Home Physical Sciences Decay Mode Solutions for the Supersymmetric Cylindrical KdV Equation
Article Publicly Available

Decay Mode Solutions for the Supersymmetric Cylindrical KdV Equation

  • Shufang Deng EMAIL logo , Weili Qin and Guiqiong Xu
Published/Copyright: May 26, 2016

Abstract

Supersymmetric cylindrical KdV equation is presented. Decay mode solutions for the supersymmetric KdV equation are derived by supersymmetric Hirota operator.

1 Introduction

Supersymmetric integrable systems have been studied extensively during the past decade. Thus, a large number of well-known integrable equations have been extended into the supersymmetric context, such as KdV equation, KP hierarchy, and Boussinesq equation [13]. It has been shown that these supersymmetric integrable systems possess the Bäcklund transformation, the Hamiltonian formalism, Darboux transformation, bilinear form, and multi-soliton solutions [48]. The bilinear form of supersymmetric integrable was introduced by Carstea [8], it requires an extension of the Hirota bilinear operator [9, 10] to the supersymmetric systems. In recent years, Carstea, Liu, and Zhang have done a lot of work on the supersymmetric equations [8, 1113]. However, to our knowledge, the supersymmetric equation for the variable coefficient KdV equation has been considered rarely so far.

The cylindrical KdV

(1)ut+6uux+uxxx+u2t=0 (1)

was first studied by Maxon and Viecelli [14]. They solved (1) numerically and obtained that (1) has cylindrical soliton solutions which consists of a pulse moving inward rapidly at an ever increasing speed, leaving behind a flat wake that moves inward at sound speed. In 1980 [15], Nakamura showed analytical soliton solution by Hirota method that agrees with numerical and experiment results.

In this article, we consider the cylindrical KdV equation. We will show the supersymmetric cylindrical KdV equation and obtain the decay mode soliton solutions for it.

2 Supersymmetrical KdV Equation

The supersymmetric extension of a nonlinear evolution equation refers to a system coupled equations for a bosonic u(x, t) and a fermionic field η(t, x), which reduces to the initial equation in the limit where the fermionic field is zero (the bosonic limit). We extend the classical space (x, t) to a large space (superspace) (t, x, θ), where θ is a Grassmann variable and also to extend the pair of fields (u, η) to a large fermionic or bosonic superfield ϕ(t, x, θ). In order to have a nontrivial extension for the cylindrical KdV, we choose ϕ to be fermionic, having the expansion

(2)ϕ=η(t,x)+θu(x,t). (2)

We consider the space supersymmetric invariance xxϵθ, θθ+ϵ (ϵ is an anticommuting parameter). Multiplying θ in (1), each term in the space supersymmetry is

(3)utϕt,uux(Dϕ)ϕx   or   (Dϕx)ϕ,uxxxϕxxx, (3)

where D=θ+θx is the superderivative and

(4)(Dϕx)ϕ=uxη+θuuxθηηxx, (4)
(5)(Dϕ)ϕx=uηx+θuux. (5)

Reduce to the initial equation in the limit where the fermionic field is zero (the bosonic limit). Cylindrical KdV can be extended to the supersymmetric

(6)ϕt+ϕxxx+3(Dϕ)(ϕx)+3(Dϕx)ϕ+ϕ2t=0. (6)

3 Bilinear Form and Decay Mode Solutions

In order to derive the bilinear form for the cylindrical KdV, we consider the transformation

(7)ϕ=2D(lnf)x=2D3(lnf), (7)

where f(t, x, θ) is bosonic. Equation (6) can be transformed into the following superbilinear form

(8)Fff=(SDt+SDx3+D2t)ff=0, (8)

here

(9)SDxnDtmfg=(Dθ1Dθ2)(x1x2)n(t1t2)mf(x1,t1,θ1)g(x2,t2,θ2)|x1=x2,t1=t2,θ1=θ2. (9)

We are going to derive the supersoliton solutions through the classical perturbative method. Expanding f into power series of a small parameter ϵ as

(10)f=1+ϵf(1)+ϵ2f(2)+ϵ3f(3)+. (10)

Substituting (10) into (8) and equating coefficients

(11)Ff(1)1=0, (11)
(12)2Ff(2)1=Ff(1)f(1), (12)
(13)Ff(3)1=Ff(1)f(2), (13)
(14). (14)

In order to obtain the solutions, we introduce the quantity aij defined by

(15)aij=ρiρj(12t)1/3[ω(zi)ω(zj)ω(zi)ω(zj)]/(zizj), (15)

where ρi, ρj are arbitrary constant parameters

(16)zk=(xxk)(12t)1/3+(θθk)(12t)1/3,k=i,j (16)

and ω(zk), k=i, j present Airy function Ai or Bi which are two linearly independent solutions of the ordinary differential equation

ω(z)zω(z)=0.

To the definition of aij for the case i=j, we denote as follows

(17)aii=ρi2(12t)1/3[ziω(zi)ω(zi)ω(zi)ω(zi)]. (17)

By the direct calculation, we can find

(18)aili=ρiρli(12t)23ωiωli, (18)
(19)aili=ρiρli(12t)1(ωiωli+ωiωli). (19)

By the direct calculation, we have found

(20)Faii1=3(12t)2(xixi+θiθi)2(1+θ)aii, (20)
(21)Faiiajj=(1+θ){3(12t)2[(xixi+θiθi)2+(xjxj+θjθj)2]aiiajj6(12t)2(xixj+θiθj)(xixj+θiθj)aijaij+6(12t)2(xixj+θiθj)(xjxi+θjθi)aijaji}. (21)

If taking

f(1)=a11=ρ12(12t)1/3[z1ω1(z1)ω(z1)ω(z1)ω(z1)],

we can obtain the one-decay solution. To get further than one solution, we must generalise closed relation (20) and (21). For the arbitrary functions a1, a2, b1, and b2, we can directly verify the following

(22)Fa1a2b1b2=(Fa1b1)a2b2+(Fa1b2)a2b1+(Fa2b1)a1b2+(Fa2b2)a1b1(Fa1a2)b1b2(Fb1b2)a1a2+c(2,2), (22)

where

(23)c(2,2)=3{(Dx2a1b1)(SDxa2b2)+(Dx2a2b2)(SDxa1b1)+(Dx2a1b2)(SDxa2b1)+(Dx2a2b1)(SDxa1b2)(Dx2a1a2)(SDxb1b2)(Dx2b1b2)(SDxa1a2)}=12θ{a1a2b1b2+a1(a2b1b2a2b1b2a2b1b2)+a2(a1b1b2a1b1b2a1b1b2)+(ab,ba)+2a1a2b1b2}+{6a1a2,θb1b2+6a1,θa2b1b2+3a1a2(b1b2,θ+b1,θb2)+3a2a1(b1b2,θ+b1,θb2)3a1(a2b1,θb2+a2,θb1b2+a2b2,θb1+a2,θb2b1)+6a1,θa2b1b2+6a2,θa1b1b23a2(a1b1,θb2+a1,θb1b2+a1b2,θb1+a1,θb2b1)6a1,θ(a2b1b2+a2b2b1)6a2,θ(a1b2b2+a1b2b1)+6a1,θa2b1b2+6a2,θa1b1b2+(ab,ba)}. (23)

Here, prime presents differentiation with respect to x. The relation can be extended to

(24)F(a1an)(b1bn)=i,j=1nF(aibj)i,j=1,ijnF(aiaj)i,j=1,ijnF(bibj)+c(n,n), (24)

where

(25)F(aibj)=(Faibj)(aibj)1k=1nakbk,F(aiaj)=(Faiaj)(aiaj)1k=1nakbk,F(bibj)=(Fbibj)(bibj)1k=1nakbk, (25)
(26)c(n,n)=12θ{i<jaiaj+ijk,j<kaiajakijaiajbk+j<kaibjbk+2i<j<k<laiajakal2j<k<laibjbkbl+(ab,ba)+2i<j,k<laiajbkbl}+{6ijaiaj,θ+3ijkaiajak,θ3ijai(ajbk,θ+aj,θbk)+3jkaibjbk,θ+6j<k<lai,θajakal6j<k<lai,θbjbkbl6k<laibj,θbkbl+6j<kai,θbjbk6ij,kai,θajbk+6ijk,j<kai,θajak+6ij,k<lai,θajbkbl+(ab,ba)}. (26)

We can prove (24–26) by mathematical induction. We consider (n+1)×(n+1) in (24) product as F(a1···an+1)· (b1···bn+1)=F(a1···an)an+1·(b1···bn)bn+1 and apply 2×2 formula given by (22) to the right-hand side of this. Using relation (24), we see that F(a1···an+1)·(b1···bn+1) reduces to the form (24, 26) with n replaced by n+1.

Let

(27)bm+1==bn=1,(mn), (27)

and rewrite the suffix

(28)aiaili,ajajlj,,bibisi,bjbjsj, (28)

Then

(29)F|a11a1nan1ann||b11b1mbm1bmm|=F[l1l2ln(1)τ(l1,,ln)a1l1a2l2anln][s1s2sm(1)τ(s1,,sn)b1s1b2s2bmsm]=l1l2ln,s1,s2,sm(1)τ(l1,,ln)(1)τ(s1,,sm){i=1nj=1mF(ailibjsj)+(nm)i=1nF(aili1)i,j=1,i<jnF(ailiajlj)i,j=1,i<jmF(bisibjsj)(nm)j=1mF(1bjsj)+c(n,m)}, (29)

where l1l2ln and s1s2sm take sum over all possible permutation of (l1, l2, …, ln) and (s1, s2, …, sm). And c(n, m) represents the same form as (26) in which replacement of (27) and (28) is made.

In (29), c(n, m) only the term

(30)(l1ln)(s1sm)(1)τ(l1,,ln)+τ(s1,,sm)[12θ(i,j=1i<jnailiajlj+i,j=1i<jmbisibjsj)+6(i,j=1ijnailiajlj,θ+i,j=1ijmbisibjsj,θ)], (30)

remain nonvanishing after taking sum over n! and m! permutations. This is seen as follows. Of total n! or m! permutations, there always exist equal number of even and odd permutations, thus we have the relation

(31)(l1ln)(s1sm)j<k(ailibjljbksk)(1)τ(l1,,ln)+τ(s1,,sm)=(l1ln)all even permutation of(s1sm)j<k[aili(bjsjbkskbjskbksj)](1)τ(l1,,ln)+τ(s1sn)=(l1ln)all even permutation of(s1sm)[ailiρjρsjρkρsk(12t)43(ωjωsjωkωskωjωskωkωsj)](1)τ(l1,,ln)+τ(s1sn)=0. (31)

In this way, all the terms in c(n, m) except (30) and ailiajljbksk vanish. As to the remaining two types of terms, the latter also vanishes as

(32)(l1ln)(s1sm){k=1mi,j=1ijn(ailiajljbksk)}(1)τ(l1,,ln)+τ(s1,,sm)=(l1ln)all even permutation of(s1sm)k=1mi,j=1ijn[ailiajjailjajlj+ajljailiajliailj]bksk(1)τ(l1,,ln)+τ(s1sn)=0. (32)

Thus, in this way, all terms in c(n, m) except (30) vanish. By the (20) and (24), (29) can be written to

(33)F|a11a1nan1ann||b11b1mbm1bmm|=l1l2ln,s1,s2,sm(1)τ(l1,,ln)+τ(s1,,sm)(12t)2(1+θ)6{i,j=1,i<jn[2(xixli+θiθli)(xjxlj+θjθlj)ailiajlj]+i,j=1,i<jm[2(xixsi+θiθsi)(xjxsj+θjθisi)bisibjsj]i=1nj=1m(xixj+θiθj)(xlixsj+θliθsj)aijblisj+i=1nj=1m(xixsj+θiθsj)(xjxli+θjθli)aisjbjli}. (33)

Especially

(34)F|a11a1nan1ann|1=l1ln(1)τ(l1ln)12(12t)2(1+θ){i,j=1i<jn(xixli+θiθli)(xjxlj+θjθlj)ailiajlj}. (34)

Several lowest orders of these are written explicitly as

(35)Fa111=0, (35)
(36)F|a11a12a21a22|1=12(12t)2(1+θ)(x1x2+θ1θ2)2a122, (36)
(37)Fa11a22=12(12t)2(1+θ)(x1x2+θ1θ2)2a122, (37)
(38)F|a11a12a13a21a22a23a31a32a33|1=121+θ)(12t)2{(x1x2+θ1θ2)2a122a33+(x1x3+θ1θ3)2a132a22+(x2x3+θ2θ3)2a232a112[(x1x2+θ1θ2)(x1x3+θ1θ3)+(x1x3+θ1θ3)(x2x3+θ2θ3)+(x1x2+θ1θ2)(x3x2+θ3θ2)]a12a13a23}, (38)
(39)F|a11a12a21a22|a33=121+θ)(12t)2{(x1x2+θ1θ2)2a122a33(x1x3+θ1θ3)2a132a22(x2x3+θ2θ3)2a232a11+2(x1x3+θ1θ3)(x2x3+θ2θ3)a12a13a23}, (39)
(40)F|a11a12a13a21a22a23a31a32a33|a11=121+θ)(12t)2(x2x3+θ2θ3)2(a12a13a11a23)2, (40)
(41)F|a11a12a21a22||a11a13a31a33|=121+θ)(12t)2(x2x3+θ2θ3)2(a12a13a11a23)2, (41)
(42)F|a11a12a21a22||a11a12a13a21a22a23a31a32a33|=0, (42)
(43)F|a11a12a13a21a22a23a31a32a33||a11a12a13a21a22a23a31a32a33|=0. (43)

From (35–43), we can obtain the one-, two-, and three-decay mode solutions as following

(44)f=1+a11,f=1+(a11+a22)+|a11a12a21a22|=|1+a11a12a211+a22|,f=1+(a11+a22+a33)+(|a11a12a21a22|+|a22a23a32a33|+|a33a31a13a11|)+|a11a12a13a21a22a23a31a32a33|=|1+a11a12a13a211+a22a23a31a321+a33|. (44)

Generally, the N decay mode solution can be shown here

(45)f=|1+a11a12a1na211+a22a2nan1an21+ann|. (45)

Award Identifier / Grant number: 11301183

Funding statement: This work was supported by the National Natural Science Foundation of China under Grant No. 11301183.

Acknowledgments:

This work was supported by the National Natural Science Foundation of China under Grant No. 11301183.

References

[1] P. Mathieu, J. Math. Phys. 29, 2499 (1988).10.1063/1.528090Search in Google Scholar

[2] Y. I. Manin and A. O. Radul, Commun. Math. Phys. 98, 65 (1985).10.1007/BF01211044Search in Google Scholar

[3] C. M. Yung, Phy. Lett. B 309, 75 (1993).10.1016/0370-2693(93)91506-ISearch in Google Scholar

[4] M. Chaichain and P. Kulish, Phys. Lett. B 78, 413 (1978).10.1016/0370-2693(78)90473-2Search in Google Scholar

[5] A. Ibort, L. M. Alonso, and E. M. Reus, J. Math. Phys., 37, 6157 (1996).10.1063/1.531770Search in Google Scholar

[6] W. Oevel and Z. Popowicz, Commun. Math. Phys. 139, 441 (1991).10.1007/BF02101874Search in Google Scholar

[7] Q. P. Liu and M. Manas, solv-int/9806005.Search in Google Scholar

[8] A. S. Carstea, Nonlinearity 13, 1645 (2000).10.1088/0951-7715/13/5/312Search in Google Scholar

[9] R. Hirota, Direct Methods in Soliton Theory, Springer-Verlag, Berlin 2004.10.1017/CBO9780511543043Search in Google Scholar

[10] X. B. Hu, C. X. Li, J. J. C. Nimmo, and G. F. Yu, J. Phys. A: Math. Gen. 38, 195 (2005).10.1088/0305-4470/38/1/014Search in Google Scholar

[11] Q. P. Liu and Y. F. Xie, Phys. Lett. A 325, 139 (2004).10.1016/j.physleta.2004.03.047Search in Google Scholar

[12] Q. P. Liu, X. B. Hu, and M. X. Zhang, Nonlinearity 18, 1597 (2005).10.1088/0951-7715/18/4/009Search in Google Scholar

[13] S. Ghosh and D. Sarma, J. Nonlinear Math. Phys. 10, 526 (2003).10.2991/jnmp.2003.10.4.6Search in Google Scholar

[14] S. Maxon and J. Viecelli, Phys. Fluids 17, 1614 (1974).10.1063/1.1694941Search in Google Scholar

[15] A. Nakamura and H. H. Chen, J. Phys. Soc. Japan 50, 711 (1981).10.1143/JPSJ.50.711Search in Google Scholar

Received: 2016-3-18
Accepted: 2016-4-26
Published Online: 2016-5-26
Published in Print: 2016-7-1

©2016 by De Gruyter

Downloaded on 19.12.2025 from https://www.degruyterbrill.com/document/doi/10.1515/zna-2016-0102/html
Scroll to top button