Available online at www.sciencedirect.

com

Procedia Engineering 48 (2012) 295 – 303

MMaMS 2012

Verification of stress components determined by experimental methods

using Airy stress function
Mária Kenderováa*, František TrebuĖaa, Peter Frankovskýa
a
Technical University of Košice, Faculty of Mechanical Engineering, Letná 9, 042 00 Košice, Slovak republic

Abstract

The paper is dedicated to mathematical theory of elasticity and its relevant methods as applied for determination of plane stress state for
the purposes of verification of experimental values which were determined using PhotoStress® method as applied for chosen samples. The
paper further deals to a great extent with Airy biharmonic function as regards plane stress state in the samples with centrically placed
round holes. The introductory part of the paper includes basic algorithms for solving the problems of mathematical theory of elasticity
followed by descriptions of Airy biharmonic stress function in Cartesian and polar coordinates. In further parts of the paper were derived
equations for the calculation of radial, circumferential and shear stress in samples with centrically placed round holes. Calculations are
included in a single part of the paper together with input algorithms for calculations of radial, circumferential and shear stress. The
algorithms were performed using the syntax of programming language Matlab in MATLAB/Simulink. Subsequently, the results of radial
and circumferential stress calculations were depicted as 2D pictures similar to experimental samples with visualised stress functions.
These were later used for the verification of values determined through an experiment.
© 2012 The Authors. Published by Elsevier Ltd.
© 2012 Published by Elsevier Ltd.Selection and/or peer-review under responsibility of the Branch Office of Slovak Metallurgical Society at
Selection
Faculty and/or peer-review
of Metallurgy under
and Faculty responsibility
of Mechanical of the Branch
Engineering, Office
Technical of Slovak
University Metallurgical
of Košice Society
Open access at Faculty
under CC BY-NC-NDof license
Metallurgy
. and
Faculty of Mechanical Engineering, Technical University of Košice.

Keywords: Airy biharmonic function, mathematical theory of elasticity, radial stress, shear stress, circumferential stress

Nomenclature
g acceleration due to gravity (m/s2)
q distributed load (N/mm)
R hole radius (mm)
Greek symbols
ıx, ıy, ız normal stresses (MPa)
IJxy, IJxz, IJyz shear stress (MPa)
ır radial stress (MPa)
ıij circumferential stress (MPa)
IJrij shear stress in polar coordinate system (MPa)
ȝ Poisson's ratio (-)
ij angle parameter isoclinics (°)
ȡ body density (kg/m³)

* Corresponding author. Tel.: + 421 55 602 2456

E-mail address: maria.kenderova@tuke.sk.

1877-7058 © 2012 Published by Elsevier Ltd.Selection and/or peer-review under responsibility of the Branch Office of Slovak Metallurgical Society at
Faculty of Metallurgy and Faculty of Mechanical Engineering, Technical University of Košice Open access under CC BY-NC-ND license.
doi:10.1016/j.proeng.2012.09.517
296 Mária Kenderová et al. / Procedia Engineering 48 (2012) 295 – 303

Subscripts
x, y, z Cartesian coordinates (mm)
r, ij polar coordinate (mm), (°)

1. Mathematical theory of elasticity

The main objective of mathematical theory of elasticity is to specify states of stress and strains in the body subjected to
examination. This theory applies a complex mathematical formalism for description of solved problems which requires
knowledge of solving differential equations. However, there are problems which cannot be solved in an exact way even if
we applied the means of mathematics. By reason of this fact various forms of analogical solutions and experimental
investigation of the issue are applied, i.e. either through direct or indirect methods. For the description of a specific problem
mathematical theory of elasticity uses rather differential forms mainly because of the possibility to solve the problem in
more general way.
When solving a problem within mathematical theory of elasticity, following conditions need to be fulfilled [1]:
• the balanced state of any loose element,
• continuum of the whole body must be kept,
• bonds of a bonded body and
• physical relations for material deformation must be taken into consideration.

Today, there are three algorithms for solving problems in mathematical theory of elasticity [1]:
• solution in terms of displacement components,
• solution in terms of stress components,
• solution by means of stress functions.

1.1. Solving differential equations in terms of stress components

Solutions within the theory of elasticity in terms of stress components are based on equilibrium equations, conditions of
compatibility, or the Hooke’s Law. When we adjust the Hooke’s Law to the conditions of compatibility, we will get
compatibility equations expressed through stress components. These, together with equilibrium equations, make up six
equations through which we can specify six unknown stress components. Mathematical solution of such formulated problem
is in case of general bodies more complex.

1.2. Solving differential equations in terms of strain components

When solving a problem here, it is appropriate to express differential equations in terms of displacement components
(Lamé’s equations). Lamé’s equations can be derived by the incorporation of stress components expressed by means of
displacement components while considering the relation for relative change in volume. The form of Lamé’s equations after
these adjustments is as follows [1]:

1 ∂ε Y 1 ∂ε Z
∇2 v + ⋅ v + =0, ∇2 w + ⋅ v + = 0. (1)
1− 2 ⋅ μ ∂ y G 1− 2⋅ μ ∂ z G

1.3. Solving differential equations by means of Airy stress function

In our case we decided to make use of Airy stress function in order to solve the stress state. Stress function is defined as a
function which derivations allow us to specify all stress components in the body under examination while considering
specific relations. Maxwell in relation to triaxial stress state introduced three stress functions φ1 , φ2 , φ3 from which stress
components can be determined using following relations [1]:
∂ 2 φ3 ∂ 2 φ 2 ∂ 2 φ3
σx = + , τ xy = − ,
∂ y2 ∂ z2 ∂x∂y
∂ 2 φ1 ∂ 2 φ3 ∂ 2 φ1 (2)
σy = + , τ yz = − ,
∂ z 2 ∂ x2 ∂y ∂z
∂ 2 φ2 ∂ 2 φ1 ∂ 2 φ2
σz = + , τ zx = − .
∂ x2 ∂ y 2 ∂z ∂x
 Mária Kenderová et al. / Procedia Engineering 48 (2012) 295 – 303 297

Through the implementation of these equations into equilibrium equations in Cartesian coordinates while ignoring body
forces we will find out that these are relevant for random functions φi ( x, y, z ) . The solution presupposes that the condition
of compatibility is fulfilled for each of the chosen functions. Nevertheless, space solution and determination of the load the
tested body is exposed to is rather complicated. For determination of plane stress state only one stress function φ ( x, y ) is
required. This stress function is known as Airy stress function. Plane stress state components hence can be calculated using
following relations [1]:
∂2 φ ∂2 φ ∂2 φ
σx = , σy = , τ xy = − −C ⋅x . (3)
∂ y2 ∂ x2 ∂x∂y
The constant C within shear stress denotes self-weight body force on the Y-axis. Body force is a multiple of acceleration
due to gravity g and body density ρ . Equilibrium equations for plane stress state can be expressed in differential form while
considering body force on the Y-axis [1]:

∂ σ x ∂ τ xy ∂ τ yx ∂σ y
+ =0, + +Y = 0. (4)
∂x ∂y ∂x ∂y

When equations Eq. (3) are incorporated into the second equation Eq. (4), derives the following equilibrium:

(5)
∂3 φ ∂3 φ
− 2 −C + 2 +C = 0.
∂x ∂y ∂x ∂y
Through the incorporation of Hooke’s Law for plane stress state into Saint
Venant compatibility equation and through derivation of stress components from the equation Eq. (3) we obtain the
following differential equation [1]:

∂4 φ ∂4 φ ∂4 φ (6)
+ 2 ⋅ + =0.
∂ x4 ∂ x2 ∂ y 2 ∂ y 4

The solution of the equation Eq. (6) specifying Airy stress function are biharmonic functions. Hence, it can be concluded
that any Airy stress function φ ( x, y ) is a biharmonic function. Nevertheless, it is appropriate to use polar coordinates so that
the solution is not too complicated. The transformation into polar coordinates can be done using transformation equations as
shown in Fig. 1.

Fig. 1 Transformation of Cartesian coordinates into polar coordinates [1].

298 Mária Kenderová et al. / Procedia Engineering 48 (2012) 295 – 303

From Fig. 1 derive following transformation relations:

x = r ⋅ cos ϕ , r = x2 + y 2 ,
y (7)
y = r ⋅ sin ϕ , ϕ = arctg .
x

The coordinate system can be transformed only when the equation Eq. (6) is rewritten into the following form [1]:

§ ∂2 ∂2 · § ∂2 φ ∂2 φ ·
¨¨ 2 + ¸¸ ⋅ ¨¨ 2 + ¸¸ = 0 . (8)
©∂x ∂ y2 ¹ ©∂x ∂ y2 ¹

In further stages we performed first partial derivatives ∂φ / ∂x , ∂r / ∂x , ∂ϕ / ∂x and then second partial derivatives ∂ 2φ / ∂x 2 and
∂ φ / ∂y .
2 2

Airy biharmonic function Eq. (8) after transformation into polar coordinates has the following form [1]:

§ ∂2 1 ∂ 1 ∂2 · § ∂2 φ 1 ∂φ 1 ∂2 φ ·
¨¨ 2 + ⋅ + 2⋅ ¸¸ ⋅ ¨¨ 2 + ⋅ + ⋅ ¸=0. (9)
©∂r r ∂ r r ∂ϕ2 ¹ © ∂r r ∂ r r 2 ∂ ϕ 2 ¸¹

Radial stress σ r , circumferential stress σ ϕ and shear stress τ rϕ can be specified from relations determined for plane
stress stare on an oblique section [1]:

σ r = σ ξ = σ x ⋅ cos 2 ϕ + σ y ⋅ sin 2 ϕ − τ xy ⋅ sin 2ϕ ,

§ π·
σ ϕ = σ ξ ¨ϕ + ¸ = σ x ⋅ sin ϕ + σ y ⋅ cos ϕ + τ xy ⋅ sin 2ϕ ,
2 2 (10)
© 2¹
σ −σy
τ rϕ = τ ξη = x ⋅ sin 2ϕ + τ xy ⋅ cos 2ϕ .
2
Equations Eq. (3) are incorporated into equations Eq. (10) and after transformation into polar coordinate system while
ignoring body forces we get the equations for stress state determination derived from Airy stress function in polar
coordinates [1]:

1 ∂2 φ 1 ∂φ ∂2 φ ∂ § 1 ∂φ ·
σr = ⋅ + ⋅ , σϕ = , τ rϕ = − ¨ ⋅ ¸. (11)
r2 ∂ϕ 2 r ∂ r ∂ r2 ∂ r ¨© r ∂ ϕ ¸¹

In the following part of the paper we present the application of Airy stress function for calculation of plane stresses.

2. Application of Airy stress function in calculations of stress state in tested samples

The calculation of radial, circumferential and shear stress was done as applied for experimental samples with changing
diameter of a round hole and distributed load q. Airy function had the following form Eq. (12) and at the same time had to
fit for biharmonic equation derived in the previous part Eq. (12) [1].

§ r2
φ = q ⋅ ¨¨ −
R2
⋅ ln r −
(r 2 − R 2 )2 ⋅ cos 2ϕ ·¸ (12)
4⋅r2 ¸
© 4 2 ¹
 Mária Kenderová et al. / Procedia Engineering 48 (2012) 295 – 303 299

Fig. 2 Possible stresses as applied for the sample with a round hole while being stressed by distributed load.

As regards the case of distributed load of a sample with a round hole centrically placed on the sample, the components of
radial, circumferential and shear stress state were derived in the following form [2]:

q ª R2 § 3⋅ R4 R2 · º
σr = ⋅ «1 − 2 + ¨¨ 1 + − 4 ⋅ 2 ¸¸ ⋅ cos 2ϕ » ,
© ¹
4
2 ¬ r r r ¼
q ª R2 § 3⋅ R4 · º (13)
σϕ = ⋅ «1 + 2 − ¨¨ 1 + ¸¸ ⋅ cos 2ϕ » ,
© ¹
4
2 ¬ r r ¼
q § 2⋅ R2 3⋅ R4 ·
τ rϕ = ⋅ ¨¨ − 1 + − ¸¸ ⋅ sin 2ϕ .
2 © r2 r4 ¹

Within experimental testing, a force load was elicited similarly as in Fig. 2. External load was detected neither around the
round hole, nor along its circumferential region and therefore we had to consider following boundary conditions [1]:

σ r (R ) = 0 , τ rϕ (R ) = 0 . (14)

2.1. Characteristic features of the samples under examination

Virtual samples used for the purposes of stress determination while using Airy stress function exhibited the same features
as the real samples under experimental testing. Samples No. 1, 2, 3 were made from polycarbonate in the size as stated in
Table 1. One side of every sample was photoelastically coated. The coating was needed since we applied experimental
method of photoelasticity. The samples differed in the diameter of the hole which was placed in the centre of each sample.
Sample No. 4 was made from non-alloy structural fine grain steel. One side of the sample was coated with photoelastic
coating PS-1 with the thickness 3, 125 mm. The application of the coating made that particular side of the sample highly
sensitive and, hence, could be subjected to stress analysis both in elastic and elastic-plastic range. For mathematical
determination of radial, circumferential and shear stress were required only input parameters for the size of the samples,
holes and distributed load [2-3].

Stress state calculation and its visualisation for every single sample included following steps:
• Calculation of radial, circumferential and shear stress was done according to equations Eq. (13) as well as for three angle
(
variants ϕ = 90 $ ,45 $ ,0 $ . )
• In MATLAB 2009 was made a graphical model of particular samples through, on the one hand, projection of constant
functions which denoted sample edges and, on the other hand, projection of the function of the circle equation which
represented centrically placed hole. As a result we obtained contours of the real samples. Radii of the holes in each
sample were being gradually changed. In case of the hole in sample No. 4 we were changing its size parameters in line
with Table 1 too.
300 Mária Kenderová et al. / Procedia Engineering 48 (2012) 295 – 303

• The values of radial stress were recorded as arithmetic vectors for all angles under examination and the values of
circumferential stress were recorded only as one arithmetic vector. The amount of the values was determined by even
division of the distance between the hole and the edge of the sample.
• Designed arithmetic vectors were graphically projected as functions and, when referring to vectors with different angles,
equations for transformation from Cartesian into polar coordinated were applied.
• A coordinate grid was activated for the whole projected area in order to read the values from graphic representation
easier.

Table 1. Size parameters of the tested samples [2]

Size parameters of the Hole radius

Distributed load q
Tested sample sample R
(N/mm)
(mm) (mm)
Sample No.1 – Polycarbonate 125 x 340 x 0.9 15 16
Sample No.2 - Polycarbonate 125 x 340 x 0.9 20 16
Sample No.3 – Polycarbonate 125 x 340 x 0.9 25 16
Sample No. 4 – steel STN 11
125 x 221 x 8 20 12
523

2.2. Graphical outputs of stress calculations

(a) (b)

Fig. 3 Distributions of radial and circumferential stresses provided that (a) in the sample No. 1 the hole radius is R=15 mm and (b) in the sample No. 2
R=20 mm [1].
 Mária Kenderová et al. / Procedia Engineering 48 (2012) 295 – 303 301

(a) (b)

Fig. 4 Distributions of radial and circumferential stresses provided that (a) in the sample No. 3 the hole radius is R=25 mm and (b) in the sample No. 4
R=20 mm [2].

Fig. 3 and Fig. 4 depict distributions of radial and circumferential stresses. At 90° angle the stress initially has an
increasing tendency followed by a decline nearing zero values. Fig. 3 (a) shows that the radial stress at 90° angle is after
decline nearing zero value fastest when compared to other samples. Sample No. 3 depicted in Fig. 4(a) exhibits the slowest
decline of radius stress at 90°. As shown in Fig. 3 and Fig. 4, radial stress at 45° and 0° starts with a relatively quick
increase while when reaching higher values r the increase slows down. As we can observe on sample No. 1 depicted in Fig.
3 (a), the highest values of radial stresses are reached at 0° angle. In Fig. 3 and Fig. 4 the hatched region at 90° angle in the
negative direction of the coordinate system was used for the projection of circumferential stress. In case of sample No. 1 up
to sample No. 3 are the values at the edge of the holes identical. Verification of the values relevant for circumferential
stresses matched the relation 3q for all four samples as it is depicted in Fig. 3 and Fig. 4 as a dimensioned region of the
circumferential stress at the hole boundary. In case of sample No. 1 is the decline of circumferential stress less regular and
steeper than for instance in case of sample No. 3.

3. Experimental determination of strain and stress fields by PhotoStress method

A plate with a round hole in its centre was subjected to diametral tensile load in a load frame. The tension was introduced
through lever mechanism and in order to read load values we used dynamometer 6418 with measuring range up to 1000N.
During minimum loading of the sample we noticed and took pictures of dark fringes known as isoclinics. Isoclinic fringes
denote regions where principal strains and principal normal stresses exhibit identical directions. Observing the isoclinics,
axes of the polarizer and the analyzer were parallel to each other. When rotating the analyzer and the polarizer in the range
between 0° and 90°, we get a complex overview of the distribution of isoclinics along the whole surface subjected to testing.
Fig. 5 depicts isoclinic fringes derived under angle parameter from 0° to 90° considering the increment up to 10°. With
reference to Fig. 5 it can be concluded that isoclinics gained under angle parameter 90° are identical [3-8].

(a) (b) (c) (d) (e)

302 Mária Kenderová et al. / Procedia Engineering 48 (2012) 295 – 303

(f) (g) (h) (i) (j)

Fig. 5 Isoclinic fringes on the plate with a round hole in its centre under different angle parameters (a) 0° (b) 10° (c) 20° (d) 30° (e) 40° (f) 50° (g) 60° (h)
70° (i) 80° (j) 90° [2].

Isochromatic fringes were observed on a loaded photoelastically coated sample during gradual loading of a plate with a
round hole placed in its centre. Axes of the polarizer and analyzer were during this process parallel to each other.
Isochromatic fringes are lines of constant principal normal stress or principal strain difference.
The sequence of a principal fringe is specified by means of a compensator on reflection polariscope. The plate was
subjected to diametral compression so as to determine the isochromatic fringe order. When analysing isochromatic fringes
along the whole surface of the plate it is necessary to know the relation between colourful fringes and stress or strain. The
analysis of isochromatic fringes was done using reflection polariscope LF/Z-2 with polarizer and analyzer being parallel to
each other.

(a) (b) (c) (d) (e)

(f) (g) (h) (i)

Fig. 6 Isochromatic fringes during gradual loading of the plate with a hole in its centre [2].

4. Conclusion

The results of radial, circumferential and shear stress calculations provide us with details on the behaviour of the sample
subjected to distributed load. A characteristic feature of radial stress in the edge region of the hole is that it equals zero even
at all required changes of the angle. As the distance of the stress under examination from the hole increases we
simultaneously observe the increase of radial stress. Radial stress at 90° angle reaching zero would imply extreme size
parameters of the sample or a very small size of the hole. However, for practical reasons such sample could not possibly be
used. On the contrary, radial stress at 0° angle is the higher the smaller the hole in the sample. This statement is similarly
true for radial stress at 45°. As for circumferential stress applied to the samples, we can conclude that the change in the
radius of the hole does not affect the magnitude of circumferential stress in the edge region of the hole. In this sense the
influence is observed in values recorded in further distance from the hole, though the difference in values is small.
Observing the hatched regions of the circumferential stress we can conclude that higher magnitudes noticed during the
decline occur when the radius of the hole is bigger even in case of samples with the same distributed load.
 Mária Kenderová et al. / Procedia Engineering 48 (2012) 295 – 303 303

Acknowledgements

This contribution is the result of the project implementation “Centre for the research of control of technical,
environmental and human risks for permanent development of production and products in mechanical engineering” (ITMS:
26220120060) supported by the Research & Development Operational Programme funded by the European Regional
Development Fund and with support from the project VEGA initiated by the Slovak Ministry of Education, project No.
1/0937/12 entitled “Development of non-traditional experimental methods for mechanical and mechatronic systems”.

References
[1] TrebuĖa, F., Šimþák, F., 2005. Elasticity, strength and plasticity in engineering (in Slovak), Emilena, Košice, Slovak Republic.
[2] Kenderová, M., 2012. Využitie polykarbonátových materiálov pri vývoji experimentálnych interferenþných metód umožĖujúcich urþenie polí napätí a
deformácií na automobilových plošných súþiastkach. Písomná práca k dizertaþnej skúške. Technical university of Košice, Košice, Slovak Republic.
[3] TrebuĖa, F., 2006. Princípy, postupy, prístroje v metóde PhotoStress, TypoPress, Košice, Slovak Republic.
[4] TrebuĖa, F., Jadlovský, J., Frankovský, P., Bakšiová, Z., Kostelníková, A., 2010. Further Possibilities of Using Software PhotoStress for Separation of
Principal Normal Stresses, Acta Mechanica Slovaca. Vol. 14, No.2, pp. 82-91.
[5] TrebuĖa, F., Frankovský, P., Bocko, J. Pástor, M., 2011. New Possibilities of using PhotoStress® Method, Acta Mechanica Slovaca. Vol. 15, No.4,
pp. 44-50.
[6] TrebuĖa, F., Frankovský, P., HuĖady, R., 2011. Optical methods and their application in experimental analysis of mechanical and mechatronic systems
Metallurgical Journal, Vol. 64, no. 7, pp. 173-178.
[7] Ramesh, K., 2000. Digital Photoelasticity - Advanced Techniques and Applications, Springer-Verlag, Berlin, Germany.
[8] Macura, P., 2010. Experimental Stress Analysis of Transducers by Means of PhotoStress Method, Acta Mechanica Slovaca. Vol. 14, No. 4, pp. 52-57.