the motion of rigid ellipsoidal particles in slow flows
TRANSCRIPT
![Page 1: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/1.jpg)
Tecfonophysics, 113 (1985) 163-183
Eisevier Science Publishers B.V., Amsterdam - Printed in The Nethertands 163
THE MOTION OF RIGID ELLIPSOIDAL PARTICLES IN SLOW FLOWS
BRETT FREEMAN
Department of Geology. University Park, Nottingham, NG7 2RD (United Kingdom) *
(Received February 8, 1984; revised version accepted September 24, 1984)
ABSTRACT
Freeman, B., 1985. The motion of rigid ellipsoidal particles in slow flows. Teczon~phy~jc~, 113: 163-183.
Preferred orientations of fabric elements in tectonites are usually interpreted with reference to two
continuum models, namety those of March (1932) and Jeffrey (1922). It is argued that the first of these is
usually inappropriate if the fabric elements are rigid inhomogeneities embedded in a deformable matrix.
Jeffrey’s model is more applicable but until recently its use has been limited to problems involving
axisymmetric particles and simple flow geometries. In this paper Jeffrey’s equations have been solved
numerically in order to investigate the behaviour of more general systems. The motions of two particle
shapes, one a prolate spheroid, the other a triaxiaf ellipsoid, are considered in parallel for four specified
flow geometries which involve pure or simple shear and combinations of both. Any departures from
particle axisymmetry or from simple flow geometries have a significant effect on particle behaviour
compared with that predicted by spheroid/simple flow solutions.
INTRODUCTION
Orientation distributions of planar and linear fabric elements in deformed rocks, such as micas in slates (Oertel, 1970; Tullis, 1976), prismatic crystals in igneous rocks (Bhattacha~a, 1966) and elongate pebbles in glacial tills (Glen et al., 1957; Allen, 1982, p. 199) are often invoked as indicators of the magnitude and geometry of the bulk deformation. The theoretical basis for this is found in the contrasting works of March (1932) and Jeffrey (1922). In the March model it is assumed that fabric elements behave as passive material markers, i.e. they rotate with angular velocities equal to those of lines and planes of equivalent position in a homogeneous and isotropic deforming medium. This is usually an inappropriate treatment for problems involving rigid i~omogeneities (such as porphyroblasts in a rnet~o~~c
* Present address: Department of Geology, University of Newcastle upon Tyne, Newcastle upon Tyne,.
NE1 7RU (U.K.).
0040-1951/85/$03.30 0 1985 Elsevier Science Publishers B.V.
![Page 2: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/2.jpg)
164
tectonite) since their rates of rotation are controlled through a tensor of third rank
(Bretherton, 1962) which is entirely dependent on particle shape. However. symme-
try arguments show that if the particle is ellipsoidal, then the shape tensor has only
three non-zero components, and these are easily evaluated given the semi-axial
lengths (Bretherton, 1962). Jeffrey (1922) derived the general equations of motion for
a rigid particle of neutral buoyancy isolated in a slow viscous flow whose streamlines
are steady at large distances from the particle. He also provided explicit solutions for
spheroidal particles in simple shear flows. Of course it is rare that geologists
encounter perfectly ellipsoidal particles but from the work of Eirich and Mark
(1937) it seems that the effects of minor surface imperfections on particle motions
may be neglected. This is borne out experimentally even in extreme cases as Bartok
and Mason (1957) have shown. (For a further discussion on the effects of non-el-
lipsoidal shapes see Ferguson ( 1979).)
Anulytical solutions of Jeffrey;; equutions
Provided that the volume concentration of particles is sufficiently small such that
the disturbed velocity field around one particle does not interfere with that around
any other, then Jeffrey’s (1922) equations constitute an appropriate framework for
modelling multi-particle behaviour. Within these constraints the existing limitations
of the model are due chiefly to the complexity of the governing equations. Analytical
solutions have been found only for axisymmetric particles in simple flows. The well
known solution for simple shear is given by Jeffrey (1922, eqns. 48 and 49):
tana=rtan[y/(r+ l/r)] (1)
tan /? = C [ cos*a: + ( l/r* ) sin2a] I” (2)
where (Y and p are the polar angles (azimuth and plunge) of the particle long axis
with respect to the plane of undisturbed flow, r is the axial ratio and y is the shear
strain. Equation (1) is clearly periodic and together (1) and (2) predict that the long
axis of a prolate spheroid will precess one of an infinite family of closed, spherical
elliptical orbits described by the orbit constant C, which depends only on the initial
particle orientation and axis ratio. This motion is quite distinct from that of a
passive material line (as in the March model): during the course of the rotation the
particle always decelerates towards the shear plane, passes through the plane and
accelerates away from it. In contrast the end of a passive material line remains in the
same plane moving at constant velocity in the direction of shear.
For flows involving no vorticity Jeffrey’s equations have been solved for pure
shear (Gay, 1966, 1968). the azimuth of the long axis being given by:
1 (r- 1) ln(cot aI) = ln(cot ai) + - ~
ln
i 1
3 1’2
2 (r+l) h, (3)
![Page 3: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/3.jpg)
165
and the plunge:
sin 201. ( i l/2
Cot #If = COt pi + & f
where X, and X, are the two principal quadratic extensions (the third being unity).
Subscripts f and i refer to the final and initial angles respectively. Tullis (1976)
presents similar equations for pure flattening strains of orientation distributions of
spheroids.
Although, as in simple shear, the material coordinates of the end of a prolate
spheroid differ from those of a corresponding passive linear marker after deforma-
tion, it is interesting to note that the angular paths swept out by both are identical
for all coaxial irrotational strains. The particle motions differ only in their angular
velocities, passive lines rotating faster than any spheroid.
Experimental verification of Jeffrey’s model
Several workers have attempted to assess the validity of Jeffrey’s model by
investigating the correspondence between the theoretically predicted behaviour and
observations obtained experimenta~y for simpie shear systems. Trevelyan and
Mason (1951), using a Couette apparatus, found an excellent agreement between the
calculated and observed rates of rotation of spheres, and verified the nature of the
orbits for prolate cylinders, though the actual periods are always less than the
predicted values for spheroids. This may not be too important geologically since the
shear strains necessary to complete an orbit are very large compared to anticipated
shear strains in rocks. More restricted shear box type experiments also show a fair
correspondence between observed and predicted behaviour for geologically realistic
strains (Ghosh and Ramberg, 1976, fig. 5), but all observations and initial orienta-
tions are restricted to the plane perpendicular to the vorticity, i.e. they are all C = co
orbits. A further consideration, with regard to the interpretation of tectonites, comes
from the notion that the matrix rheology is unlikely to be Newtonian; Ferguson
(1979) has addressed this problem in some detail and concludes that the model
should give acceptable approximations even for power law fluids provided that
strains are not too high.
Previous applications of Jeffrey ‘f model
Until recently the effects of strain on multiparticulate systems have been mod-
elled using the analytical solutions to Jeffrey’s (1922) equations (e.g., Reed and
Tryggvason, 1974; Tullis, 1976; Harvey and Ferguson, 1978). There are two major
restrictions here. Firstly discussion must be limited to plane strains of either pure or
simple shear although, clearly, many geological deformations cannot be approxi-
mated by such a simple treatment. For example, Sanderson’s (1982) differential.
transport model for strain variations in thrust sheets combines pure and simple shear
![Page 4: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/4.jpg)
in its interpretation of sidewall ramps and steep zones parallel to nappe transport
directions. Similarly interplay between pure and simple shear is important during the
development of folds, as indicated by Ramberg (1975). Secondly the particles are
required to be axisymmetric which greatly limits the geological applications. In
particular, it precludes any meaningful interpretation of fabrics composed of general
ellipsoidal (or approximately ellipsoidal) particles in the light of hitherto published
multiparticle models.
Recent theoretical work (Gierszewski and Chaffey, 1978; Hinch and Leal, 1979)
has brought attention to the rather more complex behaviour of triaxial particles in
simple shear flows. Both studies use numerical solution of Jeffrey’s (1922) equations,
and the results have a fair correspondence to the experimental results obtained from
Couette flow (Harris et al., 1979). I have used a similar numerical approach to
investigate the behaviour of isolated axi- and non-axisymmetric particles in some
general, geologically interesting, flows.
THEORY
Consider a set of right handed orthogonal Cartesian axes X,’ fixed in orientation
but able to move so that the origin of the basis is always coincident with the centre
of gravity of an ellipsoidal particle which is suspended in a slowly deforming fluid. A
second coordinate system X, is also centred at the origin of X,’ and is instantaneously
coincident with the principal axes of the particle. The relationship between X, and X,’
is given at any instant by the rotation matrix:
cos 4 cos 9 - cos e sin 9 sin 4 , cos I/ sin $ + cos e cos $3 sin I$ , sin 4 sin e
-sin$cos$-cosesin$cos#, -sinJisincp+cosecosqbcosrC,, cos+sinB
sin e sin 9 - sin e cos + cos e
(5)
so that X, = R,,X,‘.
8, + and + are the three euler angles and are the minimum number of parameters
needed to describe the orientation of the particle. They are defined from three
successive rotations (Fig. l), R,, being the matrix product of the rotation matrices
for each of the three operations. In this work the rotations are made in sequence
about Xi, X, then X, following the “x ” convention of Goldstein (1980, p. 147)
(though they are somewhat arbitrary).
At large distances from the particle the undisturbed flow is specified by the
velocity gradients tensor (spatial description):
(6)
![Page 5: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/5.jpg)
167
which satisfies the constant volume condition:
( Lri, i =j are the rates of natural strain and L:j, i #j, are the shear strain rates.)
The rate-of-deformation and vorticity tensors of the undisturbed flow in the fixed
.coordinate system are the symmetric and antisymmetric parts respectively of L:,:
(84
W
(for example, see Malvern, 1969).
To find the rate of deformation and vorticity tensors with respect to the rotating
Line of. nodes
Fig. 1. Eufer angle definitions. A’: refer to the fried axes and X, to tile rotating particle axes.
![Page 6: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/6.jpg)
X, coordinates we employ the usual rules for transformation of tensors:
E!, = R<,R,&, (9a)
and:
Q,, = R,,R,&, (9b)
Now, if the semi axial lengths of the particle are ai the only non-zero components of
Bretherton’s (1962) shape tensor are:
Jeffrey’s (1922) equations of motion (p. 169, eqn. 37) are now easily rewritten as:
where w, are the angular velocities of the particle about its own axes X,. If the euler
angles change with velocities b,& 4, then it is straightforward to formulate the w, in
those terms (Goldstein, 1980, p, 176):
w1 =BsinBsin\I,+~cosIC, /
w2 = $2 sin t? cos + - S sin $.J
W) = 4 cos 8 + $L
j (121
I
which after rearrangement gives three differential equations:
B = w,cos +LJ - w,sin 4
~=(w,sin++o,cos~)/sinB
\t = w3 - (b cos 8
(13)
These are solved numerically as an initial value problem. Throughout this analysis
the boundary conditions are set in terms of L,‘,, though it is obviously necessary to
appreciate the actual strains involved. Geologists have tended towards the use of the
deformation gradients tensor:
dx, F,:=dX; 04)
for the computation of deformations (e.g., Flinn, 1978; Sanderson, 1982) and the
quadratic stretch tensor:
D,; = F:, F;p (15)
for the calculation of finite strains (e.g., Sanderson, 1982; De Paor, 1983). The same
convention is used here, but because we are dealing with velocities, t;lIt is time
dependent with a rate of change:
c; = Lip FiJ (16)
![Page 7: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/7.jpg)
169
(Malvern, 1969, p. 163). Equation (16) are thus nine differential equations which can be solved over an interval, t - rO, to give ‘;;;.
All systems of differential equations used here have been solved to approximately eight significant figures accuracy using a variable-order, variable-step Adams method (Numerical Algorithm Group routine W02CBF).
ISOLATED PARTICLES IN SLOW FLOC’S
In this study two particle shapes have been considered, a prolate spheroid of axis ratio 3 : 1: 1 and an ellipsoid of axes 3 : 1” 1 : f . They have been chosen specifically to
have the same major axis length and identical volumes so their motions can be sensibly compared for equivalent boundary conditions. For most of the following discussion we consider the motion from nine starting orientations. These are represented as the plunge of the long axis (X,) on a stereographic projection and the subsequent motion is depicted by a smooth curve with dots at ten time unit intervals. To calculate the finite strain between any number of dots, n, we simply evaluate the quadratic stretch tensor:
The quadratic extensions are then the eigenvalues of O/j and the eigenvectors give their directions in space.
Here we use the velocity gradients:
0 0 0 LI, = 0.05 0 0
[ 1 0 0 0
The amount of shear per ten time units is L;, = y = 0.5 and ej is simply:
As noted previously the motion of an axisymmet~c particle is periodic with the ends of the spheroid describing closed spherical elliptical orbits about the pole to the plane of the undisturbed flow, i.e. the vorticity. The period is constant in time if y is constant, but is always proportional to y so that for one half rotation:
y = Iz(T + l/r) (18)
Further it appears that the particle will precess the same orbit for all time depending on C, the orbit constant, which is determined by the initial orientation, and is equal to tan a for a when /3 = 0. These are referred to as Jeffrey orbits.
![Page 8: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/8.jpg)
170
Any departure from axisymmetry has profound effects on the nature of the orbit
compared to that of a spheroid starting from the same long axis orientation (see also
Gierszewski and Chaffey, 1978; Hinch and Leal. 19’79). Figure 2 summarizes the
results for nine initial orientations and shows the corresponding Jeffrey orbits for the
prolate particle. It appears that the motion is still periodic about the vorticity. but
the periods are not constant in time. Imposed on the primary periodicity is a
secondary drift through families of Jeffrey orbits. This is a fundamental difference
between axi- and non-axisymmetric particle motions; the latter can drift through the
plane of the undisturbed flow. If the time spent in the flow is large then the motion
reveals itself to be doubly periodic; drift through the Jeffrey orbits eventually returns
the particle to its original position.
To assess the general behaviour in simple shear we first consider the degenerate
cases for which the rotations are singly periodic. Dealing only with the long axis we
can identify four initial orientations for which the corresponding Jeffrey orbit is
c‘ = 03, and of these four tjnere are two distinct periods:
8=90, +=90,0. l&=0,0
and:
e = 90, Q, = 90.0 li, = 90,90
(there are of course symmetrically equivalent positions outside the range 8, (p,
+ = 0 - 90). Secondly there are an infinite set of orientations for which the corre-
sponding orbit constant is c’ = 0, and which have identical periods:
B=O, $X=0-90, +=o-90
Now it is possible to consider the general behaviour in terms of the relationship of
the initial orientation to that of one of the degenerate cases. Figure 2 shows that for
the first few increments of shear (up to about y = 2) the orbits of the two particle
geometries are qualitatively similar. As initial orientations approach degeneracy the
similarity between the two orbits persists for greater shear magnitudes. Conversely as
the symmetry of the position with respect to the plane of the undisturbed flow
decreases, the orbits become less like the equivalent closed Jeffrey orbit. We note,
however, that for ail non-degenerate boundary conditions orbital drift is significant
by shears of y = 4. Thus approximations for triaxial particle motions based on
prolate particle theory are always likely to be misleading.
(2) Pure shear and other coaxial strains
The equations of Gay (1966, 1968) (eqns. (2) and (3) here) indicate an important
distinction between coaxial irrotational pure shear and flattening and simple shear
for axisymmetric particles in that the former are non-periodic. Indeed this can easily
be shown to be true for all coaxial strains involving any ellipsoidal particle geometry.
For coaxial strains the only non-zero components of the velocity gradients tensor are
![Page 9: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/9.jpg)
L -
,
![Page 10: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/10.jpg)
172
the diagonals, i.e. L:, f 0, i =.j. Hence there is no vorticity and L?,, = 0 (eqn. 9b).
Now, as the particle axes approach parallelism with the fixed axes, ,k; then:
1 0 0
H,, + L 1
0 1 0 0 0 1
and since E,‘, = 0, i f j (eqn. 9a), then E,, + 0 and w, become asymptotically smaller.
When o, = 6 the particle is in a position of stable equilibrium and will remain fixed
in position unless the applied flow is modified.
For axisymmetric particles the components of Bretherton’s shape tensor reduce
further to B, = 0 and B, = -B,. Therefore spin about the long axis is zero and o2
and w3 are proportional to E3, and E,, respectively and are scaled by constants of
equal modulii. Their velocities, which are clearly dependent on the axis ratio of the
particle, become maximum as B, = -B, -+ 1 and at the limit ~,/a, = m. w2 = -E,,
and w3 = E,, which are the angular velocities of a passive material line. Therefore
the axisymmetric particle motion for oblate and prolate spheroids follows trajecto-
ries which are the same as those predicted by the March model for passive planes
and lines respectively.
Triaxiality does not have such a drastic effect on the motion as in simple shear
but we should note that all the w, are non-zero so the ellipsoid motion always lies
somewhere between the two extremes of behaviour for spheroids.
(3) Simultaneous pure and simple shear (L;, = - L;,, L;, + 0)
All the flows described in this section are geometrically similar in that the plane
perpendicular to the vorticity has a pure shear superimposed on it. However, the
velocity components producing the pure shear are varied over an order of magnitude
so the ratio L;,/L;, ranges from 0.1 to 1.0, and the velocity producing simple shear
is held constant. Particle paths for this type of flow are given by Ramberg (1975, fig.
3). We note that cross sections through such a flow, made in the plane perpendicular
to the vorticity, are identical and that their area remains constant throughout
deformation so the resulting finite strain ellipsoid is k = 1 (Flinn, 1978).
The behaviour of axisymmetric particles is shown in Fig. 3a. Drift through the
Jeffrey orbits occurs even for small amounts of pure shear and as this component is
increased the periodic motion becomes subordinate to the asymptotic behaviour
described in the previous section. When the velocities for pure and simple shear are
equal the particle rotates from any starting position towards x’. This is an important
observation because it implies that any multiparticulate fabric will tend to be prolate
even though the finite strain ellipsoid is plane. Qualitatively similar results apply to the
triaxial particles in as much as the gross departures from the double periods become
more profound as the ratio of pure to simple shear velocities approach unity (Fig. 3).
More subtle modifications occur in accordance with the symmetry of the initial
orientation; in low symmetry orientations, and particularly when 6 is small, depar-
![Page 11: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/11.jpg)
(A)
e=60
30
:rl-
a
I 1 I L=(
p) I
xj
.005
0
.05
-.c10
5
0 0
.o 1
0
.05
-.Ol
0 0
.05
0
.05
-.05
0 0
1.05
1
,500
0
1.10
5
.501
0
0
I[
I.646
0 ,5
21
0 0
0
,951
0 0
,905
0 0
.607
0
0
0 1.
0
0 1. 1
0 0 1. 1 (a
>
Fig.
3.
a.
Path
s of
a
sphe
roid
in
si
mul
tane
ous
pure
an
d si
mpl
e sh
ear
for
two
star
ting
poin
ts.
Com
pare
w
ith
Fig.
2.
b, c
, d.
Pa
ths
of a
n el
lipso
id
in s
imul
tane
ous
pure
an
d si
mpl
e sh
ear
for
thre
e in
itial
or
ient
atio
ns
of
the
long
ax
is
and
thre
e in
itial
ro
tatio
ns
abou
t th
e lo
ng
axis
.
See
text
fo
r di
scus
sion
.
![Page 12: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/12.jpg)
. . ‘.. \ -\
![Page 13: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/13.jpg)
0 0 ; 0 0 r
175
-A v
00,: -
O z 0 0 0 0 0 0 ;: 0
I’ I’ I’
![Page 14: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/14.jpg)
176
,
0
1” 0 0 G
c ,----l
- -I u+j 0 “, 0
II LL (D .-
0 - :: O c
0 0 0 0 0 0 0 0 0
:: 0 0 0 0 0 0
I’ 0 ; 0
i/‘ / ---.
‘.
i ” I’ / p/ \ \
![Page 15: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/15.jpg)
111
1 I 1 I --
0 0 $ 0 0 ; 0 0
: cl _
;; h
0 ? O 2 Q cn O 0 ; 0
0 0 0
1 1
0 0 0
Q 6 0 0 q 0
![Page 16: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/16.jpg)
178
![Page 17: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/17.jpg)
179
- ._ .-
a3 II
r I
0 0 0
0 : 0 I’
t r
0 0 r
0 0 0
0 0 0
-c 0 0 G -
0 0 0
![Page 18: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/18.jpg)
180
![Page 19: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/19.jpg)
181
tures from the simple shear case become less predictable, cusps appear in the particle
trajectory and the time spent in orientations close to the Xi/X; plane becomes relatively large.
(4) Simple shear on twoperpendicularplunes with simultaneous pure shear (L;, = - L;,,
L;, = L;, f 0)
This is essentially a more general version of the previous flow. The two simple shear components are equivalent to a single shear direction on the plane whose pole is the vector [co545 co545 01, thus the pure shear defo~ation plane is at 45’ to the simple shear plane, and the vorticity of the flow is along the vector [ -cos45 cos45 01. Without the pure shear component we would expect the particle to precess about the vorticity as in previous examples. This is the case, but even when Z&i, is small, drift through the Jeffrey orbits is spectacular (Fig. 4a) for initial orientations which are close to the vorticity. However, if the long axis lies close to the plane perpendicu- lar to the vorticity then the particle rotates in quasi-stable orbits with constants fluctuating about C = 00. Similar behaviour is observed for L;, = L;,/5 but when L;, = L;, all initial orientations eventually rotate to a position a few degrees away from Xi.
Periodicity is retained for all initial orientations if the particle is triaxial (at least up to L;, = L;,). The rate of secondary drift through the orbits (Fig. 4a, b, c) is, as usual, dependent on the ratio L&/L;,. When this ratio is large the particle motion is similar to the axisymmetric one, but when L;, = L;, a rapid movement towards the X;/X; plane is followed by singly periodic rotation in a slightly elliptical spherical orbit which crosses the plane perpendicular to the vorticity twice every period.
SUMMARY
Numerical solution of Jeffrey’s (1922) equations governing the motion of a rigid ellipsoidal particle suspended in a creeping fluid facilitates the study of spheroidal and ellipsoidal particles in plane and non-plane strain flows, with or without vorticity. The treatment here covers some strain geometries which are geologically important, and deals with magnitudes of strains up to and beyond those anticipated in most geological situations. It is by no means exhaustive, but demonstrates that departures from axisymmetry and/or plane strain have important consequences for the nature of the particle motion. Of particular importance is the motion of particles in flows involving pure and simple shear. The suggestion is that multiparticle fabrics will become prolate even though the finite strain ellipsoid is plane. Computer simulation experiments have been carried out in order to investigate multiparticle behaviour in the flows already described here and will be published in a later paper,
![Page 20: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/20.jpg)
182
ACKNOWLEDGEMENTS
I thank Cohn Ferguson for reading and improving the manuscript. This work was
carried out during the receipt of a NERC research studentship at Nottingham
University.
REFERENCES
Allen, J.R.L., 1982. Sedimentary Structures, vol. 1. (Develop. Sedimental., 30A) Elsevier, Amsterdam, 593
PP. Bartok, W. and Mason, S.G., 1957. Particle motions in sheared suspensions. V. Rigid rods and collision
doublets of spheres. J. Colloid Sci., 12: 243-262.
Bhattacharyya, D.S., 1966. Orientation of mineral lineation along the flow direction in rocks.
Tectonophysics, 3: 29-33.
Bretherton, F.P., 1962. The motion of rigid particles in a shear flow at low Reynolds number. J. Fluid
Mech., 14: 284-301.
De Paor, D.G., 1983. Orthographic analysis of geological structures. 1. Deformation Theory. J. Struct.
Geol., 5: 255-277.
Eirich, F. and Mark, H., 1937. ijber L~sungs~ttelbindung durch Immobilisie~ng. Papierfabrikant, 27:
251-258.
Ferguson, C.C., 1979. Rotations of elongate rigid particles in slow non-Newtonian flows. Tectonophysics,
60: 247-262.
Flinn. D., 1978. Construction and computation of three~imensional progressive deformations. J. Geol.
Sot. London, 135: 291-305.
Gay, N.C.. 1966. Orientation of mineral lineation along the flow direction in rocks: a discussion.
Tectonophysics, 3: 559-564.
Gay. N.C., 1968. The motion of rigid particles embedded in a viscous fluid during pure shear deformation
of the fluid. Tectonophysics, 5: 81-88.
Ghosh. S.K. and Ramberg, H., 1976. Reorientation of inclusions by combination of pure and simple
shear. Tectonophysics, 34: I-70.
Gierszewski, P.L. and Chaffey, L.E., 1978. Rotations of an isolated triaxial ellipsoid suspended in slow
viscous flow. Can. J. Phys., 56: 6-11.
Glenn, J.W., Donner, J.J. and West, R.G., 1957. On the mechanism by which stones in till become
oriented. Am. J. Sci., 255: 194-205.
Gofdstein, H., 1980. Classical Mechanics. Addison-Wesley, London, 672 pp.
Harris, J.B., Nawaz., M. and Pittman, J.F.T., 1979. Low-Reynolds-number motion of particles with two or
three perpendicular planes of symmetry. J. Fluid Mech., 95: 415-429.
Harvey, P.K. and Ferguson, CC.. 1978. A computer simulation approach to textural interpretation in crystalline rocks. In: D.F. Merriam (Editor), Recent Advances in Geomathematics. Pergamon,
Oxford, pp. 201-232.
Hinch, E.J. and Leal, LG., 1979. Rotation of small non-axisymmetric particles in a simple shear flow. J. Fluid Mech., 92: 591-608.
Jeffrey, G.B., 1922. The motion of ellipsoidal particles immersed in a viscous fluid. PtOC. R. SOC. London. Ser. A, 102: 161-179.
Malvem, LX., 1969. Introduction to the Mechanics of a Continuous Medium. Prentice-Hall, Englewood
Cliffs, N.J., 713 pp.
March, A.. 1932. Mathematixhe Theorie der Regelung nach der Korngestalt bei affiner Deformation. 2.
Kristallogr., 81: 285-298.
![Page 21: The motion of rigid ellipsoidal particles in slow flows](https://reader030.vdocuments.site/reader030/viewer/2022020313/575090921a28abbf6b96be33/html5/thumbnails/21.jpg)
183
Oertel, G., 1970. Deformation of a slaty, lapillar tuff in the lake district, England. Geol. Sot. Am. Bull.,
81: 1173-1187.
Ramberg. H., 1975. Particle paths, displacement and progressive strain applicable to rocks. Tectonophysics,
28: l-37.
Reed, L.J. and Tryggvason, E., 1974. Preferred orientations of rigid particles in a viscous matrix deformed
by Pure shear and simple shear. Tectonophysics, 24: 85-98.
Sanderson, D.J., 1982. Models in strain variation in nappes and thrust sheets: a review. Tectonophysics,
88: 201-233.
Trevelyan, B.J. and Mason, S.G., 1952, Particle motions in sheared suspensions. I. Rotations. J. Colloid
Sci., 5: 345-367.
Tullis, T.E., 1976. Experiments on the origin of slatey cleavage and schistosity. Geol. Sot. Am. Bull., 87:
745-753.