HAL Id: jpa-00236688
https://hal.archives-ouvertes.fr/jpa-00236688
Submitted on 1 Jan 1962
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.
Quenching defects in solid solutions and their effect on precipitation
R.B. Nicholson
To cite this version:
R.B. Nicholson. Quenching defects in solid solutions and their effect on precipitation. J. Phys.
Radium, 1962, 23 (10), pp.824-827. �10.1051/jphysrad:019620023010082400�. �jpa-00236688�
QUENCHING DEFECTS IN SOLID SOLUTIONS AND THEIR EFFECT ON PRECIPITATION
By R. B. NICHOLSON,
Department of Metallurgy, University of Cambridge.
Résumé. 2014 Le réseau de dislocation dans les alliages d’aluminium trempés est très différent de celui qui existe dans l’aluminium pur trempé. Ceci a été étudié par microscopie électronique (trans- mission) et par dilatométrie, et cet effet est probablement lié à l’absorption de paires atome dissous
2014 lacune par les dislocations vis qui glissent, et à leur redistribution en boucles de dislocations
prismatiques.
Le vieillissement entraîne la précipitation sur les dislocations et les effets peuvent être analysés
à l’aide de la sous-structure de dislocations, du désaccord géométrique entre le précipité et le
réseau de la matrice.
Abstract. 2014 The dislocation structure of quenched aluminium alloys is very different from
quenched pure aluminium. This effect has been studied by transmission electron microscopy and
is probably associated with the absorption of vacancy-solute atom pairs on gliding screw
dislocations and their redistribution as prismatic dislocation loops.
Ageing causes precipitation on dislocations and the effects can be analysed in terms of the dis- location substructure and the misfit geometry of the precipitate and matrix lattices.
23; 1962,
The excess vacancies in a metal or alloy quen-
ched from a high temperature may anneal out in a
number of ways to produce a variety of dislocation substructures. The density of dislocations can be
quite high, e,g, 109 -1011 lines jsq . cm, and if the alloy is a supersaturated solid solution these dislo- cations can act as important sites for the segre-
gation and subsequent precipitation of solute
atoms.
The quench structures of some pure metals seem to follow the general prediction of Kuhlmann- Wilsdorf [1] that excess vacancies will collect to form small voids which subsequently collapse to give prismatic dislocation loops. However there
are a number of anomalies such as the hetero- geneous distribution of prismatic loops in quenched
aluminium [2], the relatively few prismatic loops
found in quenched copper and nickel and their
apparent association with dislocation lhies [3] and,
most strikingly, the near absence of prismatic loops in concentrated aluminium alloys and the
presence of long regular helical dislocations formed
by the absorption of vacancies on screw dislo- cations [4]. Dilatometer experiments by Taka-
mura [5] suggest that, in the case of gold, the size
of the specimen and hence the amount of plastic
deformation during quenching is an important
factor affecting the behaviour of the excess va-
cancies. Recent experiments using transmission electron microscopy [6] have shown that this may be a general result and that the mode of influence of the plastic deformation is the absorption and re-
distribution of vacancies on gliding screw dislo-
cations. We shall consider these results briefly.
Figure 1 is a typical electron micrograph of a quenched aluminium alloy. There are a large
number of prismatic dislocation loops which have
formed from homogeneously nucleated vacancy clusters and several regular helical dislocations. It is found that the proportion of vacancies which
Fie. 1. - Thin foil of AI-16 % Ag water quenched from
525 OC showing prismatic dislocation loops and helical dislocations ( X 40,000, printed X 0,79).
anneal out as helices increases in more concen-
trated alloys. One feature which is not present
in figure 1 but is frequently observed in these alloys is rows of closed dislocation loops lying along
110 > directions. Most helices and rows of
loops are observed near the triple points of grain
Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphysrad:019620023010082400
825
boundaries where plastic flow would be expected during quenching due to stress concentrations exis-
ting at these points. This suggésts that thé for-
mation of helices and rows of loops is associated
FiG. 2. - Thin foils of Al-7 % Mg alloy water quenched
from 500 °C, (a) without plastic deformation : (b) with
N 2 % plastic deformation during quenching. ( X 40,000,
printed X 0,73).
with plastic déformation during quenching. To
test this hypothesis, some Al-7 % Mg specimens
’were quenched in a jig so that (a) there was a mini-
mum amount of plastic deformation during quen-
ching and (b) the specimen7was homogeneously
deformed by about 2 % during quenching. The
results of this experiment, shown in figure 2, are particularly striking since the presence of magne- sium atoms in solid solution prevents homo-
geneous nucleation of dislocation loops from va-
cancy clusters after this treatment. Hence the loop structure arises solely from plastic defor-
mation effects, Very few loops are observed in figure 2(a) where the amount of plastic çlefor-
mation is restricted but in figure 2(b) many rows of
loops and helices are observed along traces of t 111.B
planes.
We now consider the behaviour of a glissile dislo-
cation loop expanding in a matrix supersaturated
with vacancies. The edge components of the dis- location can continue to glide conservatively while absorbing vacancies by climb but the movement of screw dislocations muts be restricted due to their interaction with the vacancies and the consequent
formation of jogs, cavities or simply loose atmos- pheres of vacancies. Eventually the dislocation must either cease to move or break free from its at-
mosphère leaving a row of defects behind. rn the former case the vacancies will distribute themselves
uniformly along the dislocation to give a helix while
in the latter case many defects will be observed in the path of the dislocation. This suggestion enables
us to give an approximate explanation of the great variety of quench structures observed in metals and
alloys of high stacking fault energy. In metals like Al, Cu and Ni, vacancies are absorbed on gliding screw dislocations activated by quenching
stresses and are re-distributed as prismatic dislo-
cation loops. Thus in Al we observe patches of homogeneously nucleated dislocation loops sepa-
rated by regions which have been active slip planes during quenching, showing irregular dislocations and dislocation loops formed by the re-distribution process. In Cu and Ni homogeneous nucleation
appears to be difficult and hence dislocation loops
ar e only observed in slip bands. In dilute Al alloys, typifield by Al-7 % Mg in figure 2, the
eif ect of the solute atoms is to restrict the dislo- cations to a single slip plane and hence the re- distributed defects are more closely aligned than in
pure Al. In concentrated Al alloys the redistri-
bution process seems to be prevented and hence
each dislocation can only move a short distance
before being immobilised by the absorption of
vacancies. Thus more dislocations must be pro- duced to relieve the quenching stresses and the
result is a high density of helical dislocations.
Another type of quenching defect can be pro- duced near particles in an alloy. If the particles
are present at the homogenization température,
the defêets usually take the form of arrays of pris.
matic dislocation loops whose sign is determined by
the rélative thermal expansion coefficients, oi thé
particle and matrix [7]. However if thé partiulë is
formed during quenching there may be a high local supersaturation of vacancies at the particle matrix
interface which is sufficient to operate a Bardeen- Herring source [8]. Figure 3 is an example of the
array of dislocation loops produced by such a
mechanism.
We now consider the effect of this quench dislo-
cation substructure on subsequent precipitation phenomena. Dislocations can affect precipitation
FIG. 3. - Thin foil of Al-7 % Mg oil quenched from 500°C showing operation of a Bardeen-Herring source at a pre-
cipitate particle (x 40,000).
FIG. 4. - Thin foil of AI-16 % Ag water quenched from
525 °C’and aged 1 day at 160 °C showing y’ precipitates
formingion ahelical dislocation ( x 80,000, printed X 0,73).
in various ways chiefly by their action in causing segregation of solute atoms or by distortion of the matrix lattice so that it approaches the structure of the precipitate. A good example of the latter effect occurs in the Al-Ag system where the c. p. h.
y’ transition precipitate nucleates on the narrow stacking faults at dislocations in the matrix [9]. A stacking fault in an f. c. c. matrix is, of course,
equivalent to thin layer of c. p. h. lattice and’hence
FIG. 5. - Thin foil of Al-4 % Cu water quenched from
540 °C and aged 1 hour at 200 °C showin5’ precipitation
of 0’ on helical dislocations (x 40,000, printed x 0,73).
FIG. 6. - Thin foil of Al-7 % Mg water quenched from
500 oc and aged 7 hours at 220°C showing precipitation
of rods on 9certain fsegments of dislocation loops.
(X 34,000,printed X :0,73).
827
this is an easy nucleation site for the y’ precipitate.
Figure 4 shows a helical dislocation which has
partly transformed to a row of y’ precipitates. In
this alloy system, dislocations appear to be the
only nucleating sites for y’ precipitates.
In the Al-Cu system, the 03B8’ transition precipitate
is nucleated in the matrix and on dislocations so
here the dislocation is only acting as a catalyst for precipitation. It is found that the dislocation only catalyses those orientations of precipitate where
FIG. 7. - Thin foil of Al-7 % Mg oil quenched from 500 °C and aged 5 hours at 220 °C showing precipitation on a
dislocation source similar to that shown in figure 3 ( X 29,200).
the Burgers vector of the dislocation can partly
accommodate the misfit between the precipitate
and the matrix [10]. Thus in figure 5, the Burgers
vector of the dislocation is a 2 (110) and precipitates parallel to (100) and (010), with misfits in the 100 > and 010 > directions, have nucleated
on the dislocations.
Figure 6 shows the formation of small rod-shaped précipitâtes on prismatic dislocation loops in an Al-Mg alloy. Precipitation has only occurred on
one segment of the loop. Since these loops are
formed by the Kuhlmann-Wilsdorf [1] reaction,
the Burgers vector makes an angle of 55° to the plane of the loop. Hence only one segment of the loop is pure edge dislocation and it is this segment
which is the most active nucleation site for preci- pitation. The dislocation loops .shown in figure 3
are pure edge dislocation and hence precipitation
occurs all round the loop (fig. 7).
Wé have shown that the dislocation substructure introduced by quenching is more complicated than
was originally thought and that plastic defor-
mation during quenching is important in deter- mining the na.ture of the substructure. Alloys
have more complicated substructures than pure metals owing to the effects of precipitates and
solute atoms in solution. On ageing, dislocations act as preferential nucleating sites for precipitation
and can also lead to the formation of some preci- pitate structures which would not be stable in a
perfect lattice.
Acknowledgement. - Much of this work has been done in collaboration with J., D. Embury and
C. M. Sargent.
REFERENCES
[1] KUHLMANN-WILSDORF (D.), Phil. Mag., 1958, 3, 125.
[2] WILSDORF (H. G. F.) and KUHLMANN-WILSDORF (D.),
J. Appl. Physics, 1960, 31, 516.
[3] SMALLMAN (R. E.), WESTMACOTT (K. H.) and COILEY (J. A.), J. Inst. Metals, 1959-1960, 88, 127.
[4] THOMAS (G.) and WHELAN (M. J.), Phil. Mag., 1959, 4, 511.
[5] TAKAMURA (J.), Acta Met., 1961, 9, 547.
[6] SARGENT (C. M.), EMBURY (J. D.) and NICHOLSON
(R. B.), Acta Met. 1962, 10, 1118.
[7] EMBURY (J. D.) and NICHOLSON (R. B.), J. Inst.
Metals., in press,
[8] EMBURY (J. D.) and NICHOLSON (R. B.), Acta Met.,
in press.
[9] NICHOLSON (R. B.) and NUTTING (J.), Acta Met., 1961, 9, 332.
[10] NICHOLSON (R. B.), Proc. Eur. Reg. Cont. on Electron
Microscopy at Delft, 1960, 1, 375, De Nederland.
Vereniging voor Electronenmicroscopie, Delft, 1961.
