Particuology 54 (2021) 1–4

Contents lists available at ScienceDirect

Particuology
journal homepage: www.elsevier.com/locate/partic

Connecting particle sphericity and circularity

John R. Grace ∗ , Arian Ebneyamini
Department of Chemical & Biological Engineering, University of British Columbia, Vancouver V6T 1Z3, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Sphericity, a measure of how much a particle’s shape deviates from spherical, is useful as a shape factor
Received 5 August 2020 when characterizing particulate materials. However, particle surface areas, required when determining
Received in revised form the sphericity, are very difficult to measure. As a result, the circularity, derivable from microscopic views,
10 September 2020
is often measured instead and assumed to be equal to the sphericity. This paper shows that the two
Accepted 14 September 2020
quantities are generally not equal for simple non-spherical shapes and provides advice on improving the
Available online 24 October 2020
estimation of sphericity from circularity.
© 2020 Chinese Society of Particuology and Institute of Process Engineering, Chinese Academy of
Keywords:
Non-spherical particles Sciences. Published by Elsevier B.V. All rights reserved.
Sphericity
Circularity
Shape factors

Introduction
Nomenclature
Natural particles are rarely truly spherical in shape. Neverthe-
a Oblate spheroid major semi-axis, m less, engineers and scientists like to treat them as if they were
a Prolate spheroid minor semi-axis, m true spheres, often utilizing a single “shape factor” to correct for
b Oblate spheroid minor semi-axis, m deviations from the spherical shape. The most common shape fac-
tor, used, for example to account for particle surface areas when
b Prolate spheroid major semi-axis, m estimating pressure drops through packed beds and in predicting
D Cylinder diameter, m minimum fluidization velocities, is the sphericity, defined as
H Cylinder height, m Surface area of a sphere of the same volume as the particle
r Sphere radius, m Sphericity = ϕ = . (1)
Actual exterior surface of the particle
r Hemisphere radius, m
The sphericity is 1 for spheres and <1 for all other particle shapes,
W, U and L Cuboid sides with W ≥ U ≥ L and L being vertical,
deviating further from 1 as shapes differ more and more from
m
spherical (Geldart, 1986; Kunii & Levenspiel, 1969). However, in
practice, for assemblies of real particles, the sphericity is very dif-
Greek letters
ficult to determine accurately as it requires careful measurement
ˇ Cylinder aspect ratio (D/H), (-)
of particle exterior surface areas, as well as particle volumes, of
 Circularity, (-)
three-dimensional particles.
 Cuboid aspect ratio (W/L), (-)
To overcome the difficulty of measuring the sphericity, it is com-
 Cuboid aspect ratio (U/L), (-)
mon to determine instead a two-dimensional analogue, called the
ϕ Sphericity, (-)
circularity, defined as
 Oblate spheroid aspect ratio (a/b), (-)
¯ Prolate spheroid aspect ratio (b/a), (-) Perimeter of sphere of the same projected area
Circularity =  = .(2)
Actual projected perimeter of the particle
Like the sphericity, the circularity is 1 for a sphere and dif-
fers more and more from this value as the shape deviates further
and further from spherical. However, unlike the sphericity, the
circularity can be determined from two-dimensional microscope
images in which both the projected area and projected perime-
∗ Corresponding author. ter of individual particles can be seen and therefore estimated,
E-mail address: [email protected] (J.R. Grace). with particles having orientations typically similar to those of most

https://doi.org/10.1016/j.partic.2020.09.006
1674-2001/© 2020 Chinese Society of Particuology and Institute of Process Engineering, Chinese Academy of Sciences. Published by Elsevier B.V. All rights reserved.
J.R. Grace, A. Ebneyamini Particuology 54 (2021) 1–4

Table 1
Equations for different geometric shapes.

Shape/orientation Volume Surface area 2D projected area 2D perimeter

4r 3
Sphere 3
4r 2 r 2 2r
3 2 2
2r
Hemisphere 3
3r r 2r
Cube/Cuboid (L-directiona ) L×W ×U 2 (L × W + U × W + L × U) W ×U 2W + 2U
D2 2
Cylinder (projected on its side)  4
H DH +  D2 DH 2H + 2D
2 2 D2
Cylinder (projected on its end)  D4 H DH +  D2   0.5  4
D
2
1+ 1− b
2a +     0.5
4 b2 a2
Oblate Spheroid (viewed from 3
a2 b 2
0.5 Ln a2 2a
b2 b2

below 
while lying flat)
a [major semi-axis]
1−
a2
1− 1−
a2
1< b [minor semi-axis]
⎛ ⎞
⎜1 +
 0.5
⎟
      b
 0.5 arcsin
2
⎝ ⎠
4 2 2 b a
Prolate Spheroid (viewed from 3
a b 2a 1− 2 ab  3 a+b − 3a + b a + 3b
 while lying flat) 
2 b
below a 1− a
2
b [major semi-axis] b
1<
a [minor semi-axis]

a
L, the shortest cuboid side, does not appear in a 2D projection.
b
(Hachmi, Sesbou, Benjelloun, & Bouanane, 2011; Weintrit, 2015).

particles undergoing sedimentation or fluidization. Hence it is com-

mon to use the circularity in place of the sphericity. However, the
two quantities are not identical: For example, for a perfect cube, it
is simple to show (Grace, 2020) that the sphericity and circularity
differ appreciably, being 0.806 and 0.886, respectively.
Li, Li, Zhao, Lu, and Meng (2012) calculated sphericities for
objects of a number of specific shapes. In this paper, we derive val-
ues of both circularity and sphericity for several different shapes,
similar to those of some real particles, in order to consider whether
there are ways of improving the estimation of sphericities based
on measured circularities. A reference figure is then provided to
enable estimation of the particle sphericity based on the class of
particle shape and the circularity.

Geometric shapes and formulae

Three- and two-dimensional projections of the geometric

shapes, considered in this work, are shown in Fig. 1. Note that par-
ticles of interest typically self-orient on their flat side, so that, for
example, the two-dimensional projections of the oblate and pro-
late spheroidal particles become a circle and ellipse, respectively.
A cylindrical object can be projected on either its side or its end,
depending primarily on the aspect ratio (D/H). This results in two
possible microscopic projection shapes for cylindrical particles, as
depicted in Fig. 1. Basic equations, derived from the definitions of
sphericity and circularity above for the given shapes and orienta-
tions, are summarized in Table 1. Fig. 1. Three-dimensional (3D) shapes considered in this work, together with their
corresponding two-dimensional (2D) projections.

Results and discussion

Higher circularities are also projected at higher sphericities, except
Table 2 summarizes the formulae derived to calculate the for those conditions where the two-dimensional projection of the
sphericity and circularity of the geometric shapes considered. Note particle becomes a perfect circle or square (e.g. cuboid with U = W,
that the circularity of several non-spherical objects, including a cylinder projected on its end, and oblate spheroid).
hemisphere, cylinder (projected on its end) and oblate spheroid, The variation of particle sphericity as a function of circularity is
is equal to unity, due to the circular two-dimensional projection plotted in Fig. 5 for the geometric shapes and orientations consid-
of these objects. In addition, sphericity and circularity of all these ered. Note that the sphericity of the cuboidal objects is a bivariate
shapes, except the cuboid, can be expressed based on univariate function of the particle circularity and aspect ratio. Thus, numer-
equations, corresponding to a unique correlation between their ous lines (for different aspect ratios) could be plotted in Fig. 5
sphericity and circularity. for cuboidal particles, which, within the typical range of particle
The variation of sphericity and circularity as functions of particle aspect ratios (W ≥ U ≥ L), would likely be located between dashed
aspect ratio are shown in Figs. 2–4 for different particle shapes. Note (U = L) and unbroken (U = W) lines. Fig. 5 also indicates that the par-
that the sphericity and circularity of the cuboid particles follow ticle circularity is always unity for spherical, hemispherical, oblate
bivariate functions (see Table 2), so that only two extreme cases spheroidal and cylindrical (if projected on its end) particles. How-
( =  and  = 1) are plotted in Fig. 2. Increasing the particle aspect ever, their sphericities can vary significantly, based on the particle
ratio leads to greater sphericity, up to an aspect ratio of unity, while shape and aspect ratio (see Figs. 3 and 4). In addition, a particle’s
further raising of the aspect ratio results in opposite behaviour. sphericity is usually smaller than its circularity. However, there are

2
J.R. Grace, A. Ebneyamini Particuology 54 (2021) 1–4

Table 2
Formulae for sphericity and circularity of different particles.

Shape/orientation Sphericity (ϕ) Circularity ()

Sphere 1 1
Hemisphere ∼ 0.84 1
 W U
 2
2( 3 ) 3 () 2
1
Cube/Cuboid (L-directiona )  = L
≥= L
≥1 4
+×+ +

 D
 8(
3ˇ2 3
)
2
1
(ˇ) 2
Cylinder (projected on its side) ˇ = H
≤1 16
2ˇ+ˇ2 1+ˇ

 D
 8(
3ˇ2
)
2
3
Cylinder (projected on its end) ˇ = ≥1 16
1
 
H 2ˇ+ˇ2
4
a [major semi-axis] 4 3
Oblate Spheroid (viewed from below while lying flat) 1 <  = b [minor semi-axis] 0.5 1
+( 2 −1)
2 2+ Ln[ ]
0.5 0.5
( 2 −1) −( 2 −1)
2 1
3 2
¯ ¯)
Prolate Spheroid (viewed from below while lying flat)
b
1 < ¯ = [major
semi-axis]
2
2
 2(

a[minor semi-axis] ¯ 0.5 [3(1+ ¯ )− (3+ ¯ )(1+3 ¯ )]

1+ arcsin[[1− 1 ] ]
2 0.5 2

( ¯ −1)
¯

a
L, the shortest cuboid side, does not appear in 2D projection.

Fig. 2. Sphericity and circularity of cuboidal particles at (a) U = W and (b) U = L.

Fig. 3. Sphericity and circularity of cylindrical particles projected: (a) on their sides; (b) on their ends.

exceptions for prolate spheroidal particles, as well as for long cylin- cularity and the observed predominant shape of given particles. In
ders (if projected from the side and D/L < 0.45, see Fig. 3(a)) and most cases, one may infer that it is reasonable to approximate the
cuboids (e.g.  ≥ 6.5, see Fig. 2(b)). two shape factors as being equal for smoothly rounded (spheroid-
As noted above, the circularity is often used in place of the like) particles, whereas angular particles and particles with sharp
sphericity because the former is much easier to measure. However, edges are likely to be better represented if the sphericity is esti-
it is clear from Fig. 5 that the sphericity can significantly differ from mated to be less than the circularity by approximately 10%–20%.
the circularity, and that they are only approximately equal in a few
cases (e.g. cylindrical objects projected on side, with 0.3 < D/L < 1,
and prolate spheroidal particles with 0.43 < a/b < 1, see Figs. 3(a)
Conclusions
and 4). Fortunately, there is a unique correlation between the par-
ticle sphericity and circularity for several common particle shapes.
Circularity and sphericity are formulated for a num-
Therefore, Fig. 5 is a useful reference to improve the estimation of
ber of simple particle shapes. It is shown that both
the particle sphericity based on an experimentally measured cir-
circularity and sphericity can be expressed as univari-

3
J.R. Grace, A. Ebneyamini Particuology 54 (2021) 1–4

ratio is shown to increase the sphericity to an aspect ratio of

unity, with further increase of the aspect ratio resulting in oppo-
site behaviour. Higher circularity usually corresponds to higher
sphericity, except for particles with circular and square two-
dimensional projections.
For most particle shapes, the sphericity is somewhat lower than
the circularity, with the difference being negligible only under very
specific conditions. This clearly indicates that circularity should not
be used in place of sphericity, especially for sharp-edge particles
(e.g. cuboids). Fig. 5, which plots sphericity as a function of circu-
larity for simple shapes and orientations of interest, can be used to
assist in approximating particle sphericity based on circularity for
different particle shapes.

Declaration of interests

The authors declare that they have no known competing finan-

cial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Acknowledgments
Fig. 4. Sphericity and circularity of spheroidal particles.

The authors acknowledge the financial support of the Natural

Science and Engineering Research Council of Canada (NSERC) via
their CREATE and Discovery Grant programs. We are also grateful
to Banafsheh Hassani for her assistance with the preparation of the
paper.

References

Geldart, D. (1986). Single particles, fixed and quiescent beds. Chapter 2. In D. Geldart
(Ed.), Gas fluidization technology. Chichester: Wiley.
Grace, J. R. (2020). ). Properties, minimum fluidization and powder groups. Chapter 2.
In J. R. Grace, X. Bi, & N. Ellis (Eds.), Essentials of fluidization technology. Weinheim:
Wiley-VCH Verlag. http://dx.doi.org/10.1002/9783527699483
Hachmi, M., Sesbou, A., Benjelloun, H., & Bouanane, F. (2011). Alternative equa-
tions to estimate the surface-to-volume ratio of different forest fuel particles.
International Journal of Wildland Fire, 20(5), 648–656. http://dx.doi.org/10.1071/
WF09042
Kunii, D., & Levenspiel, O. (1969). Fluidization engineering. New York: Wiley.
Li, T., Li, S., Zhao, J., Lu, P., & Meng, L. (2012). Sphericities of non-spherical objects.
Particuology, 10(1), 97–104. http://dx.doi.org/10.1016/j.partic.2011.07.005
Weintrit, A. (2015). The need to develop new geodesic-based computational
algorithms for marine navigation electronic devices and systems. In 2015 Inter-
national association of institutes of navigation world congress (IAIN). pp. 1–12.
Fig. 5. Sphericity vs. circularity for particles of different shapes and orientations. Prague: IEEE. http://dx.doi.org/10.1109/IAIN.2015.7352263

ate functions of the particle aspect ratio, with cuboid

particles being an exception. Elevating the particle aspect