Hyesung Kang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Hyesung Kang has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Dongsu Ryu
Ulsan National Institute of Science and Technology (UNIST)
3 shared publications - T. W. Jones
School of Physics and Astronomy University of Minnesota 116 Church St. S.E. 55455 Minneapolis MN
3 shared publications - Hyung Mok Lee
Dept. of Earth Sciences Pusan National University 609-735 Pusan Korea
2 shared publications - Joerg P. Rachen
Max-Planck-Institute for Radioastronomy D-53010 Bonn Germany
1 shared publication - Peter L. Biermann
MPI for Radioastronomy
1 shared publication - Sungsoo S. Kim1 shared publication
Latest Publications
- Open Access
Contributions to the Cosmic Ray Flux above the Ankle: Clusters of Galaxies
Motivated by the suggestion of Kang, Ryu \& Jones (1996) that particles can be accelerated to high energies via diffusive shock acceleration process at the accretion shocks formed by the infalling flow toward the clusters of galaxies, we have calculated the expected particle flux from a cosmological ensemble of clusters. We use the observed temperature distribution of local clusters and assume a simple power-law evolutionary model for the comoving density of the clusters. The shock parameters such as the shock radius and velocity are deduced from the ICM temperature using the self-similar solutions for secondary infall onto the clusters. The magnetic field strength is assume to be in equipartition with the postshock thermal energy behind the accretion shock. We also assume that the injected energy spectrum is a power-law with the exponential cutoff at the maximum energy which is calculated from the condition that the energy gain rate for diffusive shock acceleration is balanced by the loss rate due to the interactions with the cosmic background radiation. In contrast to the earlier paper we have adopted here the description of the cosmic ray diffusion by Jokipii (1987) which leads to considerably higher particle energies. Finally the injected particle spectrum at the clusters is integrated over the cosmological distance to earth by considering the energy loss due to the interactions with the cosmic background radiation. Our calculations show that the expected spectrum of high-energy protons from the cosmological ensemble of the cluster accretion shocks could match well the observed cosmic ray spectrum near $10^{19}$eV with reasonable parameters and models if about $10^{-4}$ of the infalling kinetic energy can be injected into the intergalactic space as the high energy particles.Open AccessContributions to the Cosmic Ray Flux above the Ankle: Clusters of Galaxies
Motivated by the suggestion of Kang, Ryu \& Jones (1996) that particles can be accelerated to high energies via diffusive shock acceleration process at the accretion shocks formed by the infalling flow toward the clusters of galaxies, we have calculated the expected particle flux from a cosmological ensemble of clusters. We use the observed temperature distribution of local clusters and assume a simple power-law evolutionary model for the comoving density of the clusters. The shock parameters such as the shock radius and velocity are deduced from the ICM temperature using the self-similar solutions for secondary infall onto the clusters. The magnetic field strength is assume to be in equipartition with the postshock thermal energy behind the accretion shock. We also assume that the injected energy spectrum is a power-law with the exponential cutoff at the maximum energy which is calculated from the condition that the energy gain rate for diffusive shock acceleration is balanced by the loss rate due to the interactions with the cosmic background radiation. In contrast to the earlier paper we have adopted here the description of the cosmic ray diffusion by Jokipii (1987) which leads to considerably higher particle energies. Finally the injected particle spectrum at the clusters is integrated over the cosmological distance to earth by considering the energy loss due to the interactions with the cosmic background radiation. Our calculations show that the expected spectrum of high-energy protons from the cosmological ensemble of the cluster accretion shocks could match well the observed cosmic ray spectrum near $10^{19}$eV with reasonable parameters and models if about $10^{-4}$ of the infalling kinetic energy can be injected into the intergalactic space as the high energy particles. - Open Access
Accreting Matter around Clusters of Galaxies: One-Dimensional Considerations
During the formation of the large scale structure of the Universe, matter accretes onto high density peaks. Accreting collisionless dark matter (DM) forms caustics around them, while accreting collisional baryonic matter (BM) forms accretion shocks. The properties of the accreting matter depend upon the power spectrum of the initial perturbations on a given scale as well as the background expansion in a given cosmological model. In this paper, we have calculated the accretion of DM particles in one-dimensional spherical geometry under various cosmological models including the Einstein-de Sitter universe, the open universe with $\Omega_o<1$, and the flat universe with $\Omega_{\Lambda}=1-\Omega_o$. A density parameter in the range $0.1\le \Omega_o \le 1$ has been considered. The initial perturbation characterized by a point mass at the origin has been considered. Since the accretion shock of BM is expected to form close to the first caustic of DM, the properties of the accreting BM are common with those of the DM. Hence, the accretion calculations with DM particles have been used to find the position and velocity of the accretion shock and the cluster mass inside it. The average temperature of BM has been estimated by adopting simplifying assumptions. The velocity of the accreting BM around clusters of a given temperature is smaller in a universe with smaller $\Omega_o$, but only by up to $\sim24\%$ in the models with $0.1\le \Omega_o \le 1$. Thus, it would be difficult to use that quantity to discriminate among the cosmological models. However, the accretion velocity around clusters of a given mass or a given radius depends more sensitively on the cosmological models. It is smaller in a universe with smaller $\Omega_o$ by up to $\sim41\%$ and $\sim65\%$, respectively. So, it can provide a betterOpen AccessAccreting Matter around Clusters of Galaxies: One-Dimensional Considerations
During the formation of the large scale structure of the Universe, matter accretes onto high density peaks. Accreting collisionless dark matter (DM) forms caustics around them, while accreting collisional baryonic matter (BM) forms accretion shocks. The properties of the accreting matter depend upon the power spectrum of the initial perturbations on a given scale as well as the background expansion in a given cosmological model. In this paper, we have calculated the accretion of DM particles in one-dimensional spherical geometry under various cosmological models including the Einstein-de Sitter universe, the open universe with $\Omega_o<1$, and the flat universe with $\Omega_{\Lambda}=1-\Omega_o$. A density parameter in the range $0.1\le \Omega_o \le 1$ has been considered. The initial perturbation characterized by a point mass at the origin has been considered. Since the accretion shock of BM is expected to form close to the first caustic of DM, the properties of the accreting BM are common with those of the DM. Hence, the accretion calculations with DM particles have been used to find the position and velocity of the accretion shock and the cluster mass inside it. The average temperature of BM has been estimated by adopting simplifying assumptions. The velocity of the accreting BM around clusters of a given temperature is smaller in a universe with smaller $\Omega_o$, but only by up to $\sim24\%$ in the models with $0.1\le \Omega_o \le 1$. Thus, it would be difficult to use that quantity to discriminate among the cosmological models. However, the accretion velocity around clusters of a given mass or a given radius depends more sensitively on the cosmological models. It is smaller in a universe with smaller $\Omega_o$ by up to $\sim41\%$ and $\sim65\%$, respectively. So, it can provide a better - Open Access
Diffusive Shock Acceleration in Oblique MHD Shocks: Comparison with Monte Carlo Methods and Observations
We report simulations of diffusive particle acceleration in oblique magnetohydrodynamical (MHD) shocks. These calculations are based on extension to oblique shocks of a numerical model for ``thermal leakage'' injection of particles at low energy into the cosmic-ray population. That technique, incorporated into a fully dynamical diffusion-convection formalism, was recently introduced for parallel shocks by Kang \& Jones (1995). Here, we have compared results of time dependent numerical simulations using our technique with Monte Carlo simulations by Ellison, Baring \& Jones 1995 and with {\it in situ} observations from the Ulysses spacecraft of oblique interplanetary shocks discussed by Baring \etal (1995). Through the success of these comparisons we have demonstrated that our {diffusion-convection} method and injection techniques provide a practical tool to capture essential physics of the injection process and particle acceleration at oblique MHD shocks. In addition to the diffusion-convection simulations, we have included time dependent two-fluid simulations for a couple of the shocks to demonstrate the basic validity of that formalism in the oblique shock context. Using simple models for the two-fluid closure parameters based on test-particle considerations, we find good agreement with the dynamical properties of the more detailed diffusion-convection results. We emphasize, however, that such two-fluid results can be sensitive to the properties of these closure parameters when the flows are not truly steady. Furthermore, we emphasize through example how the validity of the two-fluid formalism does not necessarily mean that {\it steady-state} two-fluid models provide a reliable tool for predicting the efficiency of particle acceleration in real shocks.
Open AccessDiffusive Shock Acceleration in Oblique MHD Shocks: Comparison with Monte Carlo Methods and Observations
We report simulations of diffusive particle acceleration in oblique magnetohydrodynamical (MHD) shocks. These calculations are based on extension to oblique shocks of a numerical model for ``thermal leakage'' injection of particles at low energy into the cosmic-ray population. That technique, incorporated into a fully dynamical diffusion-convection formalism, was recently introduced for parallel shocks by Kang \& Jones (1995). Here, we have compared results of time dependent numerical simulations using our technique with Monte Carlo simulations by Ellison, Baring \& Jones 1995 and with {\it in situ} observations from the Ulysses spacecraft of oblique interplanetary shocks discussed by Baring \etal (1995). Through the success of these comparisons we have demonstrated that our {diffusion-convection} method and injection techniques provide a practical tool to capture essential physics of the injection process and particle acceleration at oblique MHD shocks. In addition to the diffusion-convection simulations, we have included time dependent two-fluid simulations for a couple of the shocks to demonstrate the basic validity of that formalism in the oblique shock context. Using simple models for the two-fluid closure parameters based on test-particle considerations, we find good agreement with the dynamical properties of the more detailed diffusion-convection results. We emphasize, however, that such two-fluid results can be sensitive to the properties of these closure parameters when the flows are not truly steady. Furthermore, we emphasize through example how the validity of the two-fluid formalism does not necessarily mean that {\it steady-state} two-fluid models provide a reliable tool for predicting the efficiency of particle acceleration in real shocks.
- Open Access
Close Encounters between a Neutron Star and a Main-Sequence Star
We have examined consequences of strong tidal encounters between a neutron star and a normal star using SPH as a possible formation mechanism of isolated recycled pulsars in globular clusters. We have made a number of SPH simulations for close encounters between a main-sequence star of mass ranging from 0.2 to 0.7 Solar masses represented by an n=3/2 polytrope and a neutron star represented by a point mass. The outcomes of the first encounters are found to be dependent only on the dimensionless parameter eta' = (m/(m+M))^(1/2) (r_min/R_MS)^(3/2) (m/M)^(1/6), where m and M are the masses of the main-sequence star and the neutron star, respectively, r_min the minimum separation between two stars, and R_MS the size of the main-sequence star. The material from the (at least partially) disrupted star forms a disk around the neutron star. If all material in the disk is to be acctreted onto the neutron star's surface, the mass of the disk is enough to spin up the neutron star to spin period of 1 ms.
Open AccessClose Encounters between a Neutron Star and a Main-Sequence Star
We have examined consequences of strong tidal encounters between a neutron star and a normal star using SPH as a possible formation mechanism of isolated recycled pulsars in globular clusters. We have made a number of SPH simulations for close encounters between a main-sequence star of mass ranging from 0.2 to 0.7 Solar masses represented by an n=3/2 polytrope and a neutron star represented by a point mass. The outcomes of the first encounters are found to be dependent only on the dimensionless parameter eta' = (m/(m+M))^(1/2) (r_min/R_MS)^(3/2) (m/M)^(1/6), where m and M are the masses of the main-sequence star and the neutron star, respectively, r_min the minimum separation between two stars, and R_MS the size of the main-sequence star. The material from the (at least partially) disrupted star forms a disk around the neutron star. If all material in the disk is to be acctreted onto the neutron star's surface, the mass of the disk is enough to spin up the neutron star to spin period of 1 ms.
- Open Access
Supersonic Collisions between Two Gas Streams
A star around a massive black hole can be disrupted tidally by the gravity of the black hole. Then, its debris may form a precessing stream which may even collide with itself. In order to understand the dynamical effects of the stream-stream collision on the eventual accretion of the stellar debris onto the black hole, we have studied how gas flow behaves when the outgoing stream collides supersonically with the incoming stream. We have investigated the problem analytically with one-dimensional plane-parallel streams and numerically with more realistic three-dimensional streams. A shock formed around the contact surface converts the bulk of the orbital streaming kinetic energy into thermal energy. In three-dimensional simulations, the accumulated hot post-shock gas then expands adiabatically and drives another shock into the low density ambient region. Through this expansion, thermal energy is converted back to the kinetic energy associated with the expanding motion. Thus, in the end, only a small fraction of the orbital kinetic energy is actually converted to the thermal energy, while most of it is transferred to the kinetic energy of the expanding gas. Nevertheless the collision is effective in circularizing the debris orbit, because the shock efficiently transforms the ordered motion of the streams into the expanding motion in directions perpendicular to the streams. The circularization efficiency decreases, if two colliding streams have a large ratio of cross sections and a large density contrast. But even in such cases, the main shock extends beyond the overlapping contact surface and the high pressure region behind the shock keeps the stream of the larger cross section from passing freely. Thus the stream-stream collisions are still expected to circularize the stellar debris rather efficiently, unless the ratio
Open AccessSupersonic Collisions between Two Gas Streams
A star around a massive black hole can be disrupted tidally by the gravity of the black hole. Then, its debris may form a precessing stream which may even collide with itself. In order to understand the dynamical effects of the stream-stream collision on the eventual accretion of the stellar debris onto the black hole, we have studied how gas flow behaves when the outgoing stream collides supersonically with the incoming stream. We have investigated the problem analytically with one-dimensional plane-parallel streams and numerically with more realistic three-dimensional streams. A shock formed around the contact surface converts the bulk of the orbital streaming kinetic energy into thermal energy. In three-dimensional simulations, the accumulated hot post-shock gas then expands adiabatically and drives another shock into the low density ambient region. Through this expansion, thermal energy is converted back to the kinetic energy associated with the expanding motion. Thus, in the end, only a small fraction of the orbital kinetic energy is actually converted to the thermal energy, while most of it is transferred to the kinetic energy of the expanding gas. Nevertheless the collision is effective in circularizing the debris orbit, because the shock efficiently transforms the ordered motion of the streams into the expanding motion in directions perpendicular to the streams. The circularization efficiency decreases, if two colliding streams have a large ratio of cross sections and a large density contrast. But even in such cases, the main shock extends beyond the overlapping contact surface and the high pressure region behind the shock keeps the stream of the larger cross section from passing freely. Thus the stream-stream collisions are still expected to circularize the stellar debris rather efficiently, unless the ratio
- Open Access
Cluster Accretion Shocks as Possible Acceleration Sites for Ultra High Energy Protons below the Greisen Cutoff
Three-dimensional hydrodynamic simulations of large scale structure in the Universe have shown that accretion shocks form during the gravitational collapse of one-dimensional caustics, and that clusters of galaxies formed at intersections of the caustics are surrounded by these accretion shocks. Estimated speed and curvature radius of the shocks are 1000-3000 \kms and about 5 Mpc, respectively, in the $\Omega=1$ CDM universe. Assuming that energetic protons are accelerated by these accretion shocks via the first-order Fermi process and modeling particle transport around the shocks through Bohm diffusion, we suggest that protons can be accelerated up to the {\it Greisen cutoff energy} near $6\times 10^{19}$ eV, provided the mean magnetic field strength in the region around the shocks is at least of order a microgauss. We have also estimated the proton flux at earth from the Virgo cluster. Assuming a few (1-10) \% of the ram pressure of the infalling matter would be transferred to the cosmic-rays, the estimated flux for $ E \sim 10^{19}$eV is consistent with observations, so that such clusters could be plausible sources of the UHE CRs.Open AccessCluster Accretion Shocks as Possible Acceleration Sites for Ultra High Energy Protons below the Greisen Cutoff
Three-dimensional hydrodynamic simulations of large scale structure in the Universe have shown that accretion shocks form during the gravitational collapse of one-dimensional caustics, and that clusters of galaxies formed at intersections of the caustics are surrounded by these accretion shocks. Estimated speed and curvature radius of the shocks are 1000-3000 \kms and about 5 Mpc, respectively, in the $\Omega=1$ CDM universe. Assuming that energetic protons are accelerated by these accretion shocks via the first-order Fermi process and modeling particle transport around the shocks through Bohm diffusion, we suggest that protons can be accelerated up to the {\it Greisen cutoff energy} near $6\times 10^{19}$ eV, provided the mean magnetic field strength in the region around the shocks is at least of order a microgauss. We have also estimated the proton flux at earth from the Virgo cluster. Assuming a few (1-10) \% of the ram pressure of the infalling matter would be transferred to the cosmic-rays, the estimated flux for $ E \sim 10^{19}$eV is consistent with observations, so that such clusters could be plausible sources of the UHE CRs. - Open Access
Diffusive Shock Acceleration Simulations: Comparison with Particle Methods and Bow Shock Measurements
Direct comparisons of diffusive particle acceleration numerical simulations have been made against Monte Carlo and hybrid plasma simulations by Ellison {\it et. al.} (1993) and against observations at the earth's bow shock presented by Ellison {\it et. al.} (1990). Toward this end we have introduced a new numerical scheme for injection of cosmic-ray particles out of the thermal plasma, modeled by way of the diffusive scattering process itself; that is, the diffusion and acceleration across the shock front of particles out of the suprathermal tail of the Maxwellian distribution. We find that all of these computational methods (diffusion-advection, two-fluid, Monte Carlo and hybrid) are in substantial agreement on the issues they can simultaneously address, so that the essential physics of diffusive particle acceleration is adequately contained within each. This is despite the fact that each makes what appear to be very different assumptions or approximations.
Open AccessDiffusive Shock Acceleration Simulations: Comparison with Particle Methods and Bow Shock Measurements
Direct comparisons of diffusive particle acceleration numerical simulations have been made against Monte Carlo and hybrid plasma simulations by Ellison {\it et. al.} (1993) and against observations at the earth's bow shock presented by Ellison {\it et. al.} (1990). Toward this end we have introduced a new numerical scheme for injection of cosmic-ray particles out of the thermal plasma, modeled by way of the diffusive scattering process itself; that is, the diffusion and acceleration across the shock front of particles out of the suprathermal tail of the Maxwellian distribution. We find that all of these computational methods (diffusion-advection, two-fluid, Monte Carlo and hybrid) are in substantial agreement on the issues they can simultaneously address, so that the essential physics of diffusive particle acceleration is adequately contained within each. This is despite the fact that each makes what appear to be very different assumptions or approximations.