Efficient Evaluation

of Diffuse View
Factors for
Radiation

ABSTRACT
This paper presents a numerical method of evaluating view factors between
planar surfaces.
The method, which is based on Gaussian quadrature to perform the contour
integration is extended to surfaces with curved boundaries.
As an application, the shape factor between two elliptic surfaces has been
evaluated.
The following discussion will consider only the case of radiant interchange
between diffusely reflecting and absorbing surfaces with no absorption in the
intervening air.

BACKGROUND
Thermal radiation is typically as important as
convection or conduction in the overall heat balance in
buildings.
Whenever we talk about net radiation energy exchange
between surfaces, It starts with the calculation of view
factors.
The problem with view factors is not that they are
inherently difficult to compute, but that the calculation

 Radiation heat transfer between

surfaces
depends
on
the
orientation of the surfaces relative
to each other as well as their
radiation
properties
and
temperatures, as illustrated in Fig.

VIEW FACTOR/SHAPE
FACTOR/CONFIGURATION FACTOR
= The fraction of the radiation
leaving surface i that strikes surface
j directly.

DIFFUSE ASSUMPTION

The view factor based on the

assumption that the surfaces are
diffuse emitters and diffuse reflectors
is called the diffuse view factor.

And the view factor based on the

assumption that the surfaces are
diffuse
emitters
but
specular
reflectors is called the specular view
factor.

 Consider a situation involving N surfaces. Since each surface may

potentially interact with every other surface, there are N2
interactions, or view factors.
Even simplifications, such as the reciprocity relation and the fact that
a flat surface cannot view itself, reduce the number of view factors
only to N(N-1)/2, which is still of order (N2)

If it is known that some, but not which, surfaces can obstruct (or
shade or occlude) the views between surface pairs, it is necessary
to check N-2 surfaces as possible obstructing surfaces for each view
factor. This gives N(N-1)(N-2)/2 obstruction checks, that is, O(N 3).

INTRODUCTION

Since analytical solution is not possible for many of the

geometries of practical interest, accurate determination
of view factors by numerical means has been a topic of
research.

NOMENCLATURE

CONTOUR INTEGRATION METHOD

STOKES THEOREM

 Three methods which are considered in this paper to

evaluate the integrals in the equation developed above
are

Trapezoidal,
Simpson and
Gaussian Quadrature

TRAPEZOIDAL METHOD ( TZM)

simplest of all the quadrature formulas
needs two points on each contour element.

SIMPSON'S METHOD (SIM)

The function to be: integrated is approximated by a
quadratic function through three points within the limits
of integration

GAUSSIAN QUADRATURE
Gaussian quadrature is the most accurate method for a
given number of points on the elements.
An n point quadrature formula approximates
function by (2n- 1)th degree polynomial.

the

A one-point formula (GAUSS1) evaluates the integral as

accurate as the TZM with just one point on each
element.

 A two point formula (GAUSS2) approximates the

function with a polynomial of degree three and a threepoint formula (GAUSS3) with degree five.
To get the same accuracy, the SIM needs four and six
points, respectively. In general, for a desired accuracy,
SIM needs four times more computation time compared
to the Gaussian method.

GAUSSIAN QUADRATURE WITH

NONLINEAR TRANSFORMATION
The Gaussian quadrature needs the integration limits to be
between - 1 and + 1 and hence demands a transformation of
global coordinates to local ones.
Referring to Fig. 1, the global coordinates of the points on the two
contours are transformed using the following :

RESULTS AND DISCUSSION

COMPARING RESULTS

CONCLUSIONS
The present method, is capable of computing view
factors for plane surfaces which are located arbitrarily,
straight or curved and share common edges practically
with any desired accuracy. The method is
computationally very efficient.
Use of higher order quadrature is recommended
when the surfaces share a common edge or if the
surfaces are touching anywhere.

 Use of higher order transformation is recommended

when the surfaces have curved contours.

It is found that the computational effort will be less

when the number of elements is increased by quadratic
transformation than when the order of transformation
increased.

REFERRENCES

THANK YOU