The fractional mathematical model of Maxwell’s equations in an electromagnetic field and the fractional generalized thermoelastic theory associated with two relaxation times are applied to a 1D problem for a thick plate. Laplace transform is used. The solution in Laplace transform domain has been obtained using a direct method and its inversion is calculated numerically using a method based on Fourier series expansion technique. Finally, the effects of the two fractional parameters (thermo and magneto) on variable fields distributions are made. Numerical results are represented graphically.
1. Introduction
Fractional calculus (FC) is a very useful tool in describing the evolution of systems with memory, which typically are dissipative and complex systems such as glasses, biopolymers, biological cells, porous materials, amorphous semiconductors, and liquid crystals. Scaling laws and self-similar behavior are supposed to be fundamental features of complex systems. In recent decades the FC and in particular the fractional differential equations have attracted interest of researchers in several areas including mathematics, physics, chemistry, biology, engineering, and economics [1–4].
FC theory has been used successfully in thermoelasticity and thermoviscoelasticity, such that a quasi-static uncoupled theory of thermoelasticity based on the fractional heat-conduction equation was put forward by Povstenko [5]. The theory of thermal stresses based on the heat-conduction equation with the Caputo time-fractional derivative is used by Povstenko [6] to investigate thermal stresses in an infinite body with a circular cylindrical hole. Sherief et al. [7] introduced a fractional order theory of thermoelasticity. Raslan [8, 9] has solved 1D problems in the context of this theory and applied this theory to 2D problem of thick plate [10]. The fractional parameter effect of this theory on thermoelastic material with variable thermal variable thermal material properties has been studied in [11, 12]. Ezzat [13] established a model of fractional heat-conduction equation by using the Taylor series expansion of time-fractional order developed by Jumarie [14].
Hamza et al. established a new mathematical model of Maxwell’s equations in an electromagnetic field in [15] and derived a fractional model for thermoelasticity associated with two relaxation times in [16]. A model for unsteady thermoelectric magnetohydrodynamics (TEMHD) flow and heat transfer of two immiscible second-grade fluids with two fractional parameters was introduced in [17].
The previous work introduced by Hamza et al. [18] describes one-dimensional problems in the context of the theory [16] in which the electromagnetic field effects are ignored. The model in the previous work depends only upon one fractional parameter α which does not have electric or magnetic affects. Another application of this theory introduced in [16] depends on two fractional parameters α and β. The first fractional parameter α appears only in the heat equation and is absent from both the equation of motion and the constitutive equations. In this work, we solve a 1D problem for a thick plate that was not solved in the context of both theories [15, 16].
The new model here depends on two fractional parameters α and β. The first fractional parameter α appears in the heat equation, the equation of motion, and the constitutive equations.
The effects of the fractional parameters corresponding to two models [15, 16] are discussed. The solution is obtained in Laplace transformed domain using a direct approach. All the studied field variables are represented graphically.
2. The Mathematical Model
The governing fractional Maxwell’s equations in an electromagnetic field are given by [15] (1)∇×H=J+ε0∂E∂t,(2)∇×E=-μ0τβ-1β!∂βH∂tβ,(3)∇·H=0,∇·E=0,where β is the order of the fractional derivative in Caputo sense with respect to time t such that 0≤β≤1, μ0 is the magnetic permeability, and ε0 is the electric permeability. H and E are the magnetic and electric field intensities, respectively. J is the current density and τ is a positive constant.
Equation (2) is called “Fractional Faraday’s Law of Magnetic Induction” which was derived in [15]. The proof uses Faraday’s induction law and the fractional Taylor’s series expansion developed by Jumarie [14].
Ohm’s law for moving media states that(4a)J=σ0E+μ0∂u∂t×H,where σ0 is the electric conductivity of the medium (assumed to be infinite) and u is the displacement vector.
Since J is bounded and σ0 is infinite, it follows that(4b)E=-μ0∂u∂t×H.
The governing equations for generalized fractional thermoelasticity associated with two relaxation times in the absence of external body forces and heat sources are given by [16] the following.
(i) The constitutive equations:(5)σij=2μeij+λδijekk-β1δij1+τ1αα!∂α∂tαθ,where θ is the temperature of the medium, σij are the components of the stress tensor, the constants λ and μ are Lamé’s constants, and β1=αt(3λ+2μ), where αt is the coefficient of linear thermal expansion and α∈0,1 is the order of the time-fractional derivative. τ1 is a positive constant and eij are the components of the strain tensor.
(ii) The strain-displacement relations:(6)eij=12ui,j+uj,i.
(iii) The equation of motion:(7)σij,j+Fi=ρu¨i,where ρ is the density assumed independent of time t, ui is the displacement vector component, σij are the components of the stress tensor, and Fi is the component of the Lorentz force given by(8)Fi=μ0J×Hi.
(iv) The fractional heat equation:(9)kθ,ii=β1θ0e˙kk+ρCE∂∂t+τ0αα!∂1+α∂t1+αθ,where the specific heat at constant strain is CE and k is the thermal conductivity. τ0 is a positive constant and θ0 is a reference temperature assumed to be such that (θ-θ0)/θ0≪1. As usual, superimposed dots denote time derivatives. The convention of summing over repeated indices is used.
3. Formulation of the Problem
We consider a magnetothermoelastic thick plate of perfect conductivity occupying the region 0≤x≤l in an initial magnetic field H0 in y direction at a uniform reference temperature θ0. This produces an induced magnetic field h in the direction of the y-axis and an electric field E in the z-axis (perpendicular to H and u).
The x-axis is perpendicular to the surface of the plate. The upper surface (x=0) of the plate is taken to be traction-free and is subjected to a thermal shock that is a function of time. The lower surface (x=l) of the plate is taken to be thermally isolated and laid on a rigid foundation.
It is assumed that all the state functions depend on x and t only.
Thus(10)u=ux,t,0,0,H=0,H0+h,0,E=0,0,E.
In the context of generalized fractional thermoelasticity associated with two relaxation times, the constitutive equations, the equation of motion, and fractional heat equation can be expressed in our case as follows:(11)σ=σxx=2μ+λ∂u∂x-β11+τ1αα!∂α∂tαθ,exx=∂u∂x,ρ∂2u∂t2+F1=2μ+λ∂2u∂x2-β11+τ1αα!∂α∂tα∂θ∂x,k∂2θ∂x2=∂∂tβ1θ0∂u∂x+ρCE1+τ0αα!∂α∂tαθ.
Now, (1)–(4b) yield (12)J=∂h∂x-ε0∂E∂t,(13)h=-H0β!τβ-1∂1-β∂t1-β∂u∂x,(14)E=-μ0H0∂u∂t.From (13), (14), and (12), we obtain(15)J=ε0μ0H0∂2u∂t2-H0β!τβ-1∂1-β∂t1-β∂2u∂x2.From (13), (15), and (8), we obtain(16)F1=μ0H02β!τβ-1∂1-β∂t1-β∂2u∂x2-ε0μ0∂2u∂t2.We shall use the following nondimensional variables(17)x′,u′=cηx,u,t′,τ0′,τ1′,τ′=c2ηt,τ0,τ1,τ,σ′=σλ+2μ,θ′=β1θ-θ0λ+2μ,h′=hH0,E′=Eμ0H0c.
Now, using the above nondimensional variables, the system of equations of the problem will reduce to(18)1+α02∂2u∂t2=1+RHβ!τ1-β∂1-β∂t1-β∂2u∂x2-1+τ1αα!∂α∂tαθ,∂2θ∂x2=∂∂tε∂u∂x+1+τ0αα!∂α∂tαθ,exx=∂u∂x,h=-β!τ1-β∂1-β∂t1-β∂u∂x,E=-∂u∂t,σ=∂u∂x-1+τ1αα!∂α∂tαθ,where(19)c=λ+2μρ,η=ρcEk,RH=μ0H0ρc2,α0=μ0ε0H0ρ,ε=T0β12λ+2μkη.
Also, we assume that the medium is initially at rest and the undisturbed state is maintained at uniform reference temperature. Then we have(20)u=ux,0=u˙x,0=θx,0=θ˙x,0=0.
Now, taking the Laplace transform (denoted by an overbar) with parameter s of both sides of the above system of equations, we get(21)1+α02s2u-=1+RHβ!τ1-βs1-β∂2u¯∂x2-1+τ1αsαα!θ¯,(22)∂2θ∂x2=sε∂u¯∂x+1+τ0αsαα!θ¯,(23)e¯xx=∂u¯∂x,(24)h¯=-β!τ1-βs1-β∂u¯∂x,(25)E¯=-su¯,(26)σ¯=∂u¯∂x-1+τ1αsαα!θ¯.Solving (21) and (22), we obtain(27)d4dx4+L2d2dx2+L1ddx+L0θ¯,u¯=0.The corresponding characteristic equation of (27) is (28)k4+L2k2+L1k+L0=0,where (29)L2=-s1+τ0αsα/α!1+RHβ!τ1-βs1-β+1+α02s21+RHβ!τ1-βs1-β,L1=-εs1+τ1αsα/α!1+RHβ!τ1-βs1-β,L0=-s31+τ0αsα/α!1+α021+RHβ!τ1-βs1-β.The solution of (27) compatible with (22) has the form(30)θ¯=∑i=14εskiCiekix,(31)u¯=∑i=14ki2-s1+τ0αsαα!Ciekix.Substituting (30) and (31) into (26), we get (32)σ¯=∑i=14kiki2-s1+τ0αsαα!-εs1+τ1αsαα!Ciekix.Using (31), (23)–(25) become (33)e¯xx=∑i=14kiki2-s1+τ0αsαα!Ciekix,h¯=-β!τs1-β∑i=14kiki2-s1+τ0αsαα!Ciekix,E¯=-s∑i=14ki2-s1+τ0αsαα!Ciekix.We assume that the boundary conditions have the form(34)θ0,t=ft,∂θl,t∂x=0,σ0,t=ul,t=0,t>0,where f(t) is a known function of t. Equations (30)–(32) give(35)∑i=14εskiCi=f¯s,∑i=14ki2Ciekil=0,∑i=14ki2-s1+τ0αsαα!Ciekil=0,∑i=14kiki2-s1+τ0αsαα!-εs1+τ1αsαα!Ci=0.
4. Inversion of the Laplace Transform
We shall now outline the method used to invert the Laplace transforms in the above equations. Let f-(s) be the Laplace transform of a function f(t). The inversion formula for Laplace transforms can be written as [19](36)ft=eγt2π∫-∞∞eityf-γ+iydy,where γ is an arbitrary real number greater than all the real parts of the singularities of f-(s). Expanding the function h(t)=exp(-γt)f(t) in a Fourier series in the interval [0,2L], we obtain the approximate formula [19]: (37)ft≈fNt=12c0+∑k=1Nck,for 0≤t≤2L,where (38)ck=eγtLReeikπt/Lf-γ+ikπL.
Two methods are used to reduce the total error. First, the “Korrektur” method is used to reduce the discretization error. Next, ε-algorithm is used to reduce the truncation error and therefore to accelerate convergence.
The Korrektur method uses the following formula to evaluate the function f(t):(39)ft=fNKt=fNt-e-2γLfN′2L+t.
We shall now describe ε-algorithm that is used to accelerate the convergence of the series in (37). Let N be an odd natural number and let sm=∑k=1mck be the sequence of partial sums of (37). We define ε-sequence by (40)ε0,m=0,ε1,m=sm,m=1,2,3,…,εn+1,m=εn-1,m+1+1εn,m+1-εn,m,n,m=1,2,3,….It can be shown that [19] the sequence ε1,1,ε3,1,…,εN,1,… converges to f(t)-c0/2 faster than the sequence of partial sums.
5. Numerical Results and Discussion
In order to obtain the solutions for the field functions in the physical domain, we have applied the Laplace inversion formula mentioned in the above section. FORTRAN programming language was used on a personal computer. The accuracy maintained was 7 digits for the numerical program. For computational purposes, a copper-like material has been taken into consideration. The values of the material constants are taken as in Table 1.
Material properties.
k = 386 W/(m K)
αt = 1.78 (10)−5 K−1
cE = 381 J/(kg K)
η = 8886.73
μ = 3.86 (10)10 kg/(m s2)
λ = 7.76 (10)10 kg/(m s2)
ρ = 8954 kg/m3
T0 = 293 K
ε = 0.0168
τ0 = 0.025 s
τ1 = 0.025 s
τ = 0.025 s
In order to investigate the effect of time on all field variables the computations have been carried out for t=0.05, t=0.1, t=0.15, and t=0.2. The results are displayed in Figures 1–5, for the temperature, displacement, stress, and magnetic and electric field distributions, respectively.
Temperature distribution for different time for α=β=1.
Displacement distribution for different times for α=β=1.
Stress distribution for different time when α=β=1.
Induced magnetic field distribution for different time for α=β=1.
Induced electric field distribution for different time for α=β=1.
Next, to study the effect of the fractional parameter α, we take α=0.01, α=0.2, α=0.5, and α=1 when t=0.1 and β=1. This gives us Figures 6–10, for the temperature, displacement, stress, magnetic field, and electric field distributions, respectively.
Temperature distribution for different α for β=1 at t=0.1.
Displacement distribution for different α for β=1 at t=0.1.
Stress distribution for different α for β=1 at t=0.1.
Induced magnetic field distribution for different α for β=1 at t=0.1.
Induced electric field distribution for different α for β=1 at t=0.1.
The remaining figures illustrate the effect of the fractional parameter β on the field variables. Figures 11–14 describe the displacement, stress component, magnetic field, and electric field distributions for different values of β; namely, β=0.2, β=0.5, β=0.9, and β=1 when t=0.1 and α=1. Here we notice that the fractional parameter β has no effect on the temperature distribution.
Displacement distribution for different β for α=1 at t=0.1.
Stress distribution for different β for α=1 at t=0.1.
Induced magnetic field distribution for different β for α=1 at t=0.1.
Induced electric field distribution for different β for α=1 at t=0.1.
The conclusions from Figures 1–14 can be summarized as follows:
(1) All of the physical variables have a finite speed of wave propagation for all times when α=β=1. The speed of propagation for other values of α and β needs further theoretical investigation.
(2) As is apparent from the order of the differential equation, we have two waves. The locations of the two wave fronts are the same for all functions considered. These wave fronts appear as a jump (discontinuity) in the case of the temperature, stress, and the intensity of the electric field. On the other hand, the other two functions, namely, the displacement and the intensity of the magnetic fields, are continuous. The wave fronts, in this case, appear as a cusp signifying a discontinuous first derivative. Note that the first jump in the temperature is too small to appear in the graph. For α=β=1, the waves propagate into the medium from the position x=0 to fill a finite part of the region that expands with the passage of time. We note that, for t=0.2, this region has filled the entire body of the plate. The second wave has been reflected from the other side of the plate. The positions of the wave fronts for the stress and electric field distributions as well as the size of their jump discontinuities can be found in Tables 2 and 3.
Wave fronts and stress jump sizes.
Time
First wave front of the stress
Second wave front of the stress
Location
Discontinuity gap size
Location
Discontinuity gap size
0.08
x=0.0748
1.5341611
x=0.5046729
0.4026305
0.1
x=0.0935
1.5374964
x=0.6261683
0.2926325
0.15
x=0.1495327
1.5193814
x=0.9439253
0.1216954
0.2
x=0.1962617
1.5351866
x = 0.7476634 (reflected)
0.0026
Wave fronts and electric field jump sizes.
Time
First wave front of the electric field
Second wave front of the electric field
Location
Discontinuity gap size
Location
Discontinuity gap size
0.08
x=0.0748
1.5456
x=0.5046729
0.7192609
0.1
x=0.0935
1.5477
x=0.6261683
0.4947222
0.15
x=0.1495327
1.5299
x=0.9439253
0.1858727
0.2
x=0.1962617
1.541777
x=0.7476634 (reflected)
0.0069
(3) From Figure 3, we notice that for the displacement, the first wave front is located at the peaks of this function. The location of these peaks for different times is introduced in Table 4.
Wave fronts and displacement peak sizes.
Time
First wave front location (peak)
Location of 1st wave front
Maximum value at peak
0.08
0.0748
0.133331
0.1
0.093457945
0.1530982
0.15
0.1495327
0.1892859
0.2
0.1962617
0.224686
Also, we note here that the magnitude of peaks increases with time.
(4) It is observed from Figures 6–10 that the field variables θ, u, σ, h, and E are strongly affected by the fractional parameter α. Increasing the fractional parameter α produces an increase in the peaks of the displacement and decreases the stress gap size at the location of the first wave front. Similar effects of the parameter β on u, σ, h, and E are also clear from Figures 11–14.
Competing Interests
The author declares that there are no competing interests.
PodlubnyI.1999New York, NY, USAAcademic PressMR1658022KlafterJ.LimS. C.MetzlerR.2011SingaporeWorld ScientificHilferR.2000World Scientific Publishing10.1142/9789812817747MR1890104TarasovV. E.2011Higher Education PressPovstenkoY. Z.Fractional heat conduction equation and associated thermal stress20042818310210.1080/014957390523741MR21193532-s2.0-13844253772PovstenkoY. Z.Fractional radial heat conduction in an infinite medium with a cylindrical cavity and associated thermal stresses201037443644010.1016/j.mechrescom.2010.04.0062-s2.0-77955414263SheriefH. H.El-SayedA. M. A.Abd El-LatiefA. M.Fractional order theory of thermoelasticity201047226927510.1016/j.ijsolstr.2009.09.0342-s2.0-70449480851RaslanW. E.Application of fractional order theory of thermoelasticity to a 1D problem for a spherical shell201654129530410.15632/jtam-pl.54.1.2952-s2.0-84955145752RaslanW. E.Application of fractional order theory of thermoelasticity to a 1D problem for a cylindrical cavity2014664257267MR3243413RaslanW. E.Application of fractional order theory of thermoelasticity in a thick plate under axisymmetric temperature distribution201538773374310.1080/01495739.2015.10403072-s2.0-84930240006SheriefH.Abd El-LatiefA. M.Effect of variable thermal conductivity on a half-space under the fractional order theory of thermoelasticity20137418518910.1016/j.ijmecsci.2013.05.0162-s2.0-84880569691WangY.LiuD.WangQ.ZhouJ.Effect of fractional order parameter on thermoelastic behaviors of elastic medium with variable properties201528668269210.1016/s0894-9166(16)30009-x2-s2.0-84952901800EzzatM. A.Magneto-thermoelasticity with thermoelectric properties and fractional derivative heat transfer20114061303510.1016/j.physb.2010.10.0052-s2.0-78349303958JumarieG.Derivation and solutions of some fractional Black–Scholes equations in coarse-grained space and time. Application to Merton's optimal portfolio20105931142116410.1016/j.camwa.2009.05.0152-s2.0-74249118189HamzaF.Abd El-LatiefA. M.FayikM.Memory time effect on electromagnetic-thermoelastic materials201529447450110.1080/09205071.2014.10026302-s2.0-84926408002HamzaF.AbdouM.Abd El-LatiefA. M.Generalized fractional thermoelasticity associated with two relaxation times20143791080109810.1080/01495739.2014.9361962-s2.0-84904652759HamzaF.Abd El-LatiefA. M.KhatanW.Thermomechanical fractional model of two immiscible TEMHD201520151639145410.1155/2015/3914542-s2.0-84930630367HamzaF.Abd El-LatiefA. M.AbdouM.1D applications on fractional generalized thermoelasticity associated with two relaxation times201623668970310.1080/15376494.2015.10291582-s2.0-84955294565HonigG.HirdesU.A method for the numerical inversion of Laplace transforms198410111313210.1016/0377-0427(84)90075-XMR7365942-s2.0-0021375050