betaFeedback

WorksPhysics and Astronomy

Enhancing Gravitational Waves from Global Strings via a First-Order Phase Transition

Abstract

We perform the first (3+1)-dimensional lattice simulations of a cosmological first-order phase transition (PT) in which true-vacuum bubbles with random phases nucleate, expand, and collide, producing a network of global strings. We find that the string network enters the scaling regime, and the coefficient of the logarithmic growth is smaller than the value obtained for a second-order PT. From the axion spectrum, we derive the dark-matter relic abundance and the dark-radiation bound from big bang nucleosynthesis on the axion decay constant faf_a and mass mam_a. The ratio of the gravitational wave emission rate to the axion emission rate is found to be about two orders of magnitude larger than in the second-order case, and the resulting gravitational wave spectrum can be probed by pulsar timing arrays as well as by space- and ground-based interferometers.

Introduction

The pulsar timing array (PTA) collaborations have reported possible evidence for a stochastic gravitational-wave (GW) background in the nHz frequency band [1, 2, 3, 4], offering a potential insight into detecting early-universe processes. In this context, cosmic strings, which are topological defects formed by the spontaneous breaking of a U(1)U(1) symmetry [5, 6], can provide a viable explanation for the reported PTA signal [7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]. Moreover, the axion emitted during cosmic string evolution, as a pseudo-Nambu-Goldstone boson arising from the breaking of the Peccei-Quinn (PQ) symmetry invoked to solve the strong CP problem [18, 19, 20, 21, 22, 23, 24, 25], is also an attractive dark matter candidate [26, 27, 28].

Both axion and GW emissions are essential ingredients in studies of string-network evolution for axion strings [29, 30]. The scaling property and free axion emission in the string network evolution have been studied with the method of lattice field simulation for the second order PT process in recent years [31, 32, 33, 34, 35, 36, 37]. The GW emission has been investigated for string loops [38] and for the full string network [39, 40, 41], with a characteristic suppression factor of fa2/ma,pl2f_a^2/m_{a,\mathrm{pl}}^2. Apart from the second-order PT predicted by the Kibble mechanism [42, 43], it was proposed that the first-order PT can also produce the cosmic string through the geodesic rule when bubbles with random phases collide with each other [44, 45, 46]. Early simulations in (2+1)(2+1)-dimensional spacetime investigated vortex formation in first-order PTs [47, 48, 49, 50], focusing on the string production rate from bubble collisions and its dependence on parameters such as the bubble wall velocity.

This paper performs the first (3+1)(3+1)-dimensional lattice simulations to study the dynamics of global strings produced by bubble collisions, together with the emission of free axions and GWs from these strings. The log dependence in the scaling parameter is confirmed to ensure that the network enters the scaling stage. The axion decay constant faf_a and mass mam_a are subject to constraints from the dark matter relic density and dark radiation in Big Bang nucleosynthesis (BBN). We find that the ratio γ\gamma of GWs energy to free axion energy is two orders of magnitude larger than in a second-order PT, which allows us to assess the detectability of GWs from global strings in GW detectors.

The simulation setup

The action of the system under study is given by

S=−∫d4x(∂μφ∂μφ∗+V(∣φ∣)),(1)S=-\int d^4x\left(\partial_\mu\varphi\partial^\mu\varphi^*+V(|\varphi|)\right), \tag*{(1)}

with the FLRW metric convention of ds2=a2(τ)(−dτ2+dx2)ds^2=a^2(\tau)(-d\tau^2+dx^2), and the potential form is

V(∣φ∣)=12M2∣φ∣2+13δ∣φ∣3+14λ∣φ∣4(2)V(|\varphi|)=\frac{1}{2}M^2|\varphi|^2+\frac{1}{3}\delta|\varphi|^3+\frac{1}{4}\lambda|\varphi|^4 \tag*{(2)}

with δ<0\delta<0 to trigger a first-order PT. The radial mode and the phase mode is defined as φ=r+fa2eia/fa\varphi=\frac{r+f_a}{\sqrt{2}}e^{ia/f_a}, with fa=2ϕvf_a=\sqrt{2}\phi_v and ϕv\phi_v is the true vacuum value of the potential. The squared mass of the radial mode is mr2=M2/2+δϕv+3λϕv2/2m_r^2=M^2/2+\delta\phi_v+3\lambda\phi_v^2/2.

The simulation adopts a fixed-background expansion with the initial scale factor ai=1.0a_i=1.0. The field and space-time variables are rescaled as φ~=φ/f∗\tilde{\varphi}=\varphi/f_*, δx~=δx\delta\tilde{x}=\delta x, and δτ~=ω∗ δt/a\delta\tilde{\tau}=\omega_*\,\delta t/a, where f∗=ω∗=Hif_*=\omega_*=H_i. The dimensionless potential form is V~=V/f∗2ω∗2=12M~2∣φ~∣2+13δ~∣φ~∣3+\tilde{V}=V/f_*^2\omega_*^2=\frac{1}{2}\tilde{M}^2|\tilde{\varphi}|^2+\frac{1}{3}\tilde{\delta}|\tilde{\varphi}|^3+ 14λ~∣ϕ~∣4\frac{1}{4}\tilde{\lambda}|\tilde{\phi}|^4. By varying the action, the equation of motion in dimensionless form is obtained as

ϕ~′′−∇~2ϕ~+2a′aϕ~′=−a2∂V~∂∣ϕ~∣ϕ~∣ϕ~∣(3)\tilde{\phi}''-\widetilde{\nabla}^{2}\tilde{\phi}+2\frac{a'}{a}\tilde{\phi}'=-a^{2}\frac{\partial\widetilde{V}}{\partial|\tilde{\phi}|}\frac{\tilde{\phi}}{|\tilde{\phi}|} \tag*{(3)}

where the prime denotes the time derivative with respect to conformal time τ~\tilde{\tau}. With the choice of ω⋆\omega_{\star} and aia_i, the expansion in the radiation-dominated background in the simulation is given by

τ~=ττi=aai=(TTi)−1=(HHi)−12.(4)\tilde{\tau}=\frac{\tau}{\tau_i}=\frac{a}{a_i}=\left(\frac{T}{T_i}\right)^{-1}=\left(\frac{H}{H_i}\right)^{-\frac{1}{2}}. \tag*{(4)}

from the Friedmann equation. To simplify the simulation, the initial Hubble parameter is chosen to be Hi=7.11×1017 GeVH_i=7.11\times10^{17}\ \mathrm{GeV}, which equals the initial temperature Ti=HiT_i=H_i for a relativistic degree of freedom g⋆=106.75g_{\star}=106.75.

Using the public code CosmoLattice [51, 52], we evolve the equation of motion Eq. (3) in the radiation-dominated background with a second-order leap-frog algorithm. Initially, bubbles nucleate with randomly distributed centers. The magnitude of the complex field follows the radial profile

∣ϕ∣=ϕb2(1−tanh⁡R−R0lw)(5)|\phi|=\frac{\phi_b}{2}\left(1-\tanh\frac{R-R_0}{l_w}\right) \tag*{(5)}

for R<R0+lw/2R<R_0+l_w/2, where RR is the distance from the bubble center. Each bubble is assigned a single random phase uniformly distributed in [0,2π][0,2\pi]. The bubble profile is computed using FindBounce [53, 54]. Outside the bubble regions, the field values are set to zero. The simulation box consists of N=1024N=1024 grid points, and the rescaled time step is δτ~=0.01<δx~/3\delta\tilde{\tau}=0.01<\delta\tilde{x}/3 for each run. At the end of the simulation, the string core ds=m−1d_s=m^{-1} is resolved by 1.4 grid spacings. To investigate the effect of the number of bubbles, we consider two parameter setups listed in Table I, where the tilde denotes dimensionless variables rescaled by f⋆f_{\star} and ω⋆\omega_{\star}. To investigate the randomness of the bubble places in the simulation box, each parameter setup is realized five times.

M~\tilde{M}δ~\tilde{\delta}λ~\tilde{\lambda}ϕ~b\tilde{\phi}_bR~0\tilde{R}_0l~w\tilde{l}_wNbN_bL~\tilde{L}L/He−1L/H_e^{-1}
setup 10.64-3.322.370.912.481.9123848.074.18
setup 20.77-3.322.371.083.141.8411948.073.85

Table 1. The rescaled potential parameters, the bubble profile parameters, the initial bubble number, the simulation box length, and the number of Hubble lengths contained in the box at the end of the simulation.

Numerical Results

Snapshots of the bubble and string evolution for setup 1 parameters are shown in Figure 1, which illustrates the formation of cosmic strings from bubble collisions.

Snapshots of the phase field, norm of the complex field, and string network

Figure 1. Top: snapshots of the phase field at τ~=1.0\tilde{\tau}=1.0 (left), masked by ∣ϕ∣>0.5ϕv|\phi|>0.5\phi_v, and the norm of complex field at τ~=5.5\tilde{\tau}=5.5 (right), masked by ∣ϕ∣<0.1ϕv|\phi|<0.1\phi_v. Bottom: snapshots of the string network at τ~=8.5\tilde{\tau}=8.5 (left) and 11.5 (right), using the phase identification method.

The upper panels display the phase distribution at initial rescaled times τ~=1.0\tilde{\tau}=1.0 with a mask applied for ∣ϕ∣>0.5ϕv|\phi|>0.5\phi_v, and the residual true-vacuum regions. The lower panels show the identified string networks at τ~=8.5\tilde{\tau}=8.5 and 11.5, obtained via the phase identification method. One can observe that as the bubbles collide and merge, a string network appears in the remaining false-vacuum regions and persists.

The evolution of the scaling parameter is presented in Figure 2, following the method described in [55, 56]. The two first-order setups, characterized by different bubble numbers NbN_b, yield distinct initial string densities. It can be seen that, despite the effect of the initial density of global strings, the string network still enters the scaling regime. We fit the scaling parameter with ξ=αlog⁡mr/H+β\xi=\alpha\log m_r/H+\beta and obtain α=0.108±0.024\alpha=0.108\pm0.024 and β=−0.231±0.091\beta=-0.231\pm0.091 for setup1, and α=0.112±0.015\alpha=0.112\pm0.015 and β=−0.277±0.057\beta=-0.277\pm0.057 for setup2, where the errors are the standard deviations among the five realizations. The value of α\alpha is smaller than the result 0.24(2),0.254(2),0.21(2)0.24(2),0.254(2),0.21(2), and 0.205(7)/0.23(6)0.205(7)/0.23(6) in the second order PT scenario [32, 33, 35, 36].

Evolution of the scaling parameter in the simulation

Figure 2. Evolution of the scaling parameter in the simulation. Symbols with error bars are the mean and standard deviation over five realizations. The dashed lines correspond to the fitting results with shaded bands from the fit-parameter uncertainties.

We calculate the axions emitted from the global string network using the method in Refs. [55, 56]. The mean inverse comoving momentum k−1(t)k^{-1}(t) is obtained from the dimensionless parameter ϵ−1\epsilon^{-1}, and the scaling parameter is computed using the values of α\alpha and β\beta. After the string stage, the number density scales as n∝α−3n\propto\alpha^{-3} until today without considering other effects, and these axions can be regarded as part of dark matter. Moreover, if the string ends after BBN, the energy density of free axions is constrained by the upper bound of dark radiation

ΔNeff<0.46\Delta N_{\mathrm{eff}} < 0.46. see Appendix. B for details. The parameter space of axion-like particles (ALPs), constrained by dark matter relic density Ωh2<0.12\Omega h^2 < 0.12 and dark radiation, is shown in Figure 3. Compared with the constraints on axion-like particles produced in the second-order PT [56], the constraints on faf_a and mam_a from dark matter and BBN are slightly weaker.

Figure 3. Constraints on the ALPs parameter faf_a and mam_a from dark matter relic density and dark radiation. The axion case corresponds to the result in Ref. [56]. The shaded region is excluded.

We use the instantaneous emission spectrum Fg(k/H,mr/H)F_g(k/H,m_r/H) and the ratio γ\gamma to obtain the GWs spectrum during the string stage. The GW emission rate is related to the axion emission rate by Γgw≃γΓaGμ2/fa2\Gamma_{\mathrm{gw}} \simeq\gamma\Gamma_a G\mu^2/f_a^2 with Γa=a−4d(a4ρa)dt\Gamma_a = a^{-4} \frac{\mathrm{d}(a^4\rho_a)}{\mathrm{d}t} and Γgw=a−4d(a4ρgw)dt\Gamma_{\mathrm{gw}} = a^{-4} \frac{\mathrm{d}(a^4\rho_{\mathrm{gw}})}{\mathrm{d}t}. The forward difference is taken over a time interval τ~=1.0\tilde{\tau}=1.0 with initial times τ~=10.0,11.5\tilde{\tau}=10.0,11.5, and the result is shown in Figure 4. The value of γ\gamma is 15.27±0.84,21.48±2.4415.27 \pm0.84,21.48 \pm2.44 for setups 1 and 2, respectively. See Appendix. C for detailed calculation of GWs spectrum and the instantaneous emission spectrum FgF_g. After redshiftting, the present-day GWs spectrum is shown in Figure 5, with GW detectors such as LISA [57, 58], TianQin [59, 60],

The ratio $\gamma$ between the axion and GWs emission

Figure 4. The ratio γ\gamma between the axion and GWs emission.

The gravitational waves spectrum today

Figure 5. The gravitational waves spectrum today for the values of faf_a and mam_a chosen in Figure 3. The dashed and dotted lines correspond to the setup1 and setup2 inputs, respectively. The solid line corresponds to the result in Ref. [56].

Taiji [61, 62], μ\muAres [63], DECIGO [64, 65, 66, 67], BBO [68, 69, 70], SKA [71], ET [72, 73], CE [74] and LIGO [75, 76, 77]. We find that the GWs spectra are two orders of magnitude higher than those of the axion string with the second-order PTs. The reason is as follows: the emission of GWs is a small part of the string evolution, with γ∼O(0.1)\gamma\sim O(0.1) obtained in the second-order PTs [40, 56]; however, the γ\gamma factors for the two setups of the first-order PTs are two orders higher than that cases.

Conclusion and discussion

In this work, we have studied the formation and evolution of global cosmic strings produced by a cosmological first-order PT, together with the free axions and the GWs emitted by the resulting string network. As the bubbles merge, the false vacuum is compressed into sheets that shrink into line-like defects, leaving behind a string network. We find that first-order PTs yield a slightly smaller scaling parameter than the second-order scenarios. This leads to weaker constraints on the PQ scale (faf_a) and the axion-like particle mass (mam_a) from dark matter relic density and BBN. Furthermore, the GW-to-axion emission ratio is enhanced by about two orders of magnitude compared with the second-order PT. In contrast to the conventional expectation for global strings from second-order PTs [31, 32, 40], the enhanced GW channel offers an opportunity to probe the parameter space of the axion decay constant faf_a and mass mam_a for global strings using near-future GW detectors such as PTA, LISA, and μ\muAres.

Future studies may explore phase equilibration during the evolution and how it can be described. Since thermal-plasma friction is not included in our setup, the bubble walls expand freely; this allows the dependence of the string formation efficiency on the bubble wall velocity to be investigated. Local strings from a first-order PT are also worth studying to determine their formation probability and GW emission efficiency. Moreover, the axion density from strings may affect axion miniclusters and axion stars [78, 79]; we leave these aspects for future study.

Acknowledgements

This work is supported by the National Natural Science Foundation of China (NSFC) under Grants Nos. 12322505, 12547101, and 12675130. L.B. also acknowledges Chongqing Natural Science Foundation under Grant No. CSTB2024NSCQ-JQX0022 and Chongqing Talents: Exceptional Young Talents Project No. cstc2024ycjh-bgzxm0020. H.-K. G. is supported by the startup fund provided by the University of Chinese Academy of Sciences and by the NSFC under Grants No. 12547104 and No. 12475109.

Vacuum and phase evolutions

We show the evolutions of the true-vacuum fraction, ftrue=⟨∣ϕ∣⟩/φvf_{\mathrm{true}} = \langle|\phi|\rangle/\varphi_v, its time derivative, and the average phase with respect to the comoving rescaled time in Figure 6. Setup 1 has a larger initial bubble number and completes the first-order PT earlier. And, for both parameter setups, the average phases for different realizations have different departures from the zero, although the true-vacuum fractions and their time derivatives are all almost the same.

Evolution of the true-vacuum fractions, their time derivatives, and the average phases

Figure 6. Evolution of the true-vacuum fractions (top), their time derivatives (middle), and the average phases (bottom).

Free axion spectra

For the free axion emission spectrum, the energy density of free axions can be decomposed as ρa≈⟨a˙s2⟩=∫∂ρa∂∣k∣ d∣k∣\rho_a \approx\langle\dot{a}_s^2\rangle= \int\frac{\partial\rho_a}{\partial|\mathbf{k}|}\,d|\mathbf{k}|, where the masked axion field is defined as a˙s(x,t)=a˙(x,t)∣ϕ∣/∣φv∣\dot{a}_s(x,t)=\dot{a}(x,t)|\phi|/|\varphi_v|. This mask equals 0 near the string core and approaches 1 far from the string, thereby excluding the contribution from string cores. The spectrum is presented in Figure 7, where one can see that it changes little after the string enters the scaling regime.

Evolution of the axion spectra for setups 1 and 2

Figure 7. Evolution of the axion spectra for setups 1 and 2 (from top to bottom). The results are averaged over five realizations.

Under the assumption that the mean physical reciprocal momentum of free axion is proportional to the Hubble scale, the dimensionless parameters can be extracted from simulation by ϵ−1=a(t)k−1(t)/(t/2π)\epsilon^{-1}=a(t)k^{-1}(t)/(t/2\pi). The comoving mean reciprocal momentum of axion, k−1(t)k^{-1}(t), is computed as [55]

k−1(t)=∫dk2π12kΔP(k,t)∫dk2πΔP(k,t)(6)k^{-1}(t)=\frac{\int\frac{dk}{2\pi}\frac{1}{2}k\Delta P(k,t)}{\int\frac{dk}{2\pi}\Delta P(k,t)} \tag*{(6)}

where the differential spectrum is defined by ΔP(k,t)=a(t2)4P(k,t2)−a(t1)4P(k,t1)\Delta P(k,t)=a(t_2)^4P(k,t_2)-a(t_1)^4P(k,t_1), and P(k,t)P(k,t) is the free axion power spectrum at the comoving momentum space. The inverse slope parameter ϵ−1=0.494±0.023,0.430±0.073\epsilon^{-1}=0.494\pm0.023,0.430\pm0.073 is extracted from the data at begin time τ~=9.0,10.5\tilde{\tau}=9.0,10.5 and the end time τ~=11.5,12.5\tilde{\tau}=11.5,12.5.

In the evolution of the string network, the energy conversion of global strings can be described by

dρs(t)dt=−2H(t)ρs(t)−[dρs(t)dt]a−[dρs(t)dt]g(7)\frac{d\rho_s(t)}{dt}=-2H(t)\rho_s(t)-\left[\frac{d\rho_s(t)}{dt}\right]_a-\left[\frac{d\rho_s(t)}{dt}\right]_g \tag*{(7)}
dρa(t)dt=−4H(t)ρa(t)+[dρs(t)dt]a(8)\frac{d\rho_a(t)}{dt}=-4H(t)\rho_a(t)+\left[\frac{d\rho_s(t)}{dt}\right]_a \tag*{(8)}
dρgw(t)dt=−4H(t)ρgw(t)+[dρs(t)dt]g(9)\frac{d\rho_{gw}(t)}{dt}=-4H(t)\rho_{gw}(t)+\left[\frac{d\rho_s(t)}{dt}\right]_g \tag*{(9)}

where the energy of the string network is converted into free axions and gravitational waves. The string energy can be described by ρs=πfa2ln⁡(t/(ξδs))ξ/t2\rho_s=\pi f_a^2\ln(t/(\sqrt{\xi}\delta_s))\xi/t^2, and the GW emission term can be neglected when fa≪mplf_a\ll m_{\mathrm{pl}} in the axion energy calculation. The axion number density is calculated by

N(t)=∫t∗twdtk−1(t)dE(t)dt(10)N(t)=\int_{t_*}^{t_w}dt k^{-1}(t)\frac{dE(t)}{dt} \tag*{(10)}

where N(t)≡a3na(t)N(t)\equiv a^3n_a(t) and E(t)≡a4ρa(t)E(t)\equiv a^4\rho_a(t) are the comoving number and energy density of axion. The integration is performed from the time when T=faT=f_a to the time when ma=3Hm_a=3H. The effective number of neutrinos in

BBN is calculated as ΔNeff=(8/7)(11/4)4/3ρa/ργ\Delta N_{\mathrm{eff}}=(8/7)(11/4)^{4/3}\rho_a/\rho_\gamma at temperature of 1 MeV for free axion energy density.

For the QCD case, the axion mass squared is given by ma2=α(T/Λ)−nΛ4/fa2m_a^2=\alpha(T/\Lambda)^{-n}\Lambda^4/f_a^2, where we use the parameters α=1.68×10−7\alpha=1.68\times10^{-7}, Λ=400 MeV\Lambda=400\ \mathrm{MeV}, n=6.68n=6.68 from [80]. The present QCD axion energy density from strings is calculated by Ωa,string=na,0ma,0/ρcrit,0\Omega_{a,\mathrm{string}}=n_{a,0}m_{a,0}/\rho_{\mathrm{crit},0}, where the zero-temperature axion mass is ma2(T=0)=1.46×10−3Λ4/fa2m_a^2(T=0)=1.46\times10^{-3}\Lambda^4/f_a^2, and ρcrit,0\rho_{\mathrm{crit},0} is the critical density today. The total axion energy density also includes axions from the misalignment mechanism [81]. The dark matter relic density constrains the QCD axion decay constant to fa<1.2×1010,1.3×1010 GeVf_a<1.2\times10^{10},1.3\times10^{10}\ \mathrm{GeV} for setups 1-2, respectively.

We also calculate the generation ratio p=ℓs/R∗2p=\ell_s/R_*^2, where ℓs=ξ/t2\ell_s=\xi/t^2 and R∗=(Nb/V)−1/3R_*=(N_b/V)^{-1/3}, shown in Figure 8, using data taken at the onset of the scaling regime and at the time of the minimum scaling parameter. This ratio is much less than the theoretically predicted value of 0.25 for the collisions of three bubbles [46].

The ratio of strings generated from bubble collisions

Figure 8. The ratio of strings generated from bubble collisions. The star denotes data at the beginning of the scaling parameter, and the solid circle denotes data at the time corresponding to the minimum scaling parameter.

Gravitational wave spectra evaluation

GWs correspond to the transverse-traceless (TT) component of the metric perturbation hijh_{ij}, defined by hij=gij−δijh_{ij}=g_{ij}-\delta_{ij}. The equation of motion for GWs is

h¨ij+3Hh˙ij−1a2∇2hij=2mpl2a2ΠijTT(11)\ddot{h}_{ij}+3H\dot{h}_{ij}-\frac{1}{a^2}\nabla^2h_{ij}=\frac{2}{m_{\mathrm{pl}}^2a^2}\Pi^{TT}_{ij} \tag*{(11)}

where an overdot denotes differentiation with respect to cosmic time, mplm_{\mathrm{pl}} is the reduced Planck mass, and ΠijTT\Pi^{TT}_{ij} is the transverse-traceless part of the anisotropic tensor Πij\Pi_{ij}. The anisotropic tensor is given by Πij=Tij−pgij\Pi_{ij}=T_{ij}-pg_{ij}, where the energy-momentum tensor is evaluated as Tij=2Re⁡[∂iφ∂jφ∗]T_{ij}=2\operatorname{Re}[\partial_i\varphi\partial_j\varphi^*] [51], while the background pressure term does not contribute to ΠijTT\Pi^{TT}_{ij}. In Fourier space, the transverse-traceless anisotropic tensor is obtained via ΠijTT=Λij,klΠkl\Pi^{TT}_{ij}=\Lambda_{ij,kl}\Pi_{kl}, with the projection operator defined as

Λij,kl(k^)=Pik(k^)Pjl(k^)−12Pij(k^)Pkl(k^),\Lambda_{ij,kl}(\hat{\mathbf{k}})=P_{ik}(\hat{\mathbf{k}})P_{jl}(\hat{\mathbf{k}})-\frac{1}{2}P_{ij}(\hat{\mathbf{k}})P_{kl}(\hat{\mathbf{k}}),
Pij(k^)=δij−k^ik^j,(12)P_{ij}(\hat{\mathbf{k}})=\delta_{ij}-\hat{k}_{i}\hat{k}_{j}, \tag*{(12)}

where k^i=ki/k\hat{k}_{i}=k_{i}/k is the unit vector along k\mathbf{k}. For convenience in numerically computing GWs in lattice simulation, the auxiliary tensor uiju_{ij} is introduced and evolved, with the metric perturbation recovered as hij=Λij,kluijh_{ij}=\Lambda_{ij,kl}u_{ij}. The equation of motion of uiju_{ij} is

u¨ij+3Hu˙ij−1a2∇2uij=2mpl2a2Πij.(13)\ddot{u}_{ij}+3H\dot{u}_{ij}-\frac{1}{a^{2}}\nabla^{2}u_{ij}=\frac{2}{m_{\mathrm{pl}}^{2}a^{2}}\Pi_{ij}. \tag*{(13)}

The gravitational wave energy density is defined as

ρgw(t)=132πG⟨h˙ij(x,t)h˙ij(x,t)⟩,(14)\rho_{gw}(t)=\frac{1}{32\pi G}\left\langle\dot{h}_{ij}(\mathbf{x},t)\dot{h}_{ij}(\mathbf{x},t)\right\rangle, \tag*{(14)}

where ⟨⋯ ⟩\langle\cdots\rangle denotes spatial averaging. The gravitational wave spectrum generated by the bubble collisions and subsequent string dynamics is shown in Figure 9. One can see that bubble expansion and collisions cause the GW spectrum to increase significantly before the strings enter the scaling regime; after entering the scaling regime, the GW spectrum increases slowly due to string evolution.

Gravitational-wave spectra from the simulations for setups 1 and 2

Figure 9. Gravitational-wave spectra from the simulations for setups 1 and 2 (from top to bottom), averaged over five realizations. The same color solid line corresponds to the same rescaled conformal time τ\tau. The dashed line and dotted line correspond to 2πH2\pi H and the mass of the radial mode mrm_r. The black solid line correspond to 2πR∗−12\pi R_*^{-1}

The methods in [40, 56] are used to calculate the GWs during the string evolution epoch. The instantaneous emission spectrum Fg(x,y)F_g(x,y) is calculated by:

Fg(x,y)=HΓgw1a3∂∂t(a3∂ρgw∂k),(15)F_g(x,y)=\frac{H}{\Gamma_{gw}}\frac{1}{a^3}\frac{\partial}{\partial t}\left(a^3\frac{\partial\rho_{gw}}{\partial k}\right), \tag*{(15)}

where the variables are defined as x=k/Hx=k/H and y=mr/Hy=m_r/H. The spectral index is obtained by fitting Fg(x,y)∝x−qF_g(x,y)\propto x^{-q}, yielding q=1.457±0.037q=1.457\pm0.037 and 1.606±0.0661.606\pm0.066 for setups 1 and 2, respectively, in the range [8πH,mr/2][8\pi H,m_r/2], using the same differencing method and data as in the γ\gamma calculation. The results are presented in Figure 10.

Evolution of the instantaneous emission spectrum

Figure 10. Evolution of the instantaneous emission spectrum F(x,y)F(x,y) for setups 1 and 2 (from top to bottom), averaged over five realizations. The dotted black line corresponds to 8πH8\pi H, and the dashed lines correspond to mr/2m_r/2. The black dashed line corresponds to the fitting result.

Together with the axion energy, the GWs spectrum at the end time of the global string is calculated from the emission spectrum Fg(x,y)F_g(x,y) and the fitting parameters γ\gamma and qq. The GWs spectrum is then redshifted as ΩGW,0=Ωrad,0(g0/ge)1/3ΩGW,e\Omega_{GW,0}=\Omega_{\mathrm{rad},0}(g_0/g_e)^{1/3}\Omega_{GW,e}, and the physical frequency scales as k∝a−1k\propto a^{-1} with the redshift factor a04/ae4=ge1/3ρrad,e/(g01/3ρrad,0)a_0^4/a_e^4=g_e^{1/3}\rho_{\mathrm{rad},e}/(g_0^{1/3}\rho_{\mathrm{rad},0}) [56].

References

References

  1. [1]G. Agazie et al. (NANOGrav), Astrophys. J. Lett. 951, L8 (2023), arXiv:2306.16213 [astro-ph.HE].DOI
  2. [2]J. Antoniadis et al. (EPTA, InPTA:), Astron. Astrophys. 678, A50 (2023), arXiv:2306.16214 [astro-ph.HE].DOI
  3. [3]D. J. Reardon et al., Astrophys. J. Lett. 951, L6 (2023), arXiv:2306.16215 [astro-ph.HE].DOI
  4. [4]H. Xu et al., Res. Astron. Astrophys. 23, 075024 (2023), arXiv:2306.16216 [astro-ph.HE].DOI
  5. [5]T. W. B. Kibble, J. Phys. A 9, 1387 (1976).
  6. [6]A. Vilenkin and A. E. Everett, Phys. Rev. Lett. 48, 1867 (1982).
  7. [7]N. Kitajima and K. Nakayama, Phys. Lett. B 846, 138213 (2023), arXiv:2306.17390 [hep-ph].arxiv.org/abs/2306.17390
  8. [8]L. Bian, S. Ge, J. Shu, B. Wang, X.-Y. Yang, and J. Zong, Phys. Rev. D 109, L101301 (2024), arXiv:2307.02376 [astro-ph.HE].arxiv.org/abs/2307.02376
  9. [9]S. Basilakos, D. V. Nanopoulos, T. Papanikolaou, E. N. Saridakis, and C. Tzerefos, Phys. Lett. B 850, 138507 (2024), arXiv:2307.08601 [hep-th].arxiv.org/abs/2307.08601
  10. [10]Z. Wang, L. Lei, H. Jiao, L. Feng, and Y.-Z. Fan, Sci. China Phys. Mech. Astron. 66, 120403 (2023), arXiv:2306.17150 [astro-ph.HE].DOI
  11. [11]J. Ellis, M. Lewicki, C. Lin, and V. Vaskonen, Phys. Rev. D 108, 103511 (2023), arXiv:2306.17147 [astro-ph.CO].arxiv.org/abs/2306.17147
  12. [12]W. Buchmuller, V. Domcke, and K. Schmitz, JCAP 11, 020 (2023), arXiv:2307.04691 [hep-ph].arxiv.org/abs/2307.04691
  13. [13]D. G. Figueroa, M. Pieroni, A. Ricciardone, and P. Simakachorn, Phys. Rev. Lett. 132, 171002 (2024), arXiv:2307.02399 [astro-ph.CO].DOI
  14. [14]A. Afzal et al. (NANOGrav), Astrophys. J. Lett. 951, L11 (2023), [Erratum: Astrophys.J.Lett. 971, L27 (2024), Erratum: Astrophys.J. 971, L27 (2024)], arXiv:2306.16219 [astro-ph.HE].DOI
  15. [15]J. Antoniadis et al. (EPTA, InPTA), Astron. Astrophys. 685, A94 (2024), arXiv:2306.16227 [astro-ph.CO].arxiv.org/abs/2306.16227
  16. [16]H. Cheng and L. V. Visinelli, Phys. Dark Univ. 46, 101667 (2024), arXiv:2408.16334 [hep-ph].DOI
  17. [17]L. Bian, R.-G. Cai, J. Liu, X.-Y. Yang, and R. Zhou, Phys. Rev. D 103, L081301 (2021), arXiv:2009.13893 [astro-ph.CO].DOI
  18. [18]R. D. Peccei and H. R. Quinn, Phys. Rev. Lett. 38, 1440 (1977).
  19. [19]R. D. Peccei and H. R. Quinn, Phys. Rev. D 16, 1791 (1977).
  20. [20]S. Weinberg, Phys. Rev. Lett. 40, 223 (1978).
  21. [21]F. Wilczek, Phys. Rev. Lett. 40, 279 (1978).
  22. [22]J. E. Kim, Phys. Rev. Lett. 43, 103 (1979).
  23. [23]M. A. Shifman, A. I. Vainshtein, and V. I. Zakharov, Nucl. Phys. B 166, 493 (1980).
  24. [24]A. R. Zhitnitsky, Sov. J. Nucl. Phys. 31, 260 (1980).
  25. [25]M. Dine, W. Fischler, and M. Srednicki, Phys. Lett. B 104, 199 (1981).
  26. [26]J. Preskill, M. B. Wise, and F. Wilczek, Phys. Lett. B 120, 127 (1983).
  27. [27]L. F. Abbott and P. Sikivie, Phys. Lett. B 120, 133 (1983).
  28. [28]M. Dine and W. Fischler, Phys. Lett. B 120, 137 (1983).
  29. [29]P. Auclair, D. A. Steer, and T. Vachaspati, Phys. Rev. D 101, 083511 (2020), arXiv:1911.12066 [hep-ph].DOI
  30. [30]D. Matsunami, L. Pogosian, A. Saurabh, and T. Vachaspati, Phys. Rev. Lett. 122, 201301 (2019), arXiv:1903.05102 [hep-ph].DOI
  31. [31]M. Gorghetto, E. Hardy, and G. Villadoro, Journal of High Energy Physics 2018 (2018), 10.1007/jhep07(2018)151.
  32. [32]M. Gorghetto, E. Hardy, and G. Villadoro, SciPost Phys. 10, 050 (2021), arXiv:2007.04990 [hep-ph].DOI
  33. [33]M. Buschmann, J. W. Foster, A. Hook, A. Peterson, D. E. Willcox, W. Zhang, and B. R. Safdi, Nature Communications 13 (2022), 10.1038/s41467-022-28669-y.
  34. [34]J. N. Benabou, M. Buschmann, S. Kumar, Y. Park, and B. R. Safdi, Phys. Rev. D 109, 055005 (2024), arXiv:2308.01334 [hep-ph].arxiv.org/abs/2308.01334
  35. [35]J. N. Benabou, M. Buschmann, J. W. Foster, and B. R. Safdi, Phys. Rev. Lett. 134, 241003 (2025), arXiv:2412.08699 [hep-ph].arxiv.org/abs/2412.08699
  36. [36]K. Saikawa, J. Redondo, A. Vaquero, and M. Kaltschmidt, (2024), arXiv:2401.17253 [hep-ph].DOI
  37. [37]H. Kim, J. Park, and M. Son, JHEP 07, 150 (2024), arXiv:2402.00741 [hep-ph].arxiv.org/abs/2402.00741
  38. [38]J. Baeza-Ballesteros, E. J. Copeland, D. G. Figueroa, and J. Lizarraga, Phys. Rev. D 110, 043522 (2024), arXiv:2308.08456 [astro-ph.CO].
  39. [39]T. Vachaspati and A. Vilenkin, Phys. Rev. D 31, 3052 (1985).
  40. [40]M. Gorghetto, E. Hardy, and H. Nicolaescu, JCAP 06, 034 (2021), arXiv:2101.11007 [hep-ph].DOI
  41. [41]Z. Zhao, L. Bian, and J. Shu, Phys. Rev. D 114, 023539 (2026), arXiv:2507.00685 [astro-ph.CO].DOI
  42. [42]M. B. Hindmarsh and T. W. B. Kibble, Rept. Prog. Phys. 58, 477 (1995), arXiv:hep-ph/9411342.DOI
  43. [43]T. W. B. Kibble, Phys. Rept. 67, 183 (1980).
  44. [44]M. Hindmarsh, A.-C. Davis, and R. H. Brandenberger, Phys. Rev. D 49, 1944 (1994), arXiv:hep-ph/9307203.DOI
  45. [45]E. J. Copeland and P. M. Saffin, Phys. Rev. D 54, 6088 (1996), arXiv:hep-ph/9604231.DOI
  46. [46]T. W. B. Kibble and A. Vilenkin, Phys. Rev. D 52, 679 (1995), arXiv:hep-ph/9501266.DOI
  47. [47]A. M. Srivastava, Phys. Rev. D 46, 1353 (1992).
  48. [48]A. Melfo and L. Perivolaropoulos, Phys. Rev. D 52, 992 (1995), arXiv:hep-ph/9501284.DOI
  49. [49]A. Ferrera, Phys. Rev. D 57, 7130 (1998), arXiv:hep-ph/9612487.DOI
  50. [50]M. Lilley and A. Ferrera, Phys. Rev. D 64, 023520 (2001), arXiv:hep-ph/0102035.DOI
  51. [51]D. G. Figueroa, A. Florio, F. Torrenti, and W. Valkenburg, Journal of Cosmology and Astroparticle Physics 2021, 035 (2021).
  52. [52]D. G. Figueroa, A. Florio, F. Torrenti, and W. Valkenburg, Computer Physics Communications 283, 108586 (2023).
  53. [53]V. Guada, A. Maiezza, and M. Nemevšek, Phys. Rev. D 99, 056020 (2019), arXiv:1803.02227 [hep-th].DOI
  54. [54]V. Guada, M. Nemevšek, and M. Pintar, Comput. Phys. Commun. 256, 107480 (2020), arXiv:2002.00881 [hep-ph].DOI
  55. [55]T. Hiramatsu, M. Kawasaki, T. Sekiguchi, M. Yamaguchi, and J. Yokoyama, Physical Review D 83 (2011), 10.1103/physrevd.83.123531.
  56. [56]Y. Jia and L. Bian, Phys. Rev. D 111, 063552 (2025), arXiv:2412.04218 [hep-ph].DOI
  57. [57]P. Amaro-Seoane et al. (LISA), (2017), arXiv:1702.00786 [astro-ph.IM].
  58. [58]J. Baker et al., (2019), arXiv:1907.06482 [astro-ph.IM].arxiv.org/abs/1907.06482
  59. [59]J. Luo et al. (TianQin), Class. Quant. Grav. 33, 035010 (2016), arXiv:1512.02076 [astro-ph.IM].DOI
  60. [60]K. Zhou, J. Cheng, and L. Ren, (2023), arXiv:2306.14439 [gr-qc].arxiv.org/abs/2306.14439
  61. [61]W.-R. Hu and Y.-L. Wu, Natl. Sci. Rev. 4, 685 (2017).
  62. [62]W.-H. Ruan, Z.-K. Guo, R.-G. Cai, and Y.-Z. Zhang, Int. J. Mod. Phys. A 35, 2050075 (2020), arXiv:1807.09495 [gr-qc].
  63. [63]A. Sesana et al., Exper. Astron. 51, 1333 (2021), arXiv:1908.11391 [astro-ph.IM].DOI
  64. [64]N. Seto, S. Kawamura, and T. Nakamura, Phys. Rev. Lett. 87, 221103 (2001), arXiv:astro-ph/0108011.DOI
  65. [65]S. Kawamura et al., Class. Quant. Grav. 28, 094011 (2011).
  66. [66]K. Yagi and N. Seto, Phys. Rev. D 83, 044011 (2011), [Erratum: Phys.Rev.D 95, 109901 (2017)], arXiv:1101.3940 [astro-ph.CO].
  67. [67]S. Isoyama, H. Nakano, and T. Nakamura, PTEP 2018, 073E01 (2018), arXiv:1802.06977 [gr-qc].DOI
  68. [68]J. Crowder and N. J. Cornish, Phys. Rev. D 72, 083005 (2005), arXiv:gr-qc/0506015.DOI
  69. [69]V. Corbin and N. J. Cornish, Class. Quant. Grav. 23, 2435 (2006), arXiv:gr-qc/0512039.DOI
  70. [70]G. M. Harry, P. Fritschel, D. A. Shaddock, W. Folkner, and E. S. Phinney, Class. Quant. Grav. 23, 4887 (2006), [Erratum: Class.Quant.Grav. 23, 7361 (2006)].
  71. [71]G. Janssen et al., PoS AASKA14, 037 (2015), arXiv:1501.00127 [astro-ph.IM].arxiv.org/abs/1501.00127
  72. [72]M. Punturo et al., Class. Quant. Grav. 27, 194002 (2010).
  73. [73]S. Hild et al., Class. Quant. Grav. 28, 094013 (2011), arXiv:1012.0908 [gr-qc].DOI
  74. [74]B. P. Abbott et al. (LIGO Scientific), Class. Quant. Grav. 34, 044001 (2017), arXiv:1607.08697 [astro-ph.IM].DOI
  75. [75]J. Aasi et al. (LIGO Scientific, VIRGO), Class. Quant. Grav. 32, 115012 (2015), arXiv:1410.7764 [gr-qc].DOI
  76. [76]E. Thrane and J. D. Romano, Phys. Rev. D 88, 124032 (2013), arXiv:1310.5300 [astro-ph.IM].DOI
  77. [77]B. P. Abbott et al. (LIGO Scientific, Virgo), Phys. Rev. D 100, 061101 (2019), arXiv:1903.02886 [gr-qc].DOI
  78. [78]D. Ellis, D. J. E. Marsh, B. Eggemeier, J. Niemeyer, J. Redondo, and K. Dolag, Phys. Rev. D 106, 103514 (2022), arXiv:2204.13187 [hep-ph].DOI
  79. [79]B. Eggemeier, A. K. Anilkumar, and K. Dolag, Phys. Rev. D 110, 043530 (2024), arXiv:2402.18221 [astro-ph.CO].DOI
  80. [80]O. Wantz and E. Shellard, Nuclear Physics B 829, 110–160 (2010).
  81. [81]M. Kawasaki, K. Saikawa, and T. Sekiguchi, Phys. Rev. D 91, 065014 (2015).

Paper details

vlnc:W_sdxmk9k8
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 8

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents