lineno,hyperref [2] [T1]fontenc graphicx epstopdf amssymb here parskip tikz amsmath multirow

epsfig [left=2cm,right=2cm,top=2cm,bottom=2cm]geometry
lineno

1 Introduction

The problems’ category which examines the state of stress and displacements in an elastic layer
medium, due to a torsional of a disc in bonded contact, has been a subject of much interest
in mechanics and applied mathematics. The problem on the torsional of elastic half space was
considered, at first, by Reissner and Sagoci [1]. They studied the static interaction of a rigid disc
and an elastic isotropic half-spaces for which they obtained the solution by means of the spheroidal
coordinates. The same problem was solved by Sneddon [2] by a different method using the Hankel
transforms method for reduction the problem to a pair of dual integral equations. Collins [3] treated
the torsional problem of an elastic half-space by supposing the displacement at any point in the
half-space to be due to a distribution of wave sources over the part of the free surface in contact
with the disk. The solution of the forced vibration problem of elastic layer of finite thickness when
the lower face is either stress free or rigidly clamped was given by Gladwell [4] who reduced the
mixed boundary value problem to a Fredholm integral equation by Noble’s method [5]. Pak and
Saphores [6] provided an analytical formulation for the general torsional problem of a rigid disc
embedded in an isotropic half-space. The quadrature numerical was used for solving the obtained
Fredholm integral equation.
The torsional of elastic layers with a penny shaped crack was considered by some researchers. Sih
and Chen [7] studied the problem of a penny-shaped crack in layered composite under a uniform
torsional stress. The displacement and stress fields throughout the composite were obtained by
solving a standard Fredholm integral equation of the second kind. Low [8] investigated a problem
of the effects of embedded flaws in the form of an inclusion or a crack in an elastic half space
subjected to torsional deformations. The corresponding Fredholm integral equations were solved
numerically by quadrature approach. The same method was used by Dhawan [9] for solving the
problem of a rigid disc attached to an elastic half-space with an internal crack.
Recently, Cai and Zue [10] discussed the torsional vibration of a rigid disc bonded to a poro -
elastic multilayered. They used the Hankel transforms and transferring matrix method. Yu [11]
studied the forced torsional oscillations inside the multilayered solid. The elastodynamic Green’s
function of the center of rotation and a point load method were used to solve the problem. Pal and
Mandal [12] considered the forced torsional oscillations of a transversely isotropic elastic half space
under the action of an inside rigid disc. The studied problem was transformed to dual integral
equations system. Which was reduced to a fredholm integral equation. A similar problem with
the rocking rotation was solved later on by Ahmadi and Eskandari [13]. They used an appropiate
Green’s function to write the mixed boundary-value problems posed as a dual integral equation.
In this paper, we investigate the problem of a penny shaped crack in the interior of a homoge-
neous elastic medium under an axisymmetric torsion applied to a rigid disc glued inside. With the
aid of the Hankel integral transformation method. The mixed boundary-value problems are written
as a system dual of integral equations. The corresponding system of fredholm integral equations
was approached by sets of linear equations. After getting the unknown coefficients of this system
we obtain numerical results and display curves according to certain pertinent parameters.

2 Formulation of the problem

We consider the axisymmetric torsion of a rigid coaxial disc of a radius b situated on plane z = h
in an infinite, isotropic and homogeneous elastic medium, containing a penny shaped crack in the
region 0 < r < a, z = 0 . The faces of the crack are supposed stress free while the dic rotates with
an angle ω about the z axis passing through their centers as shown in fig.1.
As the problem is an axisymmetrical one, the displacements and stresses are a function only of
(r, z). The Equation of equilibrium is given as follows ∂ 2 uθ ∂u uθ ∂ 2 uθ
The only non-zero
∂r2 + r∂rθ − + =0
r2 ∂z 2
components stresses are τθz = G ∂u
∂z ,
θ ∂
τθr = Gr ∂r uθ

r where G is the shear modulus of the
material. By means of the Hankel’s transformation integral and its inverse given by [14]
Z ∞
F (λ, z) = f (r, z)rJ1 (λr)dr (1)
0

and
Z ∞
f (r, z) = F (λ, z)λJ1 (λr)dλ (2)
0

where J1 is the Bessel function of the first kind of order one. The general solution of Eq.(1) for the
regions I(z ≤ 0), II(0 ≤ z ≤ h) and III(z ≥ h) as shown in fig.1 is expressed as
Z ∞h i
(i)
uθ (r, z) = Ai (λ)e−λz + Bi (λ)eλz J1 (λr)dλ i = 1, 2, 3 (3)
0

where Ai and Bi are unknown functions.

At infinity, the regularity conditions are given by

lim uθ (r, z) = 0, lim τθz (r, z) = 0 (4)

r,z→∞ r,z→∞

2
 (2) (1)
The boundary conditions of the problem are τθz (r, 0+ ) = τθz (r, 0− ) = 0 r < a
(3) (2)
uθ (r, h) = uθ (r, h) = ωr r ≤ b The continuity conditions of the problem in the planes z=0 and
(2) (1)
z=h can be written as uθ (r, 0+ ) − uθ (r, 0− ) = 0 r ≥ a
(2) (1)
τθz (r, 0+ ) − τθz (r, 0− ) = 0 r≥a
(3) (2)
uθ (r, h+ ) − uθ (r, h− ) = 0 r > b
(3) (2)
τθz (r, h+ ) − τθz (r, h− ) = 0 r > b By utilizing the condition Eq.(4), the expressions of displace-
(1) R∞
ments and stresses in the three regions take the following forms uθ (r, z) = 0 B1 (λ)eλz J1 (λr)dλ
(1) R∞ λz
τθz (r, z) = G 0 λB1 (λ)e J1 (λr)dλ
(2) R∞h i
uθ (r, z) = 0 A2 (λ)e−λz + B2 (λ)eλz J1 (λr)dλ
h i
(2)
G 0∞ λ −A2 (λ)e−λz + B2 (λ)eλz J1 (λr)dλ
R
τθz (r, z) =
(3) R∞ −λz J (λr)dλ
uθ (r, z) = 0 A3 (λ)e 1
(3) R∞
τθz (r, z) = −G 0 λA3 (λ)e−λz J1 (λr)dλ The unknown functions B1 (λ), A2 (λ), B2 (λ) and A3 (λ)
can be determined from the boundary and continuity conditions.
(2)
The boundary and continuity conditions Eqs. (2), (2), (2) and (2) show that τθz (r, 0+ ) −
(1)
τθz (r, 0− ) = 0 r ≥ 0
(3) (2)
uθ (r, h+ ) − uθ (r, h− ) = 0 r ≥ 0 The continuity conditions Eqs.(2) and (2)lead to B1 (λ) =
B2 (λ) − A2 (λ)
A3 (λ) = B2 (λ)e2λh + A2 (λ) From the mixed boundary conditions Eqs.(2), (2), (2 and (2),
we find the system of dual integral equations for obtained the unknown functions A2 and B2
R∞
0 λ[B2 (λ) − A2 (λ)]J1 (λr)dλ = 0, 0≤r<a
R∞
0 A2 (λ)J1 (λr)dλ = 0, r≥a
R∞ −λh + B (λ)(eλh )]J (λr)dλ
0 [A2 (λ)e 2 1 = ωr, 0≤r≤b
R∞ λh
0 λB2 (λ)e J1 (λr)dλ = 0. r>b

3 Reduction of the problem to a system of Fredholm integral

equations

The system of dual equations can be reduced to a system of Fredholm integral equations of second
√ R √
kind by introducing the auxiliary functions φ(t) and ψ(t) such that A2 (λ) = λ 0a tφ(t)J 3 (λt)dt
2
√ R √
B2 (λ) = e−λh λ 0b tψ(t)J 1 (λt)dt
2

With this choice of the new unknown functions, we find that the homogeneous equations Eq.(2)
and Eq.(2) are identically satisfied while equations Eq.(2) and Eq.(2) lead to the Fredholm’s inte-
gral equations.
By inserting A2 (λ) and B2 (λ) in the equations Eq.(2) and Eq.(2), we get

3
Ra√ 3
tφ(t)dt 0∞ λ 2 J 3 (λt)J1 (λr)dλ −
R
0 2
Rb√ R ∞ 3 −λh Ra√ R∞√
0 tψ(t)dt 0 λ 2 e J 1 (λt)J1 (λr)dλ = 0, r<a 0 tφ(t)dt 0 λe−λh J 3 (λt)J1 (λr)dλ +
2 2
Rb√ R∞√
0 tψ(t)dt 0 λJ 1 (λt)J1 (λr)dλ = ωr, r < b To find the first Fredholm integral equation,
2

we use λJ1 (λr) = 1r2 ddr[r2 J2 (λr)].

Taking into account the integral formula
Z ∞√
√ 3 p
λJ 3 (λt)J2 (λr)dλ = { 2πt 2 r2 r2 − t2 t<r 0 t>r (5)
0 2

we obtain Abel equation corresponding to equation Eq.(3)

Z b√ Z ∞√
t2 φ(t)
r Z r
2
√ dt − r2 tψ(t)dt λe−λh J 1 (λt)J2 (λr)dλ = 0 r<a (6)
π 0 r 2 − t2 0 0 2

By applying the Abel’s transform formula

Z r Z t
f (t) d rg(r)
√ dt = g(r) then f (t) = 2π √ dr (7)
0 r2 − t2 dt 0 t2 − r2
we then find from Eq.(6) that t
q hR √ R∞√ i
2 d R t √ r3 b
2 φ(t) = π dt 0 2 t −r2 0 δψ(δ)dδ 0 λe−λh J 1 (λδ)J2 (λr)dλ dr r < a For the right hand
2

side of the above equation, the integral is further simplified by using the following relationship
√ d
Z t
r3 √ 5
2π √ J2 (λr)dr = λt 2 J 3 (λt) (8)
dt 0 t2 − r2 2

we obtain the first Fredholm integral equation of second kind

√ Z b√
φ(t) + t δψ(δ)K(t, δ)dδ = 0, r<a (9)
0

where
Z ∞
K(t, δ) = − λe−λh J 3 (λt)J 1 (λδ)dλ (10)
0 2 2

Similarly, Eq.(3) can be reduced to the second Fredholm integral equation as follows
By using the formula
Z ∞√
√ p
λJ 1 (λt)J1 (λr)dλ = { 2tπ1r r2 − t2 t<r 0 t>r (11)
0 2

we obtain the following Abel equation

r Z r Z a√ Z ∞√
2 tψ(t)
1r √ dt + tφ(t)dt λe−λh J 3 (λt)J1 (λr)dλ = ωr r<b (12)
π 0 r2 − t2 0 0 2

By applying the Abel’s transform formula to the last equation, we obtain tψ(t) =
√ d R t r2 h R √ R √ i
2π dt 0 √t2 −r2 ωr − 0a δφ(δ)dδ 0∞ λe−λh J 3 (λδ)J1 (λr)dλ dr r < b Using the following re-
2

lationship

r3
Z t
d
√ dr = 2t2 (13)
dt 0 t2 − r 2

4
 √ d
Z t 2
r J1 (λr) √
2π √ dr = t λtJ 1 (λt) (14)
dt 0 t2 − r2 2

we finally get the second Fredholm integral equation of second kind

√ Z a√ 4ω
ψ(t) + t δφ(δ)L(t, δ)dδ = √ t, 0<t<b (15)
0 2π
with the kernel
Z ∞
L(t, δ) = λe−λh J 1 (λt)J 3 (λδ)dλ (16)
0 2 2

The system (9) and (15) can be written in the dimensionless form as follows
By putting




 δ = as,


0 < δ < a; t = au







0<t<a





 δ = bs,






 0 < δ < b; t = bu


 0<t<b


(17)
√
Next, we multiply the above two equations of the system, respectively by 2π4aωφ(au) and
√
2π4bωψ(bu) and using the following substitutions
 √ √
2π 2π
Φ(u) = 4aω φ(au) Ψ(u) = 4bω ψ(bu)



 c = ba λ = xa H = ha
(18)
√ √
2π 2π
Φ(u) = 4aω φ(au) Ψ(u) = 4bω ψ(bu)
√ √ R √
c = baλ = xa H = ha we obtain Φ(u) + c2 c u 01 sΨ(s)K(u, s)ds = 0, u < 1
q √ R √
Ψ(u) + 1c u 01 sΦ(s)L(u, s)ds = u, u < 1 where K( u, s) =- 0∞ xe−xH J 3 (xu)J 1 (xcs)dx
R
2 2
R∞
L(u, s) = 0 xe−xH J 1 (xcu)J 3 (xs)dx
2 2

4 Numerical Results and Discussion

The midpoint rule is used to find the numerical solution for the system given by Eq.(3) and Eq.(3).
Dividing the interval [0, 1] into N equal sub-intervals, so the midpoints are

u = um = 2m − 1N s = un = 2n − 1N m, n = 1, 2...N

and introducing the following notations Φ(um ) = Φm Ψ(um ) = Ψm

K(um , un ) = Kmn L(um , un ) = Lmn we obtain the following systems of infinite algebraic equations

5
 √
c2 c √ √
um N
P
Φm + N n=1 un Ψn Kmn = 0, m = 1, 2...N
√ √
1√
um N
P
Ψm + N c n=1 un Φn Lmn = um , m = 1, 2...N where
R∞
Kmn (c, H) = − 0 xe−xH J 3 (xum )J 1 (xcun )dx
2 2
R∞
Lmn (c, H) = 0 xe−xH J 1 (xcum )J 3 (xun )dx Next, we evaluate numerically the infinite integral K
2 2

and L given by Eq.(4) and Eq.(4) by simpson rule.

After solving the above system, the unknown coeffcients can be obtained then we ge the numerical
approximation of the unknown functions B1 , A2 , B2 and A3 given by Eq.(2), Eq.(2), Eq.(3) and
N
4a2 ω √ X √ √
Eq.(3) B1 (x) = √ x um [e−xH c2 cΨm J 1 (xcum ) − Φm J 3 (xum )]
N 2π m=1
2 2

2 N
4a ω √ X √
A2 (x) = √ x um Φm J 3 (xum )
N 2π m=1
2
√ N
4b2 cω √ X √
B2 (x) = e−xH √ x um Ψm J 1 (xcum )
N 2π m=1
2

N
4a2 ω √ X √ √
A3 (x) = √ x um [exH c2 cΨm J 1 (xcum ) + Φm J 3 (xum )] (18)
N 2π m=1
2 2

4.1 Stress intensity factor

The stress intensity factor at the edge of the crack is defined by

(2)
KaIII = limr→a+
p
2π(r − a)τθz (r, z)|z=0 On the plane z = 0 for r ≥ a, the expression of stress
is given by
Z ∞" Z a√ Z b√ #
(2) 3 3
−λh
τθz (r, 0) =G −λ 2 tφ(t)J 3 (λt)dt + e λ 2 tψ(t)J 1 (λt)dt J1 (λr)dλ (19)
0 0 2 0 2

Using the relation

J1 (λr) = −1λddrJ0 (λr) (20)

we obtain
Z a√ Z ∞ Z b√ Z ∞
(2) 1 3
τθz (r, 0) =G tφ(t)dt λ J 3 (λt)J0 (λr)dλ+G
2 tψ(t)dt e−λh λ 2 J 1 (λt)]J1 (λr)dλ(21)
0 0 2 0 0 2

We use the following asymptotic behavior of the Bessel function of the first kind, for large values
of λ
√
Jν (λ) 2λπ cos(λ − π2ν − π4) (22)
√ √
then we get, J3/2 (λt) 2λtπ cos(λt − π) = − 2λtπ cos(λt)
Next, we use the following integral formula to replace the infinite integral in the Eq.(21)
Z ∞
cos(λt)J0 (λr)dλ = {
0
(23) we obtain
(2) √ d a φ(t)
Z Z b√ Z ∞
3
τθz (r, 0) = − 2πG √ dt + G tψ(t)dt e−λh λ 2 J 1 (λt)]J1 (λr)dλ (24)
dr 0 r 2 − t2 0 0 2

6
Now integrating by parts, we get
√ Z b√
tφ0 (t)
Z a Z ∞
(2) G 2 p 3
τθz (r, 0) = √ [aφ(a)r r2 − a2 − √ dt]+G tψ(t)dt e−λh (λ) 2 J 1 (λt)]J1 (λr)dλ(25)
π 0 r r 2 − t2 0 0 2

The stress intensity factor at r = a, may be calculated as

√
a
q G 2 p
KIII = lim 2π(r − a) √ (aφ(a)r r2 − a2 ) (26)
r→a+ π
By using the following transformations
√
φ(a) = 4aω 2πΦN , ρ = ra

we obtain

a √ √
KIII = 4Gω a πΦN (27)

a
Fig.2 presents the results of the effect of the normalized disc size on the stress intensity factor KIII
for different disc locations H = 0.5, 1, 1.5 and 2. It is observed that the values of the stress intensity
factor versus c increase, attain its maximum values and decrease.
The effect of the axial distance between the crack and the disc H on the stress intensity factor is
also shown in Fig.9. The increase of the height H induces the decrease of stress intensity factor for
all the values of parameter c.

4.2 Displacement and stress fields

By substituting the Eqs.(4)-(4) into the expressions of the displacements and the stresses Eqs.(2)-
(2), we get the numerical results of displacements and stresses for the three regions.
(i)
The results for the variation of the normalized displacement uθ (ρ, ξ)/ωa and stress τ (i) (ρ, ξ)/Gωa
with ρ = r/a are shown graphically in Fig.3 to Fig.8 for the different value of the fixed parameters
Z = z/a, H = h/a and c = b/a.
In figure Fig.3 to Fig.5, the normalized displacements are shown for different normalized axial
distances ξ = 0, H/4, H/2, 3H/4,c = 0.5, 1 and H = 1. We notice that the displacements in the
three regions increase at first, reach maximum values at ρ = c in region 2 and 3 and then decrease
out of the disc band with increasing ρ.

7
 Fig.6 to Fig.8 show the variations of the normalized stress versus normalized radius ρ = for
different normalized axial distances ξ with fixed parameters H and c. It is concluded from Fig.6
that the magnitude of the stress in the first region is the lowest for the three others and that the
stress are initially rises, attains its maximum values and with the increase in the value of ρ the
stress go on decreasing. In the regions II and III, the variation of the corrsesponding stresses
τ (2) (ρ, ξ)/Gωa and τ (3) (ρ, ξ)/Gωa are shown in Fig.7 and Fig.8. It can be inferred that the stresses
increase up to ρ = c, then it decreases and will die out.

8
5 Conclusion

In this study, the axisymmetric torsion problem of a rigid disc embedded in the interior of a
homogeneous elastic material is analytically addressed. The medium is weakened by a penny-
shaped crack located parallel to the disc. Using the Hankel integral transformation method, the
doubly mixed boundary value problem is reduced to a system of dual integral equations, which
are reduced, to a Fredholm integral equations system of the second kind. The presented graphs
show the variation of the displacements, the stresses in the three regions and the stress intensity
factor at the edge of crack for some dimensionless parameters. The numerical results show that
the discontinuities around the crack and the inclusion cause a large increase in the stresses which
decay with distance from the disc-loaded. Furthermore, it can be seen the dependence of the stress
intensity factor on the disc size and the distance between the crack and the disc.

References

References

[1] Reissner, E; Sagoci, H.F. Forced torsion oscillation of an half-space I, Int. J. Appl. Phys., 15
(1944), 652 - 654.

[2] Sneddon, I.N. Note on a boundary value problem of Reissner and Sagoci, Int. J. Appl. Phys.,
18 (1947), 130 - 132.

[3] Collins, W.D. The forced torsional oscillations of an elastic halfspace and an elastic stratum,
Pro. London. Math.Society., 12 (1962), 226 - 244.

[4] Gladwell, G.M.L.The forced torsional vibration of an elastic stratum , Int. J. Eng. Sci.,
7(1969), 1011 - 1024.

[5] Noble, B. The solution of Bessel function dual integral equations by a multiplying factor
method, Pro. Camb. Eng. Sci., 59 (1963), 351 - 362.

[6] Pak, R. Y. S; Saphores, J.D. M. Torsion of a rigid disc in a half-space, Int. J. Engng Sci., 29
(1991), 1 - 12.

[7] Sih, G. C; Chen, E. P. Torsion of a laminar composite debonded over a penny-shaped area, J.
Franklin Inst., 293 (1972), 251 - 261.

[8] Low,R. D. On the torsion of elastic half space with embedded penny-shaped flaws, J. Appl.
Mech., 39 (1972), 786 -790.

9
 [width=10cm, height=8cm]1

Figure 1: Geometry and coordinate system

[scale=0.65]KIIIA

a
Figure 2: Variation of normalized stress intensity factor KIII

[width=8cm, height=6cm]H1C1.eps [width=8cm, height=6cm]H1C05.eps

[width=8cm, height=6cm]U2H1C1.eps [width=8cm, height=6cm]U2H1C05.eps

[width=8cm, height=6cm]UH1C1.eps [width=8cm, height=6cm]UH1C05.eps

[width=8cm, height=6cm]TH1C1.eps [width=8cm, height=6cm]TH1C05.eps

[width=8cm, height=6cm]T2H1C1.eps [width=8cm, height=6cm]T2H1C05.eps

[width=8cm, height=6cm]T3H1C1.eps [width=8cm, height=6cm]T3H1C05.eps

10
 [9] Dhawan, G. K.On the torsion of elastic half-space with penny-shaped crack, Defense. Sci. J.,
24 (1974), 15 - 22.

[10] Wang, G. C ; Chen .L . Torsional oscillations of a rigid disc bonded to multilayered poroelastic
medium, J. Zhejiang University.Sci A., 6(3) (2005), 213-221.

[11] Yu, H.Y; Forced torsional oscillations of multilayered solids,Int.J.Eng.Sci., 46 (2008), 250-259.

[12] Pal, P.C; Mandal.D; Sen.B. Torsional Oscillations of a Rigid Disc Embedded in a Transversely
Isotropic Elastic Half-Space,Adv. Theor. Appl. Mech., 4 (2011), 177-188.

[13] Ahmadi, S.F; Eskandari. M. Rocking rotation of a rigid disk embedded in a transversely
isotropic half-space ,Civil Eng. Infra. J., 47 (2014), 125-138.

[14] Debnath, L; and Bhatta. Integral transforms and their applications,Seventh edition, Chapman
Hall ., /CRC, (2007), 703.

11