See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/324220609

Solving SchrÖdinger equation for a particle in one-dimensional lattice: An

homotopy perturbation approach

Article  in  Romanian Reports in Physics · March 2019

CITATIONS READS

3 570

3 authors, including:

Tijana S Kevkić Vladica Stojanović

Faculty of Natural Science and Mathematics, University of Priština, Kosovska Mitr… University of Kosovska Mitrovica, Faculty of Sciences and Mathematics
24 PUBLICATIONS   33 CITATIONS    47 PUBLICATIONS   90 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

ECF Method View project

GL functions in MOSFET Modeling View project

All content following this page was uploaded by Vladica Stojanović on 22 March 2019.

The user has requested enhancement of the downloaded file.

Romanian Reports in Physics 71, 101 (2019)

SOLVING SCHRÖDINGER EQUATION FOR A PARTICLE

IN ONE-DIMENSIONAL LATTICE:
AN HOMOTOPY PERTURBATION APPROACH

TIJANA KEVKIĆ1,a , VLADICA STOJANOVIĆ1,b , DRAGAN PETKOVIĆ1,c

1,a,c
Department of Physics, 1,b Department of Mathematics
1
Faculty of Sciences and Mathematics,
Lole Ribara 29, Kosovska Mitrovica, Serbia
Email: a [email protected], b [email protected], c [email protected]
Received February 14, 2018

Abstract. In this paper, a novel approach for an approximate solving Schrödin-

ger equation for a particle in the one-dimensional lattice with the periodic potential is
described. This approach, based on the homotopy perturbation method (HPM), gives
an approximate analytic solution which has a high degree of convergence, and at the
same time high degree of accuracy. The convergence of the HPM is examined and
formally confirmed. In addition, the efficiency of the HPM method is illustrated in two
examples.

Key words: Homotopy perturbations, Schrödinger equation, periodic potential,

wave function, periodic boundary conditions, approximation.

1. INTRODUCTION & DEFINITION OF THE PROBLEM

A nontrivial problem of the computation of electronic states in a solid can be

greatly simplified in the case of a crystal, i.e. in the case of a regular, periodic and
infinite arrangement of atoms. In this case, under the suitable assumptions, the many-
electron problem turns into much simpler problem of an electron under a periodic
potential. Periodicity can be mathematically formalized in a simple and general way
in any number of dimensions. In the following we will assume a one-dimensional
array of atoms of the same kind, regularly spaced by a distance a. Let us emphasize
that a is the lattice parameter, while an = na, where n is an integer, are the vectors
of the crystal lattice. We suppose that system has a discrete translational invariance,
that is: it is equal to itself if translated by a or multiples of a. Furthermore, crystal
potential will be labeled as V (x), where x is the position of the particle in the lat-
tice. ForP this potential which is formed by the superposition of atomic-like potentials
V (x) = n Vn (x−an ), the following symmetry holds: V (x+a) = V (x). Thanks to
such symmetry the infinite line which represents considered one-dimensional space
can be decomposed into finite space regions of length a, periodically repeated. That
finite region [−a/2, a/2] defines the so-called unit cell.
(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG
Article no. 101 Tijana Kevkić, Vladica Stojanović, Dragan Petković 2

For an electron moving in the potential V (x), the one-dimensional, time inde-
pendent Schrödinger equation is given by:
d2 ψ(x) 2m
+ 2 (E − V (x)) ψ(x) = 0. (1)
dx2 }
Here, ψ(x) is the wave function, in general complex, m is mass of electron, E is its
total energy, and } is reduced Planck’s constant. In the following, the aforementioned
idealized modeling will be applied to a crystal of finite length L = N a, where N is the
number of unit cells, i.e. it will be assumed that x ∈ [−L/2, L/2]. It can be achieved
by introducing the periodic boundary conditions (PBC) ψ(−L/2) = ψ(L/2) = 0.
Further, also, will be assumed that wave function satisfies the probability density
normalized condition:
Z L/2 Z L/2
∗
ψ(x)ψ (x)dx = |ψ(x)|2 dx = 1.
−L/2 −L/2

In spite of the PBC which enable that the system can be treated as a one-
dimensional segment of finite length L, still retaining the discrete translational in-
variance, the solution of Eq.(1) for a periodic potential is not a simple problem.
There is still a need for finding a number of single-particle states equal to at least
half the number of electrons in the system. Due to the translational symmetry of
the problem, solution of Eq.(1) satisfies the Bloch’s condition ψ(x + a) = ψ(x)eika ,
with ψ(x) = U (x)eikx , where k is the wave vector and U (x) is a periodic function
of lattice period a. Also, it can be easily verified that wave function obeys the con-
dition of translational symmetry ψ(x) = ψ(x + L). Thus, the Bloch’s condition and
last equality imply that only values of k such that exp(ikL) = 1 are compatible with
PBC, that is, k must be an integer multiple of 2π/L. However, the time-independent
Schrödinger Eq.(1) for periodic potential has not been solved exactly even for in-
finitely large periodic systems despite the simplification that Bloch’s theorem pro-
vides. Usually, its solutions for one dimensional finite systems are obtained by using
various numerical techniques in some special cases [1, 2]. Here, we used the homo-
topy perturbations method (HPM) as a novel approach to find approximate analytical
solutions of the Schrödinger equation for arbitrary one-dimensional periodic poten-
tials. The HPM, firstly introduced by He [3–6], is an approximate-analytical method,
different from other similar ones, because it provides a simple way to adjust and con-
trol the convergence region of the solution series by choosing the proper values for
auxiliary parameters. In recent years, the HPM has been intensive developed and it
was the subject of extensive studies [7–10]. This method is also found a very useful
application in solving the various kind of nonlinear equations, mostly in the physical
sciences [11–15]. Especially, the HPM has been used in solving the different forms
of the nonlinear Schrödinger equation [16, 17].
(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG
3 Solving Schrödinger equation for a particle in one-dimensional lattice Article no. 101

2. HPM SOLUTION OF THE SCHRÖDINGER EQUATION

In order to solve the Eq.(1) using HPM, let us consider the following homotopy
equation:
 2
∂ Ψ(x; p) d2 ψ0 (x)
  2 
∂ Ψ(x; p) 2m
(1 − p) − +pξ + 2 (E − V (x)) Ψ(x; p) = 0,
∂x2 dx2 ∂x2 }
(2)
where p ∈ [0, 1] is the embedding (homotopy) parameter, and ξ 6= 0 is the auxil-
iary parameter. When p = 0, the homotopy Eq.(2) has the so-called initial solution
Ψ(x; 0) = ψ0 (x), which can be chosen arbitrarily, but such that satisfies the PBC
ψ0 (± L/2) = 0. On the other hand, when p = 1, Eq.(2) becomes equivalent to the
Schrödinger Eq.(1), whereby we assume that the same PBC are valid.
The basic assumption of the HPM is that solution of homotopy Eq.(2) can be
expressed as the power series in p:
∞
X
Ψ(x; p) = pj ψj (x). (3)
j=0

According to this, the solution of the Schrödinger Eq.(1) can be obtain as:
∞
X
ψ(x) = lim Ψ(x; p) = ψj (x), (4)
p→1−
j=0

provided that the series in Eq.(4) converges. Substituting Eq.(3) in the homotopy
Eq.(2), it follows:

∞ ∞
X d2 ψj (x) X d2 ψj−1 (x) d2 ψ0 (x)
pj + (ξ − 1) p j
+ p
dx2 dx2 dx2
j=1 j=1
∞ (5)
2m ξ
X j
+ 2 E − V (x) p ψj−1 (x) = 0.
}
j=1

Now, by equating in Eq.(5) the expressions with identical powers pj , j = 1, 2, . . . , the

following differential equations get ones:
d2 ψ1 (x)
 2 
d ψ0 (x) 2m

+ξ + 2 E − V (x) ψ0 (x) = 0, (6a)
dx2 dx2 }
d2 ψj (x) d2 ψj−1 (x) 2m ξ

2
+ (ξ − 1) 2
+ 2 E − V (x) ψj−1 (x) = 0, (6b)
dx dx }
where j = 2, 3, . . . . Also, we assume that the PBCs ψj (±L/2) = 0 hold for each
j = 1, 2, . . . In that way, the Eqs.(6a)-(6b) can be solved recursively on ψj (x) using
(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG
Article no. 101 Tijana Kevkić, Vladica Stojanović, Dragan Petković 4

the double integration:

 Z Z
2m ξ
 ψ1 (x) = 2

 dx (V (x) − E) ψ0 (x)dx − ξ ψ0 (x),
}
Z Z (7)
2m ξ
dx (V (x) − E) ψj−1 (x)dx + (1 − ξ)ψj−1 (x),

 ψj (x) = 2

}
where j = 2, 3, . . . Finally, according to Eq.(4), the HPM approximations of the un-
known wave function ψ(x) will be:
n
X
ψbn (x) := ψj (x), n = 0, 1, 2, . . . (8)
j=0
Z L/2  2
where the normalized condition ψn (x) dx = 1 holds. In the following, some
b 
−L/2
sufficient conditions for the existence and convergence of HPM approximations has
been given.
Theorem 1 Let {ψj (x)}∞ j=0 be the sequence of functions defined, for an arbitrary
L > 0, on the interval [−L/2, L/2] by the recurrence relations in Eqs.(7). In addition,
assume that the following conditions are satisfied:
(i) The function ψ0 (x) is an uniformly bounded on [−L/2, L/2], i.e. there
exists constant M > 0 such that the inequality
kψ0 (x)k := max |ψ0 (x)| ≤ M
|x|≤L/2

holds.
(ii) Potential V (x) is a periodic, integrable function on [−L/2, L/2], with pe-
riod a ∈ (0, L) and Fourier expansion:
X
V (x) = v` ei`x , (9)
`∈G
Z a/2
1
V (x)e−i`x dx are the Fourier
 2πn
where G = a | n = 0, ±1, ±2, . . . and v` =
a −a/2
coefficients of the function V (x).
2m X |v` |
(iii) For some value of ξ 6= 0, series C := satisfies the condition
}2 `2
`∈G
0 < ξ(1 − C) < 1.
Then, for an arbitrary L > 0, the sequence {ψbn (x)}∞n=0 , defined by Eq.(8),
uniformly converges on [−L/2, L/2] to the solution of the Schrödinger Eq.(1).

(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG

5 Solving Schrödinger equation for a particle in one-dimensional lattice Article no. 101

Proof. According to assumptions (i)–(ii) of the theorem, as well as Eqs.(7), it

follows:
 Z Z !
2m X 
|ψ1 (x)| ≤ M ξ 1 + 2  v` dx ei`x dx ≤ M ξ(C + 1),

}  
`∈L
 Z Z !
2m ξ X i`x

|ψ2 (x)| ≤ M ξ(C + 1) 1 − ξ + 2  v` dx e dx

}  
`∈L
h i
≤ M ξ(C + 1) 1 + ξ(C − 1) , etc.
In general, using the induction method, it can be easily proved that inequalities:
h ij−1
|ψj (x)| ≤ M ξ(C + 1) 1 + ξ(C − 1) (10)

hold for each j = 1, 2, . . . Now, for an arbitrary but fixed x ∈ [−L/2, L/2], let r(x)
be the radius of convergence of the power series in Eq.(3). Applying the Cauchy-
Hadamard theorem, as well as Eqs.(10) and the assumption (iii), we find that:
" #−1
h i−1/j h i−1+1/j
r(x) = lim sup |ψj (x)|1/j ≥ lim M ξ(C + 1) 1 + ξ(C − 1)
j→∞ j→∞
h i−1
≥ 1 + ξ(C − 1) > 1.
for some ξ 6= 0. Thus, this power series converges at p = 1.
On the other hand, according to Eqs.(7)-(10), for the HPM approximations
{ψn (x)}∞
b
n=0 we obtain:
 
ψ0 (x) = |ψ0 (x)| ≤ M,
b 
    h i
ψ1 (x) ≤ ψb0 (x) + |ψ1 (x)| = M 1 + ξ(C + 1) ,
b   
    h
i
ψ2 (x) ≤ ψb1 (x) + |ψ2 (x)| = M 1 + ξ(C + 1) 1 + (1 + ξ(C − 1)) ,
b   

and, in general,
1 − (1 + ξ(C − 1))n
   
ψn (x) ≤ M 1 + ξ(C + 1) , n = 1, 2, . . . . (11)
b 
ξ(1 − C)
In the limit case, when n → ∞, inequalities (11) imply:
 
∞
1 − (1 + ξ(C − 1))n
    
X 
ψj (x) = lim ψn (x) ≤ M lim 1 + ξ(C + 1)
  b 

 j=0  n→∞ n→∞ ξ(1 − C)
2M
= < +∞.
1−C
(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG
Article no. 101 Tijana Kevkić, Vladica Stojanović, Dragan Petković 6

Thus, the power series in Eq.(3) is absolutely convergent at p = 1, and uniformly

converges on [−L/2, L/2], for an arbitrary L > 0. According to Abel’s theorem, it
follows that homotopy function Ψ(x, p), defined by Eq.(3), P∞ is continuous from the
left at p = 1. Therefore, the Eq.(4) holds, i.e. the series j=0 ψj (x) is a solution of
Eq.(1). 
Remark 1 Thanks to the appropriate choices of the initial wave function ψ0 (x), the
condition (i) of the previous theorem can be easily fulfilled. Moreover, the condi-
tion (ii) is provided for an arbitrary periodic and integrable function V (x). Finally,
according the well-known results of the Fourier theory (see, for instance [18]), for
the Fourier coefficients v` of the function V (x) is 2
R v` /` = w` , where w` are
R valid
the Fourier coefficients of the function W (x) := dx V (x)dx. Obviously, W (x)
is also periodic and continuous on [−L/2, L/2], for an arbitrary L > 0, so the series
C is (always) convergent. Thus, the condition (iii) can be satisfied for some values
of the auxiliary parameter ξ 6= 0.

3. VALIDATION OF THE HPM PROCEDURE

To justify application of the aforementioned HPM procedure in solving the

Schrödinger equation with periodic potential two important cases will be considered:
A) Kronig-Penney model representing a periodic arrangement of rectangular
potential wells and potential barriers [20]. Assuming that potential barrier width is
a − b and the width of potential well is b, the potential can be described as:
+∞  !
X x − na
V (x) := V0 1 − Π ,
n=−∞
b
where a is the period, V0 is the amplitude of potential (i.e. the depth of well or the
height of the barrier), and
(
1, |x| ≤ 1/2,
Π(x) :=
0, |x| > 0,
is the so-called rectangle function. To obtaining HPM approximations of the wave
function ψ(x) on [−L/2, L/2], we have taken the auxiliary parameter p ξ = 1, and the
initial approximation ψ0 (x) = U0 (x) exp(ikx). Here, U0 (x) = 2/L sin(N πx/L)
with N = 1, 2, . . . , is the periodic wave function corresponding also to the solution of
the Schrödinger Eq.(1) with the so-called empty potential V (x) ≡ 0 (see, for instance
[21]).
To demonstrate the implementation of the HPM procedure in the considered
case of rectangular wells and barriers we have chosen next values: the period a = 2,
the width of barriers and wells a − b = b = 1, as L = 12 for the length of crystal (all
(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG
7 Solving Schrödinger equation for a particle in one-dimensional lattice Article no. 101

of these values are referred to E-10m), as it is shown in Fig. 1 (above left plot). From
L = N a follows that the number of unit cells is N = 6, while the wave number is
k = 2π/L = π/6. Taking these values as well as V0 = π/a = π/2 for amplitude of
potential, numerical computation gives the corresponding value of electron energy
E = 2.621eV. The convergence of thus obtained the HPM approximations ψbn (x), of
the order n = 0, 1, 2, . . . , 6, is shown in Fig. 1, where their density functions |ψbn (x)|2 ,
as well as the real and imaginary parts are plotted. As it can be easily seen, for n ≥ 1,
the HPM approximations {ψbn (x)} show a high degree of convergence, which is also
formally confirmed (see Table 1, below).

Fig. 1 – Graphs of the Kronig-Penney potential (above left), as well as the HPM-approximations
ψbn (x) of the wave function ψ(x): density functions (above right), real parts (bellow left) and
imaginary parts (bellow right).

Notice that in original Kronig-Penney model there are two wave functions cal-
culated by solving Schrödinger equation separately for the regions of barrier and
well. Using HPM procedure does not require the separate solving of Schrödinger
equation in mentioned regions, but along the whole length of observed interval. We
can conclude that HPM approximations obtained here are appropriate for the both of
potential regions: they take the form of wave function for empty potential in the well,
while ones represent the transmitted particles through barrier region.
(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG
Article no. 101 Tijana Kevkić, Vladica Stojanović, Dragan Petković 8

B) Considered, as the second case, the periodic potential given by following

function V (x) := V0 cos(ωx). This function, in general, corresponds to the well-
known Mathieu equation, which has found important applications in describing, for
instance, the flux lattice in superconductors, the wave motion in optical lattice, or
in some other periodic media [22]. With the aim to illustrate the application of the
HPM procedure to the case of a particle in considered periodic potential, we have
chosen that value of the amplitude of the potential be V0 = 1, as well as ω = 2 which
implies the period a = π (Fig. 2, above left plot). For comparison with the previ-
ous considered Kronig-Penney model, also is assumed that N = 6, i.e. the length of
the observed interval is L = 6π, and value of the wave vector is k = 2π/L = 1/3,
which implies the appropriate energy E(k) = 2.209eV. Further on, the HPM ap-
proximations were computed with the same auxiliary parameter ξ = 1 and the initial
approximation ψ0 (x). The HPM approximations {ψbn (x)}, when n ≥ 1, obtained
in the case of this periodic potential also fast converge to the exact solution of the
Eq.(1), as it can be easily seen in Fig. 2.

Fig. 2 – Graphs of the Mathieu’s periodic function (above left), as well as the HPM-approximations:
density functions (above right), real parts (bellow left) and imaginary parts (bellow right).

In addition, for each of the HPM approximation ψbn (x), n = 1, 2, . . . , 6, it was

computed the maximum approximation errors:
 
 d2 ψb (x) 2m 
n
Err(ψbn ) := max  + [E − V (x)] ψ
bn (x) ,

|x|≤L/2  dx2 }2 
(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG
9 Solving Schrödinger equation for a particle in one-dimensional lattice Article no. 101

as well as the maximum iteration differences:

 
ψn − ψbn−1 := max ψbn (x) − ψbn−1 (x) .
b  
|x|≤L/2

Let us emphasize that both of these values represent the measures of convergence
of the HPM approximations {ψbn (x)}, in comparison to their approximations order
n ≥ 1. These measures obtained for both of the aforementioned series of the HPM
approximations, labeled as A and B, respectively, are given in Table 1. The whole
computation procedure has been realized and implemented in the software package
MATHEMATICA 11.0. As it can be easily seen, the HPM approximations converge
somewhat faster in the case of B-series. However, it should be pointed out that
the both series of the HPM approximations have approximately the same level of
accuracy, of the order 10−5 , which they achieve already for n ≥ 5.

Table 1

Maximum approximation errors and iteration differences of the HPM approximations {ψbn (x)}.

Err(ψbn ) ψn − ψbn−1
b
Approximation
order (n) A-series B-series A-series B-series
1 1.82E-02 1.72E-02 7.44E-01 3.53E-01
2 2.26E-03 1.65E-03 3.92E-02 5.79E-03
3 7.92E-04 2.30E-04 1.05E-03 4.88E-04
4 1.64E-04 9.77E-05 2.76E-04 8.76E-05
5 9.02E-05 2.50E-05 8.31E-05 5.14E-05
6 8.48E-05 2.13E-05 8.27E-06 6.07E-06

4. CONCLUSION

In this work, the electron moving in one dimensional lattice with an arbitrary
periodic potential is considered. By using periodic boundary conditions (PBC) and
Bloch’s theorem the system has been treated as the one-dimensional one of finite
length which shows the discrete translational invariance. For approximate analytical
solving the appropriate Schrödinger equation, the Homotopy Perturbation Method
(HPM) has been proposed. The existence and convergence of the so-called HPM
approximations was discussed and shown in general case. Also, two illustrative ex-
amples were given to confirm the suitability and the applicability of the HPM in
solving this kind of problem.
(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG
 Article no. 101 Tijana Kevkić, Vladica Stojanović, Dragan Petković 10

REFERENCES

1. R.J. Lombard, R. Mezhoud, The Linear PT-Symmetric Complex Potential, Rom. J. Phys. 62, 112
(2017).
2. R.J. Lombard, R. Mezhoud, R. Yekken, Complex Potentials With Real Eigenvalues and the Inverse
Problem, Rom. J. Phys. 63, 101 (2018).
3. J.-H. He, Homotopy perturbation technique, Comput. Methods Appl. Mech. Engrg. 178, 257-262
(1999).
4. J.-H. He, A coupling method of a homotopy technique and a perturbation technique for non-linear
problems, Int. J. Non-Linear Mech. 35, 37-43 (2000).
5. J.-H. He, Homotopy perturbation method: a new nonlinear analytical technique, Appl. Math.
Comput. 135, 73-79 (2003).
6. J.-H. He, An elementary introduction to the homotopy perturbation method, Comput. Math. Appl.
57, 410-412 (2009).
7. M.A. Noor, W.A. Khan, New iterative methods for solving nonlinear equation by using homotopy
perturbation method, Appl. Math. Comput. 219, 3565-3574 (2012).
8. A.A. Hemeda, Homotopy perturbation method for solving systems of nonlinear coupled equations,
Appl. Math. Sci. 96:6, 4787-4800 (2012).
9. M.-F. Zhang, Y.-Q. Liu, X.-S. Zhou, Efficient homotopy perturbation method for fractional non-
linear equations using Sumudu transform, Therm. Sci., 19(4), 1167-171 (2015).
10. X.-J. Yang, H.M. Srivastava, C. Cattani, Local fractional homotopy perturbation method for solv-
ing fractal partial differential equations arising in mathematical physics, Rom. Rep. Phys., 67(3),
752-761 (2015).
11. M. Zeb, T. Haroon, A.M. Siddiqui, Homotopy perturbation solution for flow of a third-grade fluid
in helical screw rheometer, U.P.B. Sci. Bull., Series A 76(4), 179-190 (2014).
12. P. K. Roy, A. Mallick, Thermal analysis of straight rectangular fin using homotopy perturbation
method, Alexandria Engrg. J. 55(3), 2269-2277 (2016).
13. K. Grysa, A. Maciag, Temperature dependent thermal conductivity determination and source iden-
tification for nonlinear heat conduction by means of the Trefftz and homotopy perturbation meth-
ods, Int. J. Heat Mass Tran. 100, 627-633 (2016).
14. T. Kevkić, V. Stojanović, D. Randjelović, Application of homotopy perturbation method in solving
coupled Schrödinger and Poisson equation in accumulation layer, Rom. J. Phys. 62, 122 (2017).
15. V. Stojanović, T. Kevkić, G. Jelić, D. Randjelović, Determination of invariant measures: An ap-
proach based on homotopy perturbations, U.P.B. Sci. Bull., Series A 2, 119-128 (2018).
16. D. Baleanu, A. K. Golmankhaneh, A.K. Golmankhaneh, Solving of the Fractional Non-Linear
and Linear Schrödinger Equations by Homotopy Perturbation Method, Rom. J. Phys. 54(9-10),
823-832 (2009).
17. S. Al-Jaber, Solution of the radial N-dimensional Schrödinger equation using homotopy perturba-
tion method, Rom. J. Phys. 58(3-4), 247-259 (2013).
18. J.D. Hamilton, Time Series Analysis, Princeton University Press, Princeton, New Jersey (1994).
19. B. Kollmitzer, P. Hadley, Thermodynamic properties of separable square-wave potentials, Physica
B 406, 4373-4380 (2011).
20. Y. Zhang, Y. Wang, Energy Band Analysis of MQW Structure Based on Kronig-Penny Model, J.
Modern Phys. 4, 968-973 (2013).
21. J. Sólyom, Fundamentals of the Physics of Solids, Vol. II: Electronic Properties, Springer, New
York (2004).
22. L. Ruby, Applications of the Mathieu equation, Am. J. Phys. 64(1), 39-44 (1996).

(c) 2019 RRP 71(0) 101 - v.2.0*2019.2.11 —ATG

View publication stats