International Journal of Scientific & Engineering Research, Volume 4, Issue 12, December-2013 1530

ISSN 2229-5518

Generalised Gaussian Quadrature over a

Sphere
K. T. Shivaram

Abstract — This paper presents a Generalised Gaussian quadrature method for the evaluation of volume integral
𝐼 = ∭𝑣 𝑓(𝑥, 𝑦, 𝑧)𝑑𝑥 𝑑𝑦 𝑑𝑧 , where 𝑓(𝑥, 𝑦, 𝑧) is arbitrary function and 𝑣 refers to the volume of spherical region bounded by �(𝑥, 𝑦, 𝑧)/−𝑎 ≤ 𝑥 ≤
𝑎, −√𝑎2 − 𝑥 2 ≤ 𝑦 ≤ √𝑎2 − 𝑥 2 , −�𝑎2 − 𝑥 2 − 𝑦 2 ≤ 𝑧 ≤ �𝑎2 − 𝑥 2 − 𝑦 2 �, volume integral is convert to surface integral by using Gauss divergence
theorem then we have applied the Generalised Gaussian quadrature rules over a circle region to evaluate the typical volume integrals over the
spherical region with various values of 𝑎. The efficacy of this method is finally shown by numerical examples.

Index Terms— Finite element method , Generalised Gaussian Quadrature , spherical region.

——————————  ——————————

1. Introduction

IJSER
T HE finite element method has proven to be an
efficient tool for the numerical analysis of two- or A good overview of various method for evaluating
volume integrals is given by Lee and Requicha [8]
three-dimensional structures of whatever complexity,
in mechanical, thermal or other physical problems. It is evaluation of volume integrals by transforming the
widely recognized that computational cost increases volume integral to a surface integral and then into a
greatly with structure complexity, being larger with parametric line integral is given by Timmer and Stern
three-dimensional analyses than with two-dimensional [4], Cattani and Paoluzzi [2] gave a symbolic solution
ones. It is therefore desirable to devise simplified to both volume and surface integration of polynomials
approaches that may provide a reduction in overall by using a triangulation of solid boundary, Nagaraja
computational effort. An example of considerable and Rathod [9] have discussed the volume integral of
importance is the study of bodies of revolution where a a function is express to sum of four integrals over the
three dimensional problem is solved by a two- unit triangle by using gauss divergence theorem,
dimensional analysis. In particular, they are used for shivaram [11] evaluation of surface integral of arbitrary
Problems involving calculations Volume, center of function over circle region by using generalized
mass, moment of inertia and other geometric properties Gaussian quadrature rule.
of rigid homogeneous solids frequently arise in a large
number of engineering applications, in The paper is organized as follows. In Section II volume
CAD/CAE/CAM applications in geometric modeling of the sphere is equal to 8 times the volume in the first
as well as in robotics and similar problems in other octant. In Section III we will introduce the Generalized
areas of engineering which are very difficult to analyze Gaussian quadrature formula over a circle region of
using analytical techniques, These problems can be various values a. and In Section IV we compare the
solved using the finite element method. numerical results with some illustrative examples.

————————————————

K.T. Shivaram is currently pursuing Ph.D in Mathematics

Bangalore University, Bangalore , India,.
E-mail: [email protected]

IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 4, Issue 12, December-2013 1531
ISSN 2229-5518

2 GENERALISED GAUSSIAN QUADRATURE OVER A 3 FORMULATION OF INTEGRALS OVER A

SPHERICAL REGION QUARTER- CIRCLE REGION

The Numerical integration of an arbitrary function f

Z
over a quarter circle is given by

𝑎 � 𝑎 2 −𝑥 2
I = ∬C f(x, y)dx dy = ∫0 ∫0 𝑓(𝑥, 𝑦)𝑑𝑦 𝑑𝑥 (1)

Y Where a is the radius of the circle

The integral of the eqn.(1) can be transformed to the

X square {( 𝜉, 𝜂)/ 0 ≤ 𝜉 ≤ 1,0 ≤ 𝜂 ≤ 1}. Transformation is

Z 𝑥= 𝑎𝜉 𝑎𝑛𝑑 𝑦 = 𝑎 𝜂 �1 − 𝜉 2 (2)
C
We have

IJSER
o Y 𝑎
I=∫0 ∫0
�𝑎 2 −𝑥 2
𝑓(𝑥, 𝑦)𝑑𝑦 𝑑𝑥
X B 1 1
= ∫0 ∫0 𝑓�𝑥(𝜉 , 𝜂), 𝑦(𝜉 , 𝜂)� 𝐽 𝑑𝜉 𝑑𝜂 (3)
A
Fig. 1 a) Volume of the Spherical region Where J (𝜉 , 𝜂) is the Jacobians of the transformation

b) OABC is piecewise smooth and is comprised of 𝜕𝑥 𝜕𝑦

𝐽(𝜉 , 𝜂) = �𝜕𝜉 𝜕𝜉
� = 𝑎2 �1 − 𝜉 2
four surfaces 𝜕𝑥 𝜕𝑦
𝜕𝜂 𝜕𝜂

From eqn.(3) , we can write as

The Numerical integration of an arbitrary
1 1
function 𝑓(𝑥, 𝑦, 𝑧) over a Spherical region is given by
𝐼 = � � 𝑓 �𝑎 𝜉 , 𝑎 𝜂 �1 − 𝜉 2 � 𝑎2 �1 − 𝜉 2 𝑑𝜉 𝑑𝜂
0 0
�𝑎 2 −𝑥 2 �𝑎 2 −𝑥 2 −𝑦2
𝑎
𝐼=� � � 𝑓(𝑥, 𝑦, 𝑧)𝑑𝑧 𝑑𝑦 𝑑𝑥 = ∑𝑛𝑖=1 ∑𝑛𝑗=1 𝑎2 �1 − 𝜉 2 𝑤𝑖 𝑤𝑗 𝑓(𝑥�𝜉𝑖 , 𝜂𝑗 � , 𝑦�𝜉𝑖 , 𝜂𝑗 �) (4)
−𝑎
−�𝑎 2 −𝑥 2 −�𝑎 2 −𝑥 2 −𝑦 2
�𝑎 2 −𝑥 2 �𝑎 2 −𝑥 2−𝑦 2
Where 𝜉𝑖 , 𝜂𝑗 are Gaussian points and 𝑤𝑖 , 𝑤𝑗 are
𝑎
corresponding weights. We can rewrite eqn. (4) as
=8 � � � 𝑓(𝑥, 𝑦, 𝑧)𝑑𝑧 𝑑𝑦 𝑑𝑥
0
0 0
I = ∑𝑁=𝑛×𝑛
𝑘 𝑊𝑘 𝑓(𝑥𝑘 , 𝑦𝑘 ) (5)
Generalised Gaussian quadrature rule for integrating
Where 𝑊𝑘 = 𝑎2 �1 − 𝜉 2 𝑤𝑖 𝑤𝑗 , (5a)
of volume integral bounded by spherical region
V = �(𝑥, 𝑦, 𝑧)/0 ≤ 𝑥 ≤ 𝑎, 0 ≤ 𝑦 ≤ √𝑎2 − 𝑥 2 , 0 ≤ 𝑧 ≤ 𝑥𝑘 = 𝑎 𝜉 , (5b)
�𝑎2 − 𝑥 2 − 𝑦 2 � with 𝑎 = 0.5, 1, 3 and these volume
integral convert to surface integral using Gauss 𝑦𝑘 = 𝑎 𝜂 �1 − 𝜉 2 , (5c)
divergence theorem.
if 𝑖, 𝑗, 𝑘 = 1,2,3, − − − − ,

IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 4, Issue 12, December-2013 1532
ISSN 2229-5518

we find out new Gaussian points( 𝑥𝑘 , 𝑦𝑘 ) and weights 0.457879042 0.014749598 0.002734699
coefficients 𝑊𝑘 of various order N =5,10,15,20 by using 0.002826114 0.142476426 0.001524313
eqn. (5a), (5b) and (5c) and Tabulated in Table 1 and 2 0.036715186 0.142094059 0.009440868
0.142478702 0.136571561 0.020111952
TABLE 1
0.309741132 0.111847299 0.019911726
Gaussian points and weights coefficient over the region 0.457879042 0.057237994 0.006061324
with a=1 and N = 5 0.002826114 0.309736184 0.001842738
0.036715186 0.308904938 0.011413041
𝑥𝑘 𝑦𝑘 𝑊𝑘
0.0056522282 0.0056521379 0.0004429668 0.142478702 0.296899321 0.024313287
0.0734303717 0.0056369691 0.0027435258 0.309741132 0.243150088 0.024071235
0.2849574044 0.0054178878 0.0058445536 0.124432359
0.457879042 0.007327519
0.6194822640 0.0044370593 0.0057863677
0.9157580830 0.0022706706 0.0017614267 0.002826114 0.457871727 0.001096133
0.0056522282 0.0734291987 0.0027509084 0.036715186 0.456642927 0.006788925
0.0734303717 0.0732321351 0.0170378176 0.142478702 0.438895459 0.014462499
0.2849574044 0.0703859617 0.0362957910
0.309741132 0.359439925 0.014318516
0.6194822640 0.0576436240 0.0359344457
0.9157580830 0.0294991960 0.0109387957 0.457879042 0.183943827 0.004358696
0.0056522282 0.2849528525 0.0060972512
0.0377634717

IJSER
0.0734303717 0.2841881181
0.2849574044 0.2731431219 0.0804478077 4 NUMERICAL RESULT
0.6194822640 0.2236945980 0.0796469039
0.9157580830 0.1144759879 0.0242452942 Exact value N Computed
0.0056522282 0.6194723684 0.0073709513 value
0.0734303717 0.6178098764 0.0456521641
0.2849574044 0.5937986412 0.0972531486 5 0.000325512
0.5 0.5 2 − x 2 0.5 2 − x 2 − y 2
0.6194822640 0.4863001763 0.0962849380 ∫ ∫ ∫ xyz dzdydx 0.000328739
10
0.9157580830 0.2488647182 0.0293100741 1) 0 0 0
0.0056522282 0.9157434547 0.0043845316 0.5 0.5 2 − x 2 15
0.000325477
0.0734303717 0.9132858532 0.0271557016 1 1
= ∫ ∫ xy ( − x 2 − y 2 )dydx 0.000325502
0.2849574044 0.8777909181 0.0578499953 0 0 2 4 20
0.6194822640 0.7188798502 0.0572740656
0.9157580830 0.3678876548 0.0174347841 = 0.0003255208333
2) 5 0.048666861
0.5 0.5 − x 2 2 0.5 − x − y
2 2 2
0.048305638
TABLE 2 ∫ ∫ ∫ x + y + z dzdydx 10
0 0 0 0.048519698
15
Gaussian points and weights coefficient over the
region with a=0.5 and N = 5 0.048518793
20

𝑥𝑘 𝑦𝑘 𝑊𝑘
0.002826114 0.002826069 0.000110742 0.5 0.52 − x 2
2 1
0.002818485
= ∫ ∫ [ ( 1 - 4x 2 − 4 y 2 + x
0.036715186 0.000685881 0 0 3 2
0.142478702 0.002708944 0.001461138 3 3
2
0.309741132 0.00221853 0.001446592 + y) − (x + y) 2 ]dydx
2

0.001135335
3
0.457879042 0.000440357 = 0.04851863601
0.002826114 0.036714599 0.000687727
5 0.449923809
0.036715186 0.036616068 0.004259454 1 12 − x 2 12 − x 2 − y 2
x2 y + z
0.142478702 0.035192981 ∫ ∫ ∫ [ ]dzdydx 0.446724537
0.009073948
3) 0 0 0 x+ y 10

0.309741132 0.028821812 0.008983611 0.448256687

15

IJSER © 2013
http://www.ijser.org
 International Journal of Scientific & Engineering Research, Volume 4, Issue 12, December-2013 1533
ISSN 2229-5518

5 CONCLUSIONS
20
0.448827578
In this paper Volume integral(triple) of arbitrary
1− x2 − y2
1 12 − x 2
=∫ ∫ [ function over a spherical region �(𝑥, 𝑦, 𝑧)/−𝑎 ≤ 𝑥 ≤
0 0 2( x + y )
𝑎, −√𝑎2 − 𝑥 2 ≤ 𝑦 ≤ √𝑎2 − 𝑥 2 , −�𝑎2 − 𝑥 2 − 𝑦 2 ≤ 𝑧 ≤
x2 y 1− x2 − y2
+ ]dydx �𝑎2 − 𝑥 2 − 𝑦 2 � with 𝑎 = 0.5, 1, 2 convert to double
( x + y)
= 0.4488168268 integral by using Gauss divergence theorem. We have

4) 5 0.02500014 applied Generalised Gaussian quadrature rule to

1 1 −x 2 2 1 −x − y
2 2 2
xyz 0.02513677 evaluate the typical integrals The results obtained are
∫ ∫ ∫
10
dzdydx
0 0 0 x +y +z
2 2 2
15
0.02499524 in excellent agreement with the exact value.
0.02500001
1 12 − x 2 20
=∫ ∫ [ xy − xy x + y ]dydx
2 2

0 0
REFERENCES
= 0.02500000000
[1] C. Cattani, A. Paoluzzi, Boundary Integration over linear
5) 5 6.07503109 polyhedra. Comput-Aid. Des. 22(1990) pp. 130-135

IJSER
3 32 − x 2 32 − x 2 − y 2
xyz 6.08239570
∫ ∫ ∫ dzdydx 10
[2] C. Cattani, and A. Paoluzzi, Symbolic analysis of linear
0 0 0 x +y +z
2 2 2
15
6.07504415 polyhedra. Enrng Comput. 6 (1990) pp. 17-29

6.07499423 [3] F. Bernardini, Integrating of polynomials over n-dimensional

3 32 − x 2 20
polyhedra.Comput– Aid. Des.23(1991) pp. 51-58
=∫ ∫ [3xy − xy x + y ]dydx
2 2

0 0
= 6.075000000 [4] H.G.Timmer, J.M. Stern, Computation of global geometric
properties of solid objects. Comput.–Aid. Des. 12(1980) pp.
5 47.8608227 301-304
3 32 − x 2 32 − x 2 − y 2
6) ∫ ∫ ∫ ( x + y + z )dzdydx 10 47.6743991 [5] H.T. Rathod, K.V. Nagaraja, B. Venkatesudu, N.L.Ramesh,
0 0 0
Gauss Legendre Quadrature over a Triangle, J. Indian Inst.
47.7121540
15 Sci. 84 (2004) pp. 183-188
47.7120449
20 [6] H.T. Rathod, B. Venkatesudu, K.V. Nagaraja, Symmetric
Gauss Legendre quadrature formulas for composite
numerical integration over a tetrahedral surface, Journal of
Bulletin of Mathematics, Vol. 24, pp.51-79, (2006 )

3 32 − x 2 [x 9 − x 2 − y 2 [7] H.T. Rathod, K.V. Nagaraja, B. Venkatesudu, Numerical

=∫ ∫ integration of some functions over an arbitrary linear
0 0
+ y 9 − x2 − y2 tetrahedra in Euclidean three-dimensional space, Applied
Mathematics and Computation, Vol. 191, pp 397-409(2007)
-1 2 9
(x + y 2 ) + ]dydx
2 2 [8] Y.T.Lee, and A.A.G. Requicha, Algorithms for computing the
volume and other integral properties of solids I: known
= 47.71293843 methods and open issues. Commun. ACM 25(1982) pp.635-
641

[9] K.V. Nagaraja and H.T. Rathod , Symmetric Gauss Legendre

quadrature rules for numerical integration over an arbitrary
linear tetrahedra in euclidean three-Dimensional Space, Int.
Journal of Math. Analysis, Vol. 4, 2010, pp. 921-928

IJSER © 2013
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 4, Issue 12, December-2013 1534
ISSN 2229-5518

[10] O.C. Zienkiewicz, The Finite Element Method, McGraw Hill

London,3rd Edn.(1977)

[11] K.T. Shivaram, Generalised Gaussian Quadrature over a

circle, Int. Journal of research in Aeronautical and
Mechanical Engg. Vol 1.Issue 5,Sep2013, pp.19-24

[12] Kendallatkinson “ Numerical integration on the sphere

J. Austral. Math Soc (Series B) (1982), pp. 332-347

[13] Guergana Petrova “Cubature formulae for spheres, simplices

and balls” Journal of Computational and Applied
Mathematics 162 (2004) pp. 483–496

[14] Christopher A. Feuchter “Numerical Integration over a

sphere” Office of Research Analyses, Holloman AFB, New
mexico Jul 1967

IJSER

IJSER © 2013
http://www.ijser.org