This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

11institutetext: School of Physics, Peking University, Beijing 100871, China
School of Physics, Xi’an Jiaotong University, Xi’an 710049, China

Kink scattering in a Lorentz-violating ϕ6\phi^{6} model

Haobo Yan111E-mail: [email protected] (corresponding author)
Abstract

The role played by a Lorentz-violating term on the outcomes of kink scattering in the ϕ6\phi^{6} model is investigated by using the Fourier spectral method. Impacts of the Lorentz-violating term on the critical velocities, the location of two-bounce windows, and the maximal values of various types of energy densities are analyzed. Some novel features of kink-antikink collisions are discussed. The interactions between three and four kinks are also considered.

1 Introduction

In (1+1)(1+1)-dimensional nonlinear scalar field models with more than one degenerated vacua, there may exist a type of classical solution called kink, which smoothly connects different vacua and has localized energy density. In the past decades, kink and its higher-dimensional extension, domain wall, have been extensively studied in both condensed matter physics and high energy physics [1].

A related topic that has been repeatedly discussed recently, is the scattering of non-integrable kinks. Unlike the collision of integrable kinks where the kinks simply pass through each other, a pair of incoming non-integrable kinks (for instance, ϕ4\phi^{4} kinks) can be bounced back after a finite time of the collision or form a bound state called bion/oscillon, depending on their initial velocities, say v0v_{0} and v0-v_{0} for symmetric collisions [2, 3, 4, 5]. The velocity intervals within which kinks are bounced back after nn collisions are called nn-bounce windows (nnBWs). For the ϕ4\phi^{4} model, the nn-bounce windows form an interesting fractal structure [6].

Bounce windows have also been found in many other non-integrable kink models [7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]. A well-accepted explanation for the appearance of bounce windows is the resonant energy exchange between the translational and vibrational modes of the kink [5, 9]. This idea is shown recently using a collective coordinate model [26]. In addition to double kink collisions, the simultaneous collisions of multi-kink at the same point have also been investigated in many works [27, 28, 29, 30, 31, 32]. Such collisions can produce a very large energy density around the collision point, and the properties of the high energy region are important for predicting the maximal number of kink-antikink pairs formed in wave train collisions [33]. In some models [27, 34], one can also pinpoint the location of bounce windows by observing the behaviors of maximal energy densities, which vary smoothly within the bounce windows but become chaos elsewhere. Other interesting topics include deformed ϕ4\phi^{4} theory [17], deformed sine-Gordon model[35], etc.

Despite many interesting discoveries, non-integrable kink collisions are rarely considered in Lorentz-violating models. Considering the fact that Lorentz symmetry breaking naturally arises in many fundamental theories, such as canonical and loop quantum gravity [36, 37, 38], noncommutative quantum field theory [39], and might have observational implications [40, 41] (see [42] for more references), it would be interesting to consider kink collisions in Lorentz-violating non-integrable scalar field models. However, the breaking of Lorentz symmetry hinders one from deriving moving kink solutions by simply boosting static ones, which are necessary for the construction of the initial configuration. Therefore, the study of kink collisions was limited to the models that preserve the Lorentz invariance for a long time, until Bazeia and others proposed a model in which there is a simple transformation relationship between the static and dynamic kink solutions [43].

In this paper, we study the kink interactions in a Lorentz-violating ϕ6\phi^{6} model using the traveling kink solution reported in Ref. [43]. In the following section, the Lorentz-violating ϕ6\phi^{6} model is considered and the kink solutions in the Lorentz-violating ϕ6\phi^{6} model are derived. Then we analyze the linear perturbation modes of both the single kink and double kink solutions. The numerical simulation of the kink collisions, as well as the dependencies of several physical quantities on the Lorentz-violating parameter, is conducted subsequently.

2 Model and solution

The (1+1)(1+1)-dimensional Lorentz-violating ϕ6\phi^{6} theory is defined by the Lagrangian density:

=12ημνμϕνϕ+12κμνμϕνϕ12ϕ2(1ϕ2)2,\displaystyle\mathcal{L}=\frac{1}{2}\eta^{\mu\nu}\partial_{\mu}\phi\partial_{\nu}\phi+\frac{1}{2}\kappa^{\mu\nu}\partial_{\mu}\phi\partial_{\nu}\phi-\frac{1}{2}\phi^{2}(1-\phi^{2})^{2}, (1)

where we introduced the conventional Minkowski metric ημν(1001)\eta^{\mu\nu}\equiv\begin{pmatrix}1&0\\ 0&-1\end{pmatrix} and a tensor κμν(0αα0)\kappa^{\mu\nu}\equiv\begin{pmatrix}0&\alpha\\ \alpha&0\end{pmatrix}, representing the Lorentz violation.

Since the ϕ6\phi^{6} model contains three degenerate vacua {1,0,1}{\{-1,0,1\}}, kinks and antikinks take two types, each of which is connected to neighboring vacua and has to be studied separately.

The Lagrangian (Eq. 1) yields the EoM for the field

2ϕt22ϕx2+2α2ϕxt+3ϕ54ϕ3+ϕ=0\frac{\partial^{2}\phi}{\partial t^{2}}-\frac{\partial^{2}\phi}{\partial x^{2}}+2\alpha\frac{\partial^{2}\phi}{\partial x\partial t}+3\phi^{5}-4\phi^{3}+\phi=0 (2)

The static kink solution interpolating (0,1)(0,1) is found to be the same as the ordinary ϕ6\phi^{6} model solution

ϕK(x)=ϕs(0,1)(x)=1+tanh(x)2.\phi_{K}(x)=\phi_{s(0,1)}(x)=\sqrt{\frac{1+\tanh(x)}{2}}. (3)

The corresponding antikink is

ϕK¯(x)=ϕ(1,0)(x)=ϕK(x),\phi_{\bar{K}}(x)=\phi_{(1,0)}(x)=\phi_{K}(-x), (4)

and kinks belong to the other topological sector are

ϕ(0,1)(x)=ϕK(x),ϕ(1,0)(x)=ϕK(x).\phi_{(0,-1)}(x)=-\phi_{K}(x),\phi_{(-1,0)}(x)=-\phi_{K}(-x). (5)

To obtain a moving kink solution, one has to resort to a more sophisticated method, for the ordinary Lorentz boost simply does not hold in this Lorentz-violating model. However, one can still carry out a boost-like approach by deforming the Lorentz factor γ1/1v2+2αv\gamma\equiv 1/\sqrt{1-v^{2}+2\alpha v} [43]. The traveling kink solution can then be expressed as

ϕ(0,1)(x0,t)=1+tanh[γ(xx0v0t)]2,\phi_{(0,1)}(x_{0},t)=\sqrt{\frac{1+\tanh[\gamma(x-x_{0}-v_{0}t)]}{2}}, (6)

where x0x_{0} and v0v_{0} are the initial position and velocity of the kink.

As is known, the breaking of the Lorentz invariance breaks the symmetry (the width, the energy density, etc) between moving kink and antikink, and kink collisions would be asymmetrical. The spatial configurations of kink-antikink (KK¯K\bar{K}) and antikink-kink (K¯K\bar{K}K) pairs at α=0,1,2\alpha=0,1,2 are shown in Fig. 1. As α\alpha increases, the asymmetry between the kink and the antikink increases. Note that the energy density distribution of the same kink is also different if it is moving oppositely, as indicated by the dashed lines.

Refer to caption
Figure 1: Configurations of KK¯K\bar{K} and K¯K\bar{K}K topology for v0=0.2v_{0}=0.2, x0=10x_{0}=10, and α=0,1,2\alpha=0,1,2. The antikinks are always more energetic than kinks, and this asymmetry grows as α\alpha increases. The dashed lines show the difference of maximal energy densities between right-moving and left-moving (anti)kinks.

3 Linear stability and the vibrational mode

To prove the stability of the kink solutions and get a better understanding of the bounce windows, the linear perturbative states of the static kink solutions should be examined. Consider a small perturbation δϕ(x,t)\delta\phi(x,t) around a background (a static kink or a two-kink configuration), by taking the variation of the action with respect to δϕ(x,t)\delta\phi(x,t) and Fourier expand the field ϕ(x,t)\phi(x,t), we immediately arrive at a Schrödinger-like equation:

fn=ω~n2fn,\mathscr{H}f_{n}=\tilde{\omega}_{n}^{2}f_{n}, (7)

where a factorizable Hamiltonian d2dx2+1θd2θdx2=(ddx+1θdθdx)(ddx+1θdθdx)\mathscr{H}\equiv-\frac{{\mathrm{d}}^{2}}{{\mathrm{d}}x^{2}}+\frac{1}{\theta}\frac{{\mathrm{d}}^{2}\theta}{{\mathrm{d}}x^{2}}=\left(\frac{{\mathrm{d}}}{{\mathrm{d}}x}+\frac{1}{\theta}\frac{{\mathrm{d}}\theta}{{\mathrm{d}}x}\right)\left(-\frac{{\mathrm{d}}}{{\mathrm{d}}x}+\frac{1}{\theta}\frac{{\mathrm{d}}\theta}{{\mathrm{d}}x}\right) (with θ=xϕs\theta=\partial_{x}\phi_{s}) is introduced, and ω~n2(1+α2)ωn2\tilde{\omega}_{n}^{2}\equiv\left(1+\alpha^{2}\right)\omega_{n}^{2} is then always non-negative according to the supersymmetric quantum mechanics, thus proving our claim that the static solutions are stable against small perturbations. Note that fn(x)f_{n}(x) is only the expansion coefficient of perturbation, and gn(x)fn(x)eiαωnxg_{n}(x)\equiv f_{n}(x)\mathrm{e}^{{i}\alpha\omega_{n}x} is the complete eigenfunction.

For the single kink solution ϕs(0,1)(x)=1+tanh(x)2\phi_{s(0,1)}(x)=\sqrt{\frac{1+\tanh(x)}{2}}, the effective potential in the Schrödinger-like equation is Veff=x3ϕsxϕs=154(tanh(x)+1)26(tanh(x)+1)+1V_{\mathrm{eff}}=\frac{\partial_{x}^{3}\phi_{s}}{\partial_{x}\phi_{s}}=\frac{15}{4}(\tanh(x)+1)^{2}-6(\tanh(x)+1)+1 and only translational mode survives.

For the two-kink static configurations,

{ϕKK¯(x,0)=ϕK¯(xx0)+ϕK(x+x0)1,ϕK¯K(x,0)=ϕK(xx0)+ϕK¯(x+x0),\begin{cases}\phi_{K\bar{K}}(x,0)=\phi_{\bar{K}}(x-x_{0})+\phi_{K}(x+x_{0})-1,\\ \phi_{\bar{K}K}(x,0)=\phi_{K}(x-x_{0})+\phi_{\bar{K}}(x+x_{0}),\end{cases} (8)

the effective potential becomes Veff2K=15ϕ2K412ϕ2K2+1V_{\mathrm{eff}}^{2K}=15\phi_{2K}^{4}-12\phi_{2K}^{2}+1, where ϕ2K\phi_{2K} denotes either KK¯K\bar{K} or K¯K\bar{K}K pair. It turns out that the KK¯K\bar{K} pair still supports only translational mode, but K¯K\bar{K}K configuration supports many of the vibrational modes, just as the case in the ordinary ϕ6\phi^{6} model [9]. The eigenvalues ω~n2\tilde{\omega}_{n}^{2} depend both on the half-separation x0x_{0} and the Lorentz-violating parameter α\alpha. The eigenfunctions gn(x)g_{n}(x) for the first five modes of K¯K\bar{K}K pair with half-separation x0=10x_{0}=10 (1414 modes in total) are shown in Fig. 2. The α\alpha-dependence of the vibrational modes indicates that the kink collision results will be affected by the Lorentz violation. Various novel kink-collision phenomena will be discussed in the following sections.

Refer to caption
(a) α=0\alpha=0
Refer to caption
(b) α=1\alpha=1
Refer to caption
(c) α=2\alpha=2
Figure 2: Eigenfunctions of modes of K¯K\bar{K}K static solution for α=0\alpha=0 (ordinary ϕ6\phi^{6} model) and α=1,2\alpha=1,2. 0n40\leq n\leq 4 stands for the zero mode and the nthn\textsuperscript{th} excited state, and x0=10x_{0}=10 is the half-separation distance. The spatial wave function gn(x)fn(x)eiαωnxg_{n}(x)\equiv f_{n}(x)e^{i\alpha\omega_{n}x} gets more nodes under Lorentz violation.

4 Simulation and results

In this section, we study the asymmetrical kink collisions. Due to the absence of exact kink scattering solutions of the non-integrable model, one has to resort to numerical simulation. The initial conditions of kink scattering are set by superpositions of individual kinks and antikinks, which is Eq. 8 and

{ϕ˙KK¯(x,0)=ϕ˙K¯(xx0)+ϕ˙K(x+x0),ϕ˙K¯K(x,0)=ϕ˙K(xx0)+ϕ˙K¯(x+x0).\begin{cases}\dot{\phi}_{K\bar{K}}(x,0)=\dot{\phi}_{\bar{K}}(x-x_{0})+\dot{\phi}_{K}(x+x_{0}),\\ \dot{\phi}_{\bar{K}K}(x,0)=\dot{\phi}_{K}(x-x_{0})+\dot{\phi}_{\bar{K}}(x+x_{0}).\end{cases} (9)

Since a single kink has an exponentially decaying tail, the overlap in the initial conditions can be safely ignored at a large distance. Initial conditions for more kinks can be obtained in the same way.

The Fourier spectrum method [44] with periodic boundary conditions is used to solve the EoM numerically, and one may control the validity of the simulation by insisting on energy conservation. The total energy of the configuration is calculated by

E(t)=12[(ϕt)2+(ϕx)2+12ϕ2(1ϕ2)2]dx,E(t)=\frac{1}{2}\int_{-\infty}^{\infty}\left[\left(\frac{\partial\phi}{\partial t}\right)^{2}+\left(\frac{\partial\phi}{\partial x}\right)^{2}+\frac{1}{2}\phi^{2}\left(1-\phi^{2}\right)^{2}\right]\mathrm{d}x, (10)

and can be approximated by

EthE(v0)+E(v0),E_{\mathrm{th}}\approx E\left(v_{0}\right)+E\left(-v_{0}\right), (11)

where E(v0)=γ(1+αv0)MKE\left(v_{0}\right)=\gamma\left(1+\alpha v_{0}\right)M_{K} is the energy of a moving kink and the mass of the kink is MK=[(xϕ(0,1))2/2+V]𝑑x=14M_{K}=\int[(\partial_{x}\phi_{(0,1)})^{2}/2+V]dx=\frac{1}{4}. The relative error between EthE_{\mathrm{th}} and the calculated energy EnumE_{\mathrm{num}} will be checked to guarantee the validity of simulation results. The EoM is discretized, with the time step chosen automatically by MATLAB ode45 solver, typically around 0.0020.002, to ensure the relative error of energy to be less than 10910^{-9}. The space step is chosen to be 0.10.1 and 0.050.05 to guarantee the results are convergent and no significant change happens with further precision improvement.

4.1 Impacts on the collision structure

It is important to look at how the Lorentz-violating term affects the global distribution of BW and vcv_{c}. The KK¯K\bar{K} and K¯K\bar{K}K-type collision structures for α=0,0.5,1\alpha=0,0.5,1 are shown in Fig. 3. As α\alpha increases, the critical velocity decreases. The critical difference between this structure and that of the ϕ4\phi^{4} model is no bounce window is found in KK¯K\bar{K} collisions, and no bounce window higher than the 2nd2\textsuperscript{nd} order in K¯K\bar{K}K collisions is found, rendering the structure of the Lorentz-violating ϕ6\phi^{6} model terminated after the second level, which is consistent with the ordinary ϕ6\phi^{6} model [9].

One can make another two observations regarding the oscillations in the KK¯K\bar{K} structure, as shown in Fig. 3(b). First, the moments of nthn\textsuperscript{th} oscillations of the bions plot a series of curves as functions of v0v_{0}. As v0v_{0} increases, the curves decrease in general, but there exist certain windows where all the curves start to rise, which is called the rising windows (RW). The existence of RWs is also confirmed in Lorentz-violating collisions. Furthermore, by zooming in on the windows one would find the RWs decompose into smaller increasing and decreasing segments. The first two RWs are highlighted, and the first is enlarged on the right. Second, in the ordinary ϕ6\phi^{6} model, the amplitudes of the oscillations are relatively preserved, while the presence of Lorentz violation damps the oscillation. This damping effect is most evident near the critical velocity, where the amplitudes of oscillations quickly decay to near zero. The field values at the origin of v0=0.15v_{0}=0.15 for α=0,0.5,1\alpha=0,0.5,1 are plotted in Fig. 4 to illustrate the damping effect. We emphasize that this is due to the asymmetry of the Lorentz-violating KK¯K\bar{K} collisions. The bions formed from the collisions deviate from x=0x=0 as time evolves, so the field values at x=0x=0 come mainly from radiations instead of the bions themselves. However, the bions formed in the K¯K\bar{K}K collisions are not found to deviate from the collision points.

Refer to caption
(a) K¯K\bar{K}K
Refer to caption
(b) KK¯K\bar{K}
Figure 3: Field values of collisions of (a) K¯K\bar{K}K and (b) KK¯K\bar{K} at the central spatial point as functions of velocities for α=0,0.5\alpha=0,0.5 and 11. As α\alpha increases, the critical velocity decreases. There is no 3BW and higher-order bounce window in this figure. Collision points at initial velocities v0=0.03315v_{0}=0.03315 have been colored white in plot (a). In plot (b), there is no BW, but there exist windows (RW) where the nthn\textsuperscript{th} oscillation moments increases as v0v_{0} increases, with the first two RWs highlighted and the first RW enlarged on the right.
Refer to caption
Figure 4: The field values at the origin of v0=0.15v_{0}=0.15 for α=0,0.5,1\alpha=0,0.5,1 of KK¯K\bar{K} collisions. The amplitudes of oscillations decay if α0\alpha\neq 0.

The typical numerical results with v0=0.03315v_{0}=0.03315 for antikink-kink collisions have been selected in Fig. 3(a) and shown in Fig. 5. The first row of figures shows all 3 types of ϕ(x,t)\phi(x,t) configurations, namely bion, two-bounce, and inelastic scattering. The second row shows the corresponding energy density configurations, where the asymmetrical nature of collisions is clearly shown by the behavior of radiation. Note that albeit the asymmetry, the scattering still happens at x=0x=0, which can be seen in this row of energy density plots and in Fig. 3 where the amplitudes of the first oscillation at x=0x=0 are almost the same for α=0,0.5\alpha=0,0.5, and 11. The field values at the origin of the first three 2BWs (with the second 2BW being a false window) at α=1\alpha=1 are presented in Fig. 6. The number nn of oscillations of ϕ(0,t)\phi(0,t) between two collisions is the same as that of the ordinary ϕ6\phi^{6} model [9], which together with Fig. 3 means that the Lorentz violation may change the positions of 2BWs and vcv_{c}, but preserve certain invariant quantities, i.e. the number of collective mode oscillations and the number of 2BWs.

Refer to caption
Refer to caption
Refer to caption
Refer to caption
(a) α=0\alpha=0, bion.
Refer to caption
(b) α=0.5\alpha=0.5, two-bounce.
Refer to caption
(c) α=1\alpha=1, inelastic scattering.
Figure 5: The numerical results for antikink-kink collision at v0=0.03315v_{0}=0.03315. The first row shows the field configuration ϕ(x,t)\phi(x,t), and the second row shows the energy density ρ(x,t)\rho(x,t). The left, middle and right columns are results of α=0\alpha=0, α=0.5\alpha=0.5, and α=1\alpha=1, and correspond to the bound state, two-bounce window, and inelastic scattering solutions, respectively.
Refer to caption
Figure 6: The field values at the origin of the first three 2BWs (including a false window) of K¯K\bar{K}K collisions at α=1\alpha=1. The numbers of oscillations are the same as those at α=0\alpha=0.

The critical velocity of KK¯K\bar{K} and K¯K\bar{K}K collisions in the Lorentz-violating ϕ6\phi^{6} model are both monotonically decreasing functions of the parameter α\alpha, as shown in the data points of Fig. 7. This is just the case in the ϕ4\phi^{4} model. And it is of great interest if these results are compared with those of the Lorentz-violating ϕ4\phi^{4} model. Despite being calculated from different scalar models or different topologies, the data of the critical velocities present similar varying trends. The following empirical formula is introduced to fit the data:

v(α)=v(0)eτα1+α2,v(\alpha)=v(0)\frac{\mathrm{e}^{-\tau\alpha}}{\sqrt{1+\alpha^{2}}}, (12)

and the corresponding fitted curves are also shown in Fig. 7.

To explain their similar behavior under Lorentz invariance breaking, notice that the energies of the kink and the antikink are no longer the same (see Fig. 1). The excessive energy of the kink (antikink) is then wasted, as a result, higher α\alpha means lower effective colliding velocity. Eq. 12 then provides a prediction for the critical velocities under Lorentz violation.

Refer to caption
Figure 7: The critical velocity vcv_{c} of ϕ4\phi^{4} and ϕ6\phi^{6} model collisions and reconstructed fitting curves as functions of the Lorentz-violating parameter α[0,2]\alpha\in[0,2]. The fitted decaying parameters τ\tau are 0.046,0.290,0.2600.046,0.290,0.260 for K¯K\bar{K}K collision, KK¯K\bar{K} collision, and ϕ4\phi^{4} collision, respectively.

The dependence of the maximal energy densities on α\alpha provides rich information about the kink-scattering behavior under Lorentz breaking. The total energy density can be written as

ρ(x,t)=k(x,t)+u(x,t)+p(x,t),\rho(x,t)=k(x,t)+u(x,t)+p(x,t), (13)

where k(x,t)=12(ϕt)2k(x,t)=\frac{1}{2}\left(\frac{\partial\phi}{\partial t}\right)^{2}, u(x,t)=12(ϕx)2u(x,t)=\frac{1}{2}\left(\frac{\partial\phi}{\partial x}\right)^{2}, and p(x,t)=V(ϕ)p(x,t)=V(\phi) are the kinetic energy density, gradient energy density, and potential energy density, respectively.

From the experience of Ref. [27, 34], one would expect this energy density structure to be chaotic unless the velocity falls in the inelastic scattering region or the bounce windows, showing fractal information about the scattering structure. However, this is not the case anymore for the ϕ6\phi^{6} model. In Fig. 8, maximal values of ρ(x,t)\rho(x,t) and its components for kink scattering, with initial conditions x0=10x_{0}=10 and v0=0.03315v_{0}=0.03315, are presented. The curve is smooth and no chaotic zone is observed, regardless of the final state of the configuration. The lacking of the single vibrational mode forbids the violent energy exchange between different energy levels, the maximal energy densities then lost their fractal chaotic behavior.

Refer to caption
Figure 8: Maximal values of ρ(x,t)\rho(x,t) and its components for v0=0.03315v_{0}=0.03315 kink scattering as functions of α\alpha.

4.2 Impacts on multi-kink collision

The authors in Ref. [28] have studied the multi-kink collision of the ordinary ϕ6\phi^{6} model in great detail. And it would be of interest to study how the results depend on the Lorentz violation. We follow the initial value settings in Ref.[28] to guarantee the collisions occur at the same point when α=0\alpha=0 (the upper row in Fig. 9). The parameter α\alpha is then slowly increased to study the impact of small Lorentz violation (the lower row in Fig. 9 shows the results of α=0\alpha=0).

Refer to caption
Refer to caption
Refer to caption
Refer to caption
(a) (0,1,0,1)(0,1,0,1)
Refer to caption
(b) (1,0,1,0)(1,0,1,0)
Refer to caption
(c) (0,1,0,1,0)(0,1,0,1,0)
Figure 9: The spacetime configuration for three- and four-kink collisions. The captions indicate the topology of the multi-kinks. The first row and the second row are fields for α=0\alpha=0 and α=0.1\alpha=0.1, respectively.

The three-kink collisions of topology KK¯KK\bar{K}K are presented in Fig. 9(a) and Fig. 9(b). In the ordinary ϕ6\phi^{6} model, the field ends up with a right-moving kink and a small-amplitude left-moving bion. When α\alpha is turned on, the velocity of the kink and bion is decreased, whilst the amplitude of the bion is increased. In this case, part of the kinetic energy is transferred to the potential energy of the bion. The K¯KK¯\bar{K}K\bar{K} collisions are similar to those of KK¯KK\bar{K}K, in the sense of forming bions. The four-kink collisions of KK¯KK¯K\bar{K}K\bar{K} are shown in Fig. 9(c). When α=0\alpha=0, the four kinks collide and end up with two kinks and radiation in between. When α=0.1\alpha=0.1, the collision creates two bions, and the bions travel slower than the kinks at α=0\alpha=0. This tendency of forming bions is valid for small Lorentz violation parameters (α<0.1\alpha<0.1). In this range, the kinks give more energy to the bions at higher α\alpha.

It is now clear that a small α\alpha could lead to a very different final state, especially for the multi-kink collisions. In the example presented in Fig. 9, a small Lorentz violation tends to suppress the speed of kinks and bions and uses the energy to build bions or make them stronger.

5 Conclusions

In this study, we were concerned mainly with the kink scattering of a Lorentz-violating ϕ6\phi^{6} model. The static and boosted kink solutions were obtained in the spirit of Ref. [43]. It was found that with Lorentz invariance broken, traveling kinks and antikinks became asymmetrical. The following analysis proved the stability of kink solutions. The eigenfunctions of the K¯K\bar{K}K linear perturbation for α=0,1,2\alpha=0,1,2, 0n40\leq n\leq 4 and separation x0=10x_{0}=10 were calculated, as shown in 2.

After the general remarks, the asymmetrical kink collisions are studied for both the KK¯K\bar{K} and (K¯K)(\bar{K}K) configurations. The Lorentz-violating ϕ6\phi^{6} model has no BW for KK¯K\bar{K} collisions and no 3BW or higher bounce windows for K¯K\bar{K}K collisions. The structure in Fig. 3 shows how vcv_{c} decreases due to the presence of α\alpha. We observed rising windows in the structure of KK¯K\bar{K} scatterings, and the damping of the field value at the origin, or equivalently, the deviation of bions from the origin under Lorentz violation. To the author’s knowledge, these two effects have not been reported yet.

We also discovered an empirical formula Eq. 12 of vcv_{c} dependencies on α\alpha for the two topologies in the ϕ6\phi^{6} model and that of the ϕ4\phi^{4} model. One can study higher-order Lorentz-violating models to generalize the formula. We then studied the impacts of α\alpha on maximal energy densities. Due to the lacking of different modes of single kinks, the dependencies have no intricate structure as that of the ϕ4\phi^{4} model.

The impact of Lorentz violation on the multi-kink collisions was also studied. We found that the presence of small α\alpha would change the collision process. In those examples, the kinetic energy is transferred to forming bions or enlarging their amplitude.

6 Acknowledgements

The author is grateful to Yuan Zhong for helpful discussions.

References

  • [1] \NameVachaspati T. \BookKinks and Domain Walls: An Introduction to Classical and Quantum Solitons (Cambridge University Press) 2006.
  • [2] \NameKudryavtsev A. E. \REVIEWJETP Lett. (USSR) (Engl. Transl.), v. 22, no. 3, pp. 82-831975.
  • [3] \NameSugiyama T. \REVIEWProgress of Theoretical Physics6119791550.
  • [4] \NameMoshir M. \REVIEWNuclear Physics B1851981318 .
  • [5] \NameCampbell D. K., Schonfeld J. F. Wingate C. A. \REVIEWPhysica D: Nonlinear Phenomena919831 .
  • [6] \NameAnninos P., Oliveira S. Matzner R. A. \REVIEWPhys. Rev. D4419911147.
  • [7] \NameAntunes N. D., Copeland E. J., Hindmarsh M. Lukas A. \REVIEWPhys. Rev. D692004065016.
  • [8] \NameHoseinmardy S. Riazi N. \REVIEWInternational Journal of Modern Physics A2520103261.
  • [9] \NameDorey P., Mersh K., Romanczukiewicz T. Shnir Y. \REVIEWPhys. Rev. Lett.1072011091602.
  • [10] \NameGomes A. R., Menezes R., Nobrega K. Z. Simas F. C. \REVIEWPhys. Rev. D902014065022.
  • [11] \NameKhare A., Christov I. C. Saxena A. \REVIEWPhys. Rev. E902014023208.
  • [12] \NameArthur R., Dorey P. Parini R. \REVIEWJournal of Physics A: Mathematical and Theoretical492016165205.
  • [13] \NameLima F. C., Simas F. C., Nobrega K. Z. Gomes A. R. \REVIEWJournal of High Energy Physics20192019147.
  • [14] \NameBelendryasova E. Gani V. A. \REVIEWCommunications in Nonlinear Science and Numerical Simulation672019414 .
  • [15] \NameChristov I. C., Decker R. J., Demirkaya A., Gani V. A., Kevrekidis P. G. Radomskiy R. V. \REVIEWPhys. Rev. D992019016010.
  • [16] \NameKhare A. Saxena A. \REVIEWJournal of Physics A: Mathematical and Theoretical522019365401.
  • [17] \NameAdam C., Oles K., Romanczukiewicz T. Wereszczynski A. \REVIEWPhys. Rev. E1022020062214.
  • [18] \NameMohammadi M. Dehghani R. \REVIEWCommunications in Nonlinear Science and Numerical Simulation942021105575.
  • [19] \NameAlonso-Izquierdo A., Queiroga-Nunes J. Nieto L. M. \REVIEWPhys. Rev. D1032021045003.
  • [20] \NameNzoupe F. N., Dikandé A. M. Tchawoua C. \REVIEWMod. Phys. Lett. A3620212150015.
  • [21] \NameDorey P., Gorina A., Perapechka I., Romańczukiewicz T. Shnir Y. \REVIEWJournal of High Energy Physics20212021.
  • [22] \NameCampos J. G. F. Mohammadi A. \REVIEWJournal of High Energy Physics20212021.
  • [23] \NameBazeia D., Gomes A. R., Nobrega K. Simas F. C. \REVIEWPhysics Letters B8032020135291.
  • [24] \NameZhong Y., Du X.-L., Jiang Z.-C., Liu Y.-X. Wang Y.-Q. \REVIEWJournal of High Energy Physics20202020153.
  • [25] \NameChristov I. C., Decker R. J., Demirkaya A., Gani V. A., Kevrekidis P. Saxena A. \REVIEWCommunications in Nonlinear Science and Numerical Simulation972021105748.
  • [26] \NameManton N. S., Oleś K., Romańczukiewicz T. Wereszczyński A. \REVIEWPhys. Rev. Lett.1272021071601.
  • [27] \NameGani V. A., Moradi Marjaneh A. Saadatmand D. \REVIEWThe European Physical Journal C792019620.
  • [28] \NameMoradi Marjaneh A., Gani V. A., Saadatmand D., Dmitriev S. V. Javidan K. \REVIEWJournal of High Energy Physics2017201728.
  • [29] \NameEkomasov E. G., Gumerov A. M., Kudryavtsev R. V., Dmitriev S. V. Nazarov V. N. \REVIEWBrazilian Journal of Physics482018576.
  • [30] \NameMarjaneh A. M., Askari A., Saadatmand D. Dmitriev S. V. \REVIEWThe European Physical Journal B91201822.
  • [31] \NameSaadatmand D., Dmitriev S. V. Kevrekidis P. G. \REVIEWPhys. Rev. D922015056005.
  • [32] \NameGani V. A., Moradi Marjaneh A. Javidan K. \REVIEWThe European Physical Journal C812021.
  • [33] \NameAskari A., Saadatmand D., Dmitriev S. V. Javidan K. \REVIEWWave Motion78201854 .
  • [34] \NameYan H., Zhong Y., Liu Y.-X. ichi Maeda K. \REVIEWPhysics Letters B8072020135542.
  • [35] \NameMukhopadhyay M., Sfakianakis E. I., Vachaspati T. Zahariade G. \BookKink-antikink scattering in a quantum vacuum (2021).
  • [36] \NameAmelino-Camelia G., Ellis J., Mavromatos N. E., Nanopoulos D. V. Sarkar S. \REVIEWNature3931998763.
  • [37] \NameGambini R. Pullin J. \REVIEWPhys. Rev. D591999124021.
  • [38] \NameAlfaro J., Morales-Técotl H. A. Urrutia L. F. \REVIEWPhys. Rev. Lett.8420002318.
  • [39] \NameAnisimov A., Banks T., Dine M. Graesser M. \REVIEWPhys. Rev. D652002085032.
  • [40] \NameColladay D. Kostelecký V. A. \REVIEWPhys. Rev. D581998116002.
  • [41] \NameKostelecký V. A. \REVIEWPhys. Rev. D692004105009.
  • [42] \NameBietenholz W. \REVIEWPhysics Reports5052011145 .
  • [43] \NameBarreto M. N., Bazeia D. Menezes R. \REVIEWPhys. Rev. D732006065015.
  • [44] \NameTrefethen L. N. \BookSpectral methods in MATLAB (SIAM) 2000.