRAGA K. and SHINDO D.,

Jpn

J.

Appl. Phys.

28

(1989)

2556.

[8] WANG

D-N-,

ISHIMASA

T.,

NISSEN

H.-U.,

HOVM6LLER S. and RHYNER J., Philos.

Mag.

B 58

(1988)

737.

[9] DEVAUD-RZEPSKI

I., CORNIER-QUIQUANDON

M. and GRATIAS

D.,

Third International

Meeting

on

Quasicrystals (Mexico, 1989): Quasicrystals

and Incommensurate Structures in Condensed

Matter,

M-J- Yacaman et aJ. Eds.

(World

Scientific,

Singapore, 1990)

p. 498.

[10] ZHANG

Z.,

WOLLGARTEN M. and URBAN

K.,

Philos.

Mag.

Lett. 61

(1990)

125.

[ll]

HATWALNE Y. and RAMASWAMY

S.,

Philos.

Mag.

Lett. 61

(1990)

169.

[12] WOLLGARTEN

M.,

GRATIAS

D.,

ZHANG Z. and URBAN

K.,

Philos.

Mag.

64

(1991)

819.

[13] CHEN

C-H-,

REMEIKA

I.P.,

ESPINOSA G-P- and COOPER

A-S-, Phys.

Rev. B 35

(1987)

7737.

[14]YU

D-P-,

STAIGER W. and KLiMAN

M.,

Philos.

Mag.,

in press.

[is]

HUMPHREYS

C-J-, Imaging

of

Dislocations,

in Dislocations in

Solids,

F-R-N- Nabarro Eds.

N-H-P-C-,

5

(1980)

1.

[16] see for

example:

HIRSCH

P-B-,

HOWIE A., NICHOLSON R-B-, PASHLEY D-W- and WHELAN

M-J-,

Electron

Microscopy

of Thin

Crystals (Butterworths, Paris, 196s);

SAADA

G., Microscopie dlectronique

des lames minces cristallines

(Masson

lr

Cie,

Paris,

1966).

[17] CAHN

J,W.,

SHECHTMAN D. and GRATIAS

D.,

J. Mater, Res, 1

(1986)

13.

[18] CROCOMBETTE J-P- and KL#MAN

M., private

communication.

Références

Documents relatifs

ABSTRACT Following the severe acute respiratory syndrome (SARS) event in 2002/2003, the World Health Organization (WHO) developed outbreak communications guidelines. With the

Support for parsing ontologies with description graphs and description graph rules is available in an extension of the OWL API 7 and HermiT can be used for reasoning with

There are at least three major reasons for incorporating daily mobility in public health research: (i) to identify specific population groups, including socially

Dans toute la feuille, K désigne un corps commutatif..

First of all, due to the rectifica- tion theorem we prove (see Theorem 2.2) that any non-vanishing smooth gradient field ∇u is isotropically realizable locally in R d , in the

Theorem 2.2.. The necessary condition ii) holds true if the conductivity matrix is not symmetric. Several other results also extend to the non-symmetric case. However, the

Given a Hilbert space H characterise those pairs of bounded normal operators A and B on H such the operator AB is normal as well.. If H is finite dimensional this problem was solved

Upon VIHCO, we present the algorithms of parallel and se- quential processing, and explore the answers through exper- iments both on synthetic and real dataset, and the results