Neutron diffraction determination of residual stress in 6061Al-15 vol%SiCw composites with
different whisker orientation
R.Fernández1, G.Bruno2, A.Borrego1, G.González-Doncel1, A.Pyzalla2
1
Dept. of Physical Metallurgy, Centro Nacional de Investigaciones Metalúrgicas (CENIM), C.S.I.C.
Av. de Gregorio del Amo 8, E-28040 Madrid, Spain
2
Hahn-Meitner Institut, Glienicker Str.100, D-14109 Berlin, Germany
PACS: 81.05.N, 61.12, 07.10.P
Abstract
The determination of residual stresses, RS, in 6061Al-15vol%SiCw composites was carried out to
correlate them with whisker orientation, distribution and aspect ratio. The composites were obtained
by a powder metallurgical, PM, route and consolidated by extrusion at four different temperatures,
Text. This is the main parameter affecting the whisker orientation/ distribution; an increase in T ext
leads to an increasing whisker alignment with extrusion axis. The results of the sin2 plots account
for a non-hydrostatic stress state in all composites. This is expected because of the deviatoric
component of the RS tensor generated by the whisker geometry. The RS can explain the Strength
Differential Effect, SDE, observed in uniaxial testing (tensile vs. compressive behaviour) of these
composites.
Introduction
Among the different proposed strengthening mechanisms related to the addition of ceramic
reinforcement to a monolithic metallic material in a Metal Matrix Composite, MMC, [1], RS play a
very important role. For example the SDE, which has been observed by several authors [2], cannot
be explained by means of well-known mechanisms like load transfer (shear-lag) and dislocation
enhancement. The SDE, however, has been related to RS [3]. The present work has been focused on
the effect of whisker orientation on RS in MMCs.
1. Materials and Experimental Procedures
The investigated composites were prepared by a powder metallurgical route from powder of
AA6061 and 15%vol -SiC single crystal whiskers. Four different composite materials extruded at
different temperatures were obtained, Table 1, [4]. Unreinforced AA6061 was also prepared. In
short, it was seen in [4] that: (i) The whisker length is similar for all materials (about 3.4  0.1m);
(ii) The volume fraction of the <111> and <100> matrix fibre texture components developed during
extrusion, depends on Textr; (iii) The fraction of whisker aligned with the extrusion axis (≤15º)
increases with Textr.
Material Textr Fv<111> Fv<100> Fv of aligned
Code
(°C) ≤15º (%) ≤15º (%) whiskers (%)
C08
525
46
20
----C32
300
28
14
26
C34
359
27
14
25
C38
498
50
16
31
C45
534
37
18
31
Table 1- Processing and microstructural
parameters of the investigated materials. The
grain size and the whisker length are very
similar in all composite materials.
The sin2 method [5] was used to obtain RS by neutron diffraction, ND, fig.1. Here, the lattice
spacing (or the strain) is plotted against sin2 Measurements were performed on the E3
diffractometer at HMI, Berlin, using a wavelength  = 1.37 Å and a gauge volume of 222 mm3
on cylinders, 6.5 mm diameter and 13 mm height. The Al- and SiC-(311) planes were used because
they are low index and low texture planes and provide for a good approximation to the isotropic
macroscopic elastic properties; i.e. RS are directly comparable to mechanical behaviour. The RS in
the materials were studied in a peak-aged condition [6]. Texture analysis by means of x-rays shows
that in the reinforcement a large fraction of randomly aligned particles is present and the <111>
fibre texture present is not so strong. However, a pronounced <111> + <100> fibre texture was
found in the matrix. For this reason only a few  directions are suitable for RS analysis. In fact, a
good signal-to noise ratio in ND could be measured only on few  angles, which were chosen to
perform the sin2 scan. For calculation of strain, the matrix C08 (unreinforced) was taken as
reference, unstressed material. Because of the axisymmetry of the extrusion process, the stress
tensor components rad and hoop are equal.
 = 0°  ax
 = 90°  rad
Incident Beam
Scattering
Plane
Sample


Extrusion
Axis
Fig.1- Description of the geometry of the ND
experiment.  (ranging from 0° to 90°) is the angle
between the scattering vector Q and the sample axis
(extrusion direction). The sample is tilted within the
scattering plane, thus using the so-called -method.
The axial and radial sample directions correspond to
the angles  = 0° and  = 90°.
Q
Detector
2
2. Results and Discussion
The results of the sin2 plot measurements, figs. 2, indicate that strains (stresses) are tensile in the
matrix and compressive in the SiC phase and that, in general, there is a non hydrostatic RS state in all
reinforced materials. This non hydrostatic stress state is, however, expected in a whisker reinforced
MMC because of the deviatoric component  = ax - hydr of the RS tensor generated by the whisker
geometry itself [7]. The latter is associated to the difference between the tension and compression
behaviour, SD. No clear effect of whisker distribution/orientation, specific of each composite, is
appreciated, as the slope of the curves, and thus the RS, are very similar.
800
200
(a)
Al Phase
0
400
C32
200
C34
C38
C45
0
-200
0
0.2
0.4
2
sin 
0.6
0.8
1
(strain)
(strain)
600
(b)
C38
C45
C32
C34
-200
-400
-600
SiC Phase
-800
0
0.2
0.4
0.6
0.8
1
2
sin 
Fig.2- strain, , vs sin2 curves for all the investigated samples. (a) matrix, (b) reinforcement. The difference
between the strains measured at  = 0° and  = 90° is proportional to the RS that causes the SDE (see table 2). The
hydrostatic stress contribution in the SiC may be related to the choice of the unstrained material.
RS are generated in the quenching process from the solid solution process (520ºC) prior to the ageing
treatment, and the extrusion temperature (when lower than the solid solution mixing temperature) can
be supposed to have little influence on RS. The tension and compression strength differences, SD/2 are
in good agreement (Table 2) with the ND measurements and calculations with a modified Eshelby
model [8], which takes into account the orientation distributions of whiskers. The anomaly shown by
the C45 material can be associated to the inhomogeneous whisker distribution observed.
Mat. SD/2
Al Eshelby SiC ’SiC
Code (MPa) (MPa) (MPa) (MPa) (MPa)
C08
0
0
C32
10
3
8
-43
-14
C34
9
11
9
-23
-62
C38
12
14
14
-69
-78
C45
26
15
12(29)
-47
-87
Table 2- Comparison between calculated and
measured stresses.  are the measured differences
between the axial and the hydrostatic RS in both
phases. Eshelby is the RS in the matrix calculated
with a modified Eshelby model. Typical errors are: 7
MPa for the matrix and 30 MPa for the SiC. ’SiC
is the calculated RS in the reinforcement using the
equilibrium condition between both phases.
This material developed a stripe like structure. If we take into account that only some of the regions
are effectively reinforced, the Eshelby value mirrors the measured SD for which case the value in
parenthesis, table 2, is obtained. The RS from Neutron Diffraction give an average value, which agrees
with an homogeneous Eshelby model.
3. Acknowledgement
HMI, Berlin, Germany, in the frame of the Access to Research Infrastructures of the EC (contract n.o
ERBFMGECT950060) and Projects: MAT97-1059-C02-01 from CICYT and 07N/0066/98 from
Consejería de Educación y Cultura of CAM, Spain.
4. References
[1]- T.Christman, A.Neeldman, S.Suresh: Acta metall. 37, 3029 (1989)
[2]- R.S.Arsenault, S.B.Wu: Mater. Sci. Eng. 96, 77 (1987)
[3]- R.S.Arsenault, M.Taya: Acta metall. 35, 651 (1987)
[4]- A.Borrego, R.Fernández, M.C.Cristina, J.Ibáñez, G.González-Doncel: submitted to Textures and
Microstructures
[5]- H.Dölle, V.Hauk: Z.Metallkde 69, 410 (1978)
[6]- A.Borrego, J.Ibáñez, V.López, M.Lieblich, G.González-Doncel: Scripta mater. 34, 471 (1996)
[7]- K.Maeda, K.Wakashima: Scripta mater. 36, 335 (1997)
[8]- P.J.Withers, W.M.Stobbs, O.B.Pedersen: Acta metall. 37, 3061 (1989).