11 visualizações

Enviado por ilyesingenieur

- General Theory of Relativity
- ME1303-GAS DYNAMICS AND JET PROPULSION.pdf
- 1106
- BOOK - Cut-off_Grades.pdf
- Amath301-syllabus
- HEAT TRANSFER
- A Note on Shock Waves and Sedov's Explosion Problem
- Numerical Analysis.pdf
- Diffusion Processes on Nonholonomic Manifolds and Stochastic Solutions for Einstein Spaces
- 14v150g
- Evaluation of Dynamic Response of a Structure When Subjected To Blast Load
- Presplitting study.pdf
- A New Breakup Regime of Liquid Drops Identified in a Continuous and Uniform Air Jet Flow 2007
- bibliografia
- AERODYNAMICS OF AIRCRAFT
- 14 Experimentgdsgdal Tests of Gravitational Theory
- R. Paul Drake- Radiative Shocks in Astrophysics and the Laboratory
- Numerical Methods Assignment 2
- C. Michaut et al- Jump conditions in hypersonic shocks: Quantitative effects of ionic excitation and radiation
- Operator Splitting

Você está na página 1de 14

Extension of the Piecewise Parabolic Method to One-Dimensional

Relativistic Hydrodynamics

JOSE

M

a

MARTI * AND EWALD MU LLER

Max-Planck-Institut fur Astrophysik, Karl-Schwarzschild-Str. 1, 85740 Garching, Germany

Received July 25, 1994; revised April 28, 1995

form of the RHD equations. One advantage of this approach

is the possibility of using numerical techniques specially de-

An extension to 1D relativistic hydrodynamics of the piecewise

parabolic method (PPM) of Colella and Woodward using an exact

signed to solve nonlinear hyperbolic systems of conservation

relativistic Riemann solver is presented. Results of several tests

laws [9]. In fact, the codes described in Refs. [9, 12, 15] and

involving ultrarelativistic ows, strong shocks and interacting dis-

the code HLLE of [14] are based on Godunov-type methods and

continuities are shown. AcomparisonwithGodunovs methoddem-

an implementation of different approximate Riemann solvers

onstrates that the main features of PPM are retained in our relativis-

(readers lacking knowledge of Godunov-type methods and Rie- tic version. 1996 Academic Press, Inc.

mann solvers should consult, e.g., the recent book of LeVeque

[16]). Judging fromthe results of several test calculations shown

1. INTRODUCTION

in these references, it can be concluded that an accurate descrip-

tion of ultrarelativistic ows with strong shock waves can be

Relativistic hydrodynamics (RHD here after) plays an im-

accomplished by writing the system of RHD in conservation

portant role in different elds of physics, e.g., in astrophysics,

form and using Riemann solvers.

cosmology, and nuclear physics. In extragalactic jets emanating

In a recent paper [17] we have derived the exact solution of

fromcore dominated radio sources associated with active galac-

the Riemann problemfor ideal gases in relativistic hydrodynam-

tic nuclei [1] or in current laboratory heavy-ion reactions [2]

ics. Similarly to the classical (Newtonian) case, the solution

even ultrarelativistic ows are encountered. The necessity of

can be obtained by solving an implicit algebraic equation which

modeling such relativistic ows which also involve strong

determines the pressure in the intermediate states that develop

shocks is triggering the development of relativistic hydro-codes.

after breakup of the initial discontinuity. Our solution extends

The rst code to solve the RHD equations on an Eulerian

previously known particular solutions (i.e., [18, 19]) to the

grid was developed by Wilson [3, 4] and collaborators [5, 6].

general case of two arbitrary initial states. It can be used to

The code is based on explicit nite differencing techniques and

construct an exact Riemann solver and allows one to formulate

uses a monotonic transport algorithm to discretize the advection

a relativistic version of Godunovs method.

terms of the RHD equations. The stabilization of the code

On the other hand, the piecewise parabolic method ([20],

across shocks is accomplished by means of a von Neumann

PPM hereafter), is a well-known higher-order extension of Go-

and Richtmyer [7] articial viscosity. This code has been widely

dunovs method being used extensively in classical hydrody-

used in numerical cosmology, axisymmetric relativistic stellar

namics. Besides the use of an exact Riemann solver, the key

collapse, accretion onto compact objects, and, more recently,

ingredients responsible for the accuracy of PPM are a parabolic

in collisions of heavy ions. The codes acccuracy decreases as

interpolation of variables inside numerical cells and special

the ow becomes strongly relativistic (ow Lorentz factor,

monotonicity constraints and discontinuity steepeners to keep

W 2; see [5]).

discontinuities sharp and free of numerical oscillations. Finally,

Norman and Winkler [8] proposed a fully implicit treatment

solving Riemann problems for states averaged over the domain

of the equations in order to overcome the numerical problems

of dependence of the interfaces makes PPM second-order accu-

in the ultrarelativistic limit (W 1). Recently, several new

rate in time. In this paper, we present an extension of PPM in

methods for numerical RHD have appeared [915] which with

its direct Eulerian version to one-dimensional RHD, in which

the exception of [11] (smoothed particle hydrodynamics) and

all the key ingredients of PPM have properly been generalized.

[10] (ux-corrected transport method for the equations of rela-

The paper is organized as follows. In Section 2 we dene

tivistic magnetohydrodynamics) are based on the conservation

the RHD equations in an Eulerian reference frame in conserva-

tion form and describe the transformation of quantities from

* Present address: Departamento de Astronom a y Astrof sica, Universidad

de Valencia, 46100 Burjassot, Valencia, Spain. the Eulerian frame to the rest (proper) frame of a uid element.

1

0021-9991/96 $12.00

Copyright 1996 by Academic Press, Inc.

All rights of reproduction in any form reserved.

2 MARTI

AND MU

LLER

This transformation is a key ingredient in our algorithm. Section equations and throughout the whole paper the speed of light is

set equal to unity. 3 is devoted to a description of the reconstruction procedure,

the determination of effective second-order-accurate (in time) From Eqs. (4)(6) the following relation between p, v, and

the conserved quantities can easily be derived: left and right states dening the Riemann problems and the

calculation of the numerical uxes using an exact relativistic

Riemann solver. Several numerical test calculations including

v

S

D p

. (9)

the RHD generalization of the interaction of two blast waves

[22] are presented in Section 4. In Appendix I, the explicit

formulae used in the interpolation step are given. Finally, the

In the non-relativistic limit (v 0, h 1) D, S, and

analytical solution corresponding to the problem of the interac-

approach their Newtonian counterparts, , v, and E

tion of two relativistic blast waves introduced in Section 4 is

v

2

/2, and Eqs. (1)(3) reduce to the classical ones,

described in Appendix II.

v

x

0 (10)

2. THE EQUATIONS OF RELATIVISTIC

HYDRODYNAMICS

v

t

(v

2

p)

x

0 (11)

In an Eulerian reference frame, the equations of RHD of a

perfect uid in one spatial (Cartesian) coordinate x can be

E

t

v(E p)

x

0. (12)

written in conservation form as

The system of Eqs. (1)(3) with denitions (4)(8) is closed

D

t

Dv

x

0 (1)

by means of an equation of state (EOS), which we shall assume

as given in the form

S

t

(Sv p)

x

0 (2)

(13) p p(, ).

(S Dv)

x

0, (3)

In order to use the exact solution of the relativistic Riemann

problem derived in [17] we restrict ourselves to an ideal gas

EOS, i.e.,

where D, S, and are the rest-mass density, the momentum

density, and the energy density in a xed frame, respectively.

(14) p ( 1),

These variables are related to quantities in the local rest frame

of the uid through

where is the adiabatic index. A very important quantity de-

rived from the EOS is the local sound velocity c, which in our

D W (4)

case is dened through

S hW

2

v (5)

hW

2

p D, (6)

c

2

p

h

. (15)

where , p, v, W, and h are the proper rest-mass density, the

In any RHD code evolving the conserved quantities D, S,

pressure, the uid velocity, the uid Lorentz factor, and the

in time, the variables p, , , v have to be computed from

specic enthalpy, respectively. The latter two quantities are

the conserved quantities at least once per time step. In our

given by

approach, like in Refs. [9, 12], this is achieved using relations

(4)(8) and (14) to construct the function

W

1

1 v

2

(7)

(16) f (p) ( 1)

p

and

with

and

given by

h 1

p

, (8)

D

W

(17)

and where is the specic internal energy. Note that in the previous

PPM METHOD FOR 1D RELATIVISTIC HYDRODYNAMICS 3

ues. Then, we concentrate on the construction of left and right

D(1 W

) p(1 W

2

)

DW

, (18)

interface states averaged over the domain of dependence of

each interface, which are used as initial states for the Riemann

problem to be solved at each interface. Finally, we discuss the

respective, where

solution of the relativistic Riemann problem and the computa-

tion of the numerical uxes.

W

1

1 (S/( D p))

2

. (19)

3.1. Interpolation Procedure

The zero of f (p) in the physically allowed domain p ]p

min

,

In our relativistic version of PPM the interpolation algorithm

[ determines the pressure, the monotonicity of f (p) in that

described in the original paper by Colella and Woodward [20]

domain ensuring the uniqueness of the solution. The lower

is applied to zone-averaged values of V (p, , v), which are

bound of the physically allowed domain, p

min

, dened by

obtained from zone averaged values of the conserved quanti-

ties u

j

.

(20) p

min

S D, Interpolating in V instead of, for example, in u has several

advantages. First, in this case the solution of Eq. (16), which

involves an iteration, has to be computed only once per zone is obtained from (9) by taking into account that (in our units)

per time step, and second, one can easily avoid the occurrence v 1. Knowing p, Eq. (9) then directly gives v, while the

of unphysical values (i.e., larger than the speed of light) for remaining state quantities are straightforwardly obtained from

the interpolated ow velocity. Third, interpolating in p and Eqs. (4)(8).

simplies the implementation of the contact discontinuity

detector (see Appendix I). 3. A RELATIVISTIC VERSION OF PPM

Like in the Newtonian version of PPM, we determine for

Equations (1)(3) can be written in the form

each zone j the quartic polynomial which has zone-averaged

values a

j2,

a

j1

, a

j

, a

j1

, a

j2

to interpolate the structure inside

the zone, where a is one of the quantities p, , or v. Using this u

t

F(u)

x

0 (21)

quartic polynomial values of a at the left and right interfaces

of the zones, a

L, j

and a

R, j

, are obtained. These reconstructed

values are then modied such that the parabolic prole, which

with

is uniquely determined by a

L, j

, a

R, j

, and a

j

, is monotonic inside

the zone (monotonization). Finally, the interpolation procedure

(22) u (D, S, )

T

is slightly modied near discontinuities to produce narrower

jumps (see Appendix I).

and

3.2. Construction of Effective Left and Right States

(23) F (Dv, Sv p, S Dv)

T

.

To obtain time-averaged uxes at an interface j ' separating

In order to solve system (21) we consider the conservative

zones j and j 1, in PPM two spatially averaged states,

difference scheme

V

j1/2,S

(S L, R, where L and R denote the left and right sides

of the interface, respectively), are constructed, which take into

account the characteristic information reaching the interface

u

n1

j

u

n

j

t

x

F

j1/2

F

j1/2

, (24)

from both sides during the time step.

In the Lagrangian version of PPM this implies to calculate

the average of V over the domain of dependence of each inter- where u

n

j

and u

n1

j

are the zone-averaged values of the state

vector u of zone j at times t t

n

and t t

n1

t

n

t, face. In its direct Eulerian version, however, the construction

of the effective states is more complicated, because the number respectively. F

j1/2

are the time-averaged numerical uxes at the

right and left interfaces of zone j. In a Godunov-type difference of characteristics reaching the interface of a zone from a given

side can vary from zero to three. In a rst step, one computes scheme appropriate left and right states are constructed from

the zone-averaged values, which are then used to calculate the the average over that part of the domain of dependence of each

interface, which lies to the left and right of the interface. The numerical uxes F

j1/2

by solving the corresponding Riemann

problem. initial guess is then corrected by subtracting that amount of

each characteristic which will not reach the interface during In the remainder of this section we discuss in detail how the

numerical uxes are calculated in our relativistic version of the time step (see Fig. 5 in [20]).

In our relativistic version of PPM the correction of the initial PPM. First, we describe the interpolation procedure used to

reconstruct the variables inside zones from zone-averaged val- guess is obtained by closely following the procedure in [20], but

4 MARTI

AND MU

LLER

TABLE I

j1/2,S

1

2C

j1/2,S

vj1/2,S

v

j1/2,S

p

j1/2,S

p

j1/2,S

C

j1/2,S

(32)

Spectral Decomposition of Matrix A

Eigenvalues Left eigenvectors Right eigenvectors

0

j1/2,S

j1/2,S

0

j1/2,S

R

j1/2,S

p

j1/2,S

p

0

j1/2,S

C

2

j1/2,S

(33)

v c

1 vc

l

0,

1

2

,

1

2hW

2

c

W

2

c

, 1, hW

2

c

otherwise.

0

v r

0

(1, 0, 0) l

0

1, 0,

1

hc

2

The quantities a

#

j1/2,S

(S L, R; a p , v) are the averages

of a over that part of the domain of dependence of the

v c

1 vc

l

0,

1

2

,

1

2hW

2

c

W

2

c

, 1, hW

2

c

interface. They are calculated using the (monotonized) parabo-

Note. Left and right eigenvectors have been chosen such that l

#

r

#

##

;

lae determined for a in the interpolation step; i.e.,

#, # , 0, .

a

#

j1/2,L

a

R, j

x

2

a

R, j

a

L, j

6

1

2

3

x

(34)

considering the characteristic speeds and Riemann invariants of

the equations of relativistic instead of Newtonian hydrodynam-

a

j

a

L, j

a

R, j

2

,

ics. We rewrite system (21) in terms of V in characteristic

form as

where

V

t

A

V

x

0, (25)

x

max(0, t

#

(V

j

))

x

j1/2

x

j1/2

(35)

where matrix A is dened as

and

a

#

j1/2,R

a

L, j1

x

2

a

R, j1

a

L, j1

6

1

2

3

x

(36)

A

v

1 v

2

c

2

v

hW

2

(1 v

2

c

2

)

0

v(1 c

2

)

1 v

2

c

2

1

hW

4

(1 v

2

c

2

)

0

hc

2

1 v

2

c

2

v(1 c

2

)

1 v

2

c

2

. (26)

a

j1

a

L, j1

a

R, j1

2

,

where

Note that the local sound speed, c, dened in (15) has explicitly

been introduced in matrix A, whose eigenvalues

#

(# ,

x

max(0, t

#

(V

j1

))

x

j3/2

x

j1/2

. (37)

0, ) and corresponding left and right eigenvectors l

#

and r

#

are given in Table I. In analogy with [20] the effective left and

Note that we have dened x in Eqs. (35) and (37), and hence right states, V

j1/2,S

(S L, R), are then found to be

a

#

j1/2,S

, slightly different from Colella and Woodward ([20]; see

their Eq. (3.5)). However, the resulting

#

j1/2,S

, the only quanti-

p

j1/2,S

p

j1/2,S

C

2

j1/2,S

(

j1/2,S

j1/2,S

) (27)

ties depending on a

#

j1/2,S

, are identical to those of [20], because

they have to fulll Eqs. (30) and (31). With our modied

v

j1/2,S

vj1/2,S

C

j1/2,S

(

j1/2,S

j1/2,S

) (28)

denition of a

#

j1/2,S

, the quantities a

j1/2,S

appearing in Eqs. (27)

j1/2,S

j1/2,S

R

j1/2,S

(

j1/2,S

0

j1/2,S

j1/2,S

) (29)

to (33) are then simply given by

a

j1/2,L

a

j1/2,L

(38) with

a

j1/2,R

a

j1/2,R

. (39)

#

j1/2,L

0 if

#

(V

j

) 0 (30)

#

j1/2,R

0 if

#

(V

j1

) 0 (31) Finally, the quantities C

j1/2,S

and R

j1/2,S

are dened according to

C

j1/2,S

j1/2,S

h

j1/2,S

W

2

j1/2,S

c

j1/2,S

(40)

and

PPM METHOD FOR 1D RELATIVISTIC HYDRODYNAMICS 5

where

S

(p) (

S

(p)) denotes the family of all states which can

R

j1/2,S

2

j1/2,S

h

j1/2,S

W

4

j1/2,S

(41)

be connected through a rarefaction (shock) with a given state

S ahead of the wave (for more details see [17]). Because the

h

j1/2,S

, W

j1/2,S

, and c

j1/2,S

being given as functions of p

j1/2,S

,

Riemann invariants

j1/2,S

, and v

j1/2,S

(see Eqs. (8), (7), and (15), respectively).

It is worthwhile to note that the Newtonian limits of

j1/2,S

and

j1/2,S

coincide with the corresponding coefcients

J

1

2

ln

1 v

1 v

d (45)

in Eq. (3.7) of [20]. However, the limit of

0

j1/2,S

differs from

its Newtonian counterpart, because contrary to [20] we have

are constant through rarefaction waves propagating to the left

used the density instead of the specic volume as a variable

(J

) or right (J

for the characteristic equations.

3.3. Solution of the Riemann Problem and Computation of the expression

the Numerical Fluxes

In Godunovs approach the numerical uxes F

j1/2

are calcu-

S

( p)

(1 v

S

)A

( p) (1 v

S

)

(1 v

S

)A

( p) (1 v

S

)

(46)

lated according to

F

j1/2

F(u

j1/2

), (42) with

where u

j1/2

is an approximation to (1/t) u(x

j1/2

, t) dt, i.e.,

A

( p)

1 c( p)

1 c( p)

1 c

S

1 c

s

2/1

(47)

the time-averaged value of the solution at x

j1/2

, which is ob-

tained solving the Riemann problem at x

j1/2

with left and right

states V

j1/2,L

and V

j1/2,R

, respectively.

the () sign of A

In a recent paper [17] we have shown that an exact solution

(47), c

S

is the sound speed of the state S, and c( p) is given by

of a Riemann problem for a polytropic gas in RHD can be

obtained, in complete analogy to the Newtonian case, by solving

a non-linear algebraic equation.

c( p)

( 1) p

( 1)

S

( p/p

S

)

1/

p

1/2

. (48)

Both in relativistic and Newtonian hydrodynamics the dis-

continuity between the two constant initial states V

L

and V

R

decays into two elementary nonlinear waves (shocks or rarefac-

The family of all states

S

( p), which can be connected

tions), one moving towards the initial left state and the other

through a shock with a given state S ahead of the wave, is

towards the initial right state. Between the waves two new

determined by the shock jump conditions. One obtains (see

states V

L

and V

R

appear, which are separated from each other

[17])

through a contact discontinuity moving along with the uid.

Across the contact discontinuity, pressure and velocity are con-

tinuous, while the density exhibits a jump. As in classical

S

( p)

h

S

W

S

v

S

p p

S

j ( p)1 V

( p)

2

through rarefaction waves and the RankineHugoniot condi-

tions across shocks provide the conditions to link the intermedi-

h

S

W

S

( p p

S

)

S

W

S

v

S

j ( p)1 V

( p)

2

1

,

ate states V

S

(S L, R) with their corresponding initial state

(49)

V

S

. In particular, one can express the velocity of the intermedi-

ate states v

S

as a function of the pressure p

S

of these states.

where the () sign corresponds to S R (S L). V

( p)

The smoothness of the velocity across the contact discontinuity

and j ( p) denote the shock velocity and the modulus of the

then gives

mass ux across the shock front, respectively. They are given by

v

L

(p

) v

R

(p

), (43)

V

( p)

2

S

W

2

S

v

S

j ( p)

2

1 (

S

/j ( p))

2

2

S

W

2

S

j ( p)

2

(50)

where p

p

L

p

R

. This is the above mentioned non-linear

algebraic equation to be solved at each interface in each time

step. The functions v

S

(p) are dened by

and

v

S

(p)

S

(p) if p p

S

S

(p) if p p

S

,

(44)

j ( p)

( p

S

p)

h

2

S

h( p)

2

p

S

p

2h

S

, (51)

6 MARTI

AND MU

LLER

v

S

c

S

1 v

S

c

S

if p

p

S

V( p

) if p

p

S

,

(55)

where V( p

ing states V

S

and V

S

. Note that

S

S

. Hence, for

S

0

or

S

0, i.e., outside rarefaction waves, the solution of the

Riemann problem at a given interface is given by

V

V

S

if

S

0

V

S

if

S

0.

(56)

If

S

0

S

holds at an interface, the solution V

has to be

FIG. 1. Graphical solution in the p-v plane of the Riemann problems

evaluated inside a rarefaction wave (see [17]). In [22] the solu- dened by the initial states V

L

( p

L

10

3

,

L

1, v

L

0.5) and V

i

R

( p

i

R

,

R

1, v

R

0) (i 1, ..., 4) with p

1

R

10

2

, p

2

R

10, p

3

R

1, p

4

R

10

1

.

tion inside the rarefaction wave is obtained by interpolating

The adiabatic index of the equation of state is , in all cases. Note the asymptotic

linearly between the states on both sides of the wave, which

behaviour of the functions when they approach v 1.

is sufciently accurate for Newtonian problems. However, in

our relativistic variant of PPM we compute the exact solution.

Finally, the numerical uxes at each interface are obtained

where the enthalpy h( p) of the state behind the shock is the

according to Eq. (42). This concludes the description of our

(unique) positive root of the quadractic equation

method, which is of second-order accuracy in space and time.

It can be degraded to a rst-order Godunov method by simply

h

2

1

( 1)( p

S

p)

p

( 1)( p

S

p)

p

(52)

setting a

L, j

a

R, j

a

j

(a p, , v) in the interpolation step.

4. NUMERICAL TESTS h

h

S

( p

S

p)

S

h

2

S

0,

Traditionally, numerical methods for RHD have been tested

which is obtained from the Taub adiabat (the relativistic version

against two kinds of problems, namely wall shocks and shock

of the Hugoniot adiabat) for an ideal gas equation of state.

tubes, giving rise to ows with large Lorentz factors and strong

In Fig. 1, the functions v

L

( p) and v

R

( p) are displayed in a

shock waves. Thus, we have simulated these problems with

p-v diagram for a particular set of Riemann problems. In the

our relativistic version of PPM, too. In addition, we have calcu-

calculations presented here, Eq. (43) is solved for the pressure

lated the more challenging problem of the interaction of discon-

p

tinuities, i.e., the relativistic version of the collision of two

section and quadratic interpolation following the procedure

blast waves proposed by Woodward and Colella [22].

described in [25]. Once p

state quantities can easily be derived [17].

4.1. Shock Heating of a Cold Fluid

Now we are ready to determine the Riemann solution at a

given interface following a procedure analogous to [26]. Let

The initial setup consists of an inowing cold (i.e., 0)

sign(v

), where v

v

L

( p

) v

R

( p

), and set

gas with coordinate velocity v

1

and Lorentz factor W

1

, which

lls the computational grid and hits a wall placed at the opposite

edge of the grid. As the gas hits the wall, it is compressed and

S

L if 1

R otherwise.

(53)

heated up, converting its momentum into internal energy and

giving rise to a shock, which starts to propagate off the wall.

Behind the shock, the gas is at rest (v

2

0) and has a specic Then, we dene

internal energy

2

W

1

1. (57)

v

s

c

S

1 v

S

c

S

if p

p

S

V( p

if p

p

S

(54)

The compression ratio between shocked and unshocked gas,

2

/

1

, follows from and

PPM METHOD FOR 1D RELATIVISTIC HYDRODYNAMICS 7

FIG. 2. Exact (solid line) and numerical proles of pressure, density, and FIG. 3. Exact (solid line) and numerical proles of pressure, density, and

owvelocity of a relativistic shock tube (Problem1; see text). The computations ow velocity for the shock heating problem with an inow velocity v

1

0.99999, when the shock has propagated 50 zones off the wall (at x 1). The were performed on an equidistant grid of 400 zones.

computations were performed on an equidistant grid of 100 zones.

wall was placed at x 1. For numerical reasons, the specic

internal energy of the inowing gas was set to a small nite value

1

1

2

, (58)

(

1

10

7

W

1

). Figure 2 shows the proles of pressure, rest-mass

density and ow velocity for a run with a gas inow velocity

where is the adiabatic index of the equation of state. v

1

0.99999after theshockhas propagated50zones off thewall.

This test problem, which sometimes is also formulated as The proles obtained for other inowvelocities are qualitatively

the collision of two identical gases moving at equal speed similar. Themeanandmaximumerrors obtainedfor thecompres-

in opposite directions in order to avoid reecting boundary sion ratio are displayed in Table II for a set of inowvelocities.

conditions, has widely been used to check the accuracy of RHD It shows that with our relativistic PPM the mean relative error

codes [5, 8, 9, 12, 14, 15, 27]. Concerning explicit schemes,

r

() never exceeds a value of 10

3

and that, in accordance with

the numerical results improved signicantly for this test prob- other codes based on a Riemann solver, the accuracy of our re-

lem, when numerical methods based on Riemann solvers sults does not exhibit any signicant dependence on the Lorentz

were introduced. factor of the inowing gas.

In our test calculations we have used a gas with an adiabatic

4.2. Relativistic Shock Tubes

index , and inow velocities ranging from nearly Newton-

ian to ultrarelativistic values. The computational grid consisted

Shock tubes represent a special class of Riemann problems

of 100 equidistant zones covering the interval x [0, 1]. The

in which the initial state on both sides of the discontinuity is

at rest. They have become a useful tool in testing numerical

codes, because their evolution involves shock waves and rar-

TABLE II

efactions. We have simulated two particular shock tube prob-

Shock Heating of a Cold Gas Moving with a Velocity v

1

and a

lems characterized by the following initial states:

Lorentz Factor W

1

PROBLEM 1.

v

1

W

1

max

r

()

r

()

0.5 1.15 7.6188 4.3E 02 1.3E 04

L

10.0,

R

1.0

p

L

13.3, p

R

0

v

L

0, v

R

0.

0.9 2.29 12.1766 5.6E 02 6.4E 04

0.99 7.09 31.3552 6.9E 02 3.9E 04

0.999 22.4 92.4651 7.5E 02 8.6E 04

0.9999 70.7 285.8498 7.6E 02 5.9E 04

PROBLEM 2.

0.99999 223.6 897.4294 7.7E 02 3.4E 04

L

1.0,

R

1.0

Note. Maximum and mean relative errors of the compression ratio ,

max

r

() and

r

(), are given after the shock has propagated 50 zones off the

p

L

10

3

, p

R

10

2

wall. The zone next to the wall, which always dominates the maximum error,

has not been considered when calculating the mean error. v

L

0, v

R

0.

8 MARTI

AND MU

LLER

TABLE III

L 1 Norm Errors of Conserved Quantities and Convergence Rates

Corresponding to Problem 1 at t 0.36 for Relativistic PPM and

Godunovs Method (Superscript G)

x E(D)

1

E(S)

1

E()

1

r E(D)

G

1

r

G

_, 2.53E 01 3.97E 01 2.72E 01 4.64E 01

,',, 1.43E 01 2.54E 01 1.67E 01 0.82 2.89E 01 0.68

',, 7.51E 02 1.19E 01 7.80E 02 0.93 1.84E 01 0.65

,',, 3.98E 02 5.77E 02 3.83E 02 0.92 1.14E 01 0.69

,',, 1.94E 02 2.96E 02 1.93E 02 1.04 7.03E 02 0.70

,,,, 1.03E 02 1.96E 02 1.28E 02 0.91 4.39E 02 0.68

Note. E(a)

1

j

x

j

a

j

A

j

, where A

j

is the exact solution at x x

j

.

In both cases the adiabatic index is 5/3 and the initial

FIG. 4. Exact (solid line) and numerical proles of pressure, density, and

discontinuity is placed at x 0.5. owvelocity of a relativistic shock tube (Problem2; see text). The computations

were performed on an equidistant grid of 400 zones.

Problem 1 was chosen because it has been considered by

several authors [5, 6, 14, 15], whose results can directly be

compared with ours. For numerical reasons the pressure of the

right state has been set to a small nite value ( p

R

0.66 the L-1 norm error for different grid resolutions, together with

10

6

). The decay of the initial discontinuity gives rise to an the convergence rate for both the relativistic PPM and the

intermediate state located between a shock wave and a rarefac- relativistic Godunov method.

It is worthwhile to note that in both shock tube problems, tion propagating to the right (i.e., positive x-direction) and left,

the accuracy obtained on the nest grid with the relativistic respectively. The uid in the intermediate state moves to the

Godunov variant of our method is already achieved by the right with a velocity v

shell

0.72. Figure 3 shows the results

relativistic PPM with a four times coarser grid.

for a grid of 400 equidistant zones. One recognizes that the

shock is smeared across 45 zones and that the largest errors

occur for the postshock density. In Table III the errors of D,

4.3. Collision of Two Relativistic Blast Waves

S, and are displayed for different grid resolutions using the

The collision of two strong blast waves [23] was used by

discrete L-1 norm. Rening the grid the convergence rate of the

Woodward and Colella [22] to compare the performance of

solution (column 5 of Table III) indicates an order of accuracy

several numerical methods (including PPM) in classical hydro-

of code of roughly 1, which is expected for problems with

dynamics. In their test calculations the initial condition consists

discontinuities and which is in good agreement with the New-

of three constant states of an ideal gas with 1.4. The

tonian version of PPM [22]. This behaviour indicates that the

density is unity and the velocity vanishes everywhere in the

main features of the method are retained in our relativistic

interval [0, 1] covered by the grid. Reecting wall conditions

version. For comparison, we have also listed in Table III the

are used at x 0 and x 1. In the left state (x 0.1) the

errors of the rst-order accurate relativistic Godunov variant

pressure p 10

3

, while in the right state (x 0.9) p 10

2

of our method. The corresponding convergence rate is only

0.66 (see Table III).

Problem 2, usually referred to as the propagation of a relativ-

TABLE IV

istic blast wave, was rst considered by Norman and Winkler

[8]. The ow pattern is similar to that of Problem 1 but some- L 1 Norm Errors of Conserved Quantities and Convergence Rates

Corresponding to Problem 2 at t 0.36 for Relativistic PPM and

what more extreme. Relativistic effects reduce the postshock

Godunovs Method (Superscript G)

state to a thin dense shell. The uid in the shell moves with

v

shell

0.960, while the shock front ahead of it (the blast wave)

x E(D)

1

E(S)

1

E()

1

r E(D)

G

1

r

G

propagates with a velocity v

S

0.986. Norman and Winkler

[8] obtained very good results with an adaptive grid of 400

,',, 6.18E 01 1.09E 01 1.10E 01 7.06E 01

',, 4.94E 01 6.61E 00 6.43E 00 0.32 6.38E 01 0.15

zones using an implicit hydro-code with articial viscosity.

,',, 3.21E 01 4.25E 00 4.10E 00 0.62 5.45E 01 0.23

Figure 4 shows the results obtained with our relativistic PPM

,',, 1.78E 01 2.71E 00 2.67E 00 0.85 4.63E 01 0.24

on a xed grid of 400 equidistant zones. As in Problem 1, the

,,,, 1.00E 01 1.83E 00 1.89E 00 0.83 3.66E 01 0.34

largest errors arise in the postshock state. To achieve a con-

Note. E(a)

1

j

x

j

a

j

A

j

, where A

j

is the exact solution at x x

j

. verged solution a grid of 2000 zones is required. Table IV gives

PPM METHOD FOR 1D RELATIVISTIC HYDRODYNAMICS 9

FIG. 5. (1) Exact (solid line) and numerical density prole of the colliding relativistic blast wave problem before the interaction of the waves. The

computations were performed on an equidistant grid of 4000 zones. (2) Same as (1) but showing the ow velocity.

holds. In the central state (0.1 x 0.9) the initial pressure interaction occurs later than in the Newtonian problem, namely

at t 0.420. For the same reason, the dense shells are thinner, p 10

2

.

The early evolution is characterized by the development of having a width at the time of the collision of x 0.008

and x 0.019 for the left and right shells, respectively. two strong blast waves, which propagate into the cold central

gas. These waves are followed by thin shells of dense material. Consequently, the interaction is limited to an extremely narrow

region of size x 0.015. Because of the narrowness of the At t 0.028, the shells collide near x 0.69, resulting in a

multiple interaction of strong shocks and rarefactions with each structures one has to use of a very ne grid to resolve the

structures properly. In the calculations presented here we have other and with contact discontinuities. Much of the wave inter-

action takes place in a narrow region of size x 0.2. This used a grid of 4000 equidistant zones.

Figure 5 shows the density and velocity proles of the ow test is considered as a very severe one, in the sense that it

contains the most challenging ingredients that can appear in prior to the shock collision at time t 0.20. The relative error

in the density of the left (right) shell never exceeds 2.0% (0.6%) one-dimensional hydrodynamics, i.e., strong shocks, narrow

structures, and interaction of discontinuities. and has a value of about 1.0% (0.5%) at the moment of shock

collision. The quality of the numerical solution is drastically We have considered the same initial conditions to test our

relativistic code. While in [22] a special version of PPM is degraded when the simulations are performed with the Godunov

variant of our method. At t 0.20 the relative errors in the used to produce the most accurate solution for the interaction,

we have relied on the exact solution of the relativistic Riemann density of the left (right) shell are about 50% (16%) and drop

only slightly to a value of about 40% (5%) at the time of problem [17] to construct the analytical solution of the inter-

acting blast wave problem. Note that this exact solution is also collision (t 0.420).

The collision of the shells produces a region of very high used in our code to calculate the numerical uxes. Based on the

exact solution of the relativistic Riemann problem, an analytical density bounded by two shocks. The density jump across the

shock propagating to the left (right) is 7.26 (12.06); i.e., the solution to the blast wave collision problem can be obtained

for epochs prior to interactions with the rarefaction waves. For value lies well above the classical limit for strong shocks (6.0

for 1.4). In Fig. 6 a snapshot of the system is displayed this reason, we have used outow boundary conditions at x

0 and x 1 (to avoid the reection and subsequent interaction after the interaction has occurred. Compared to Fig. 5 a largely

different scale had to be used in the density plot of Fig. 6 to of the rarefaction waves produced by the initial data) and

stopped our calculations after the interaction of the leading include the narrowdense newstates produced by the interaction.

Obviously, our relativistic PPM code satisfactorily resolves the shocks.

The evolution of the systemis described in detail in Appendix structure of the collision region, the maximum relative error in

the density distribution being less than 2.0%. Using the Godu- II. The (dimensionless) propagation speed of the two blast

waves is slower than in the Newtonian case, but very close to nov variant of our method, the new states are much more

smeared out and the positions of the leading shocks are wrong the speed of light (0.9776 and 0.9274 for the shock waves

propagating to the right and left, respectively). Hence, the shock (see Fig. 6).

10 MARTI

AND MU

LLER

FIG. 6. (1) Same as Fig. 5(1) but after the blast wave interaction. Note the change in scale on both axes with respect to Fig. 5(1). (2) Same as Fig. 6(1)

but showing the ow velocity.

5. CONCLUSIONS zone-averaged values a

j2

, a

j1

, a

j

, a

j1

, and a

j2

. The expression

for a

j1/2

then reads

We have presented and tested an extension to one-dimen-

sional relativistic hydrodynamics of the well-known PPM

a

j1/2

a

j

x

j

x

j

x

j1

(a

j1

a

j

)

method of Colella and Woodward [20]. The results obtained

for problems involving ultrarelativistic ows, strong shocks,

and interacting discontinuities and the comparison with Godu-

2

k1

x

jk

2x

j1

x

j

x

j

x

j1

(59)

novs method demonstrate the superior accuracy and perfor-

mance of our relativistic PPMhydrodynamics code. In the code,

for the rst time, an exact relativistic Riemann solver is used

x

j1

x

j

2x

j

x

j1

x

j2

x

j1

2x

j1

x

j

(a

j1

a

j

)

to compute the numerical uxes across zone interfaces. The

modular structure of the code very easily allows the incorpora-

tion of approximate relativistic Riemann solvers, too.

x

j

x

j1

x

j

2x

j

x

j1

m

a

j1

x

j1

x

j1

x

j2

x

j

2x

j1

m

a

j

We also provide the exact solution of the relativistic counter-

part of the collision of two blast waves [22] (see Appendix II).

This solution can be used as a challenging 1D test case to

with

calibrate relativistic hydrodynamics codes.

Finally, we mention that the method can be extended in a

m

a

j

min(a

j

, 2a

j

a

j1

, a

j1

a

j

)sign (a

j

),

straightforward manner to treat also multidimensional relativis-

if (a

j1

a

j

)(a

j

a

j1

) 0, (60)

tic ows. A particular multidimensional relativistic PPM code,

0, otherwise,

based on an approximate relativistic Riemann solver has already

been developed and successfully applied to the simulations of

where

relativistic jets [28].

a

j

x

j

x

j1

x

j

x

j1

(61)

APPENDIX I: RECONSTRUCTION PROCEDURE

In this apprendix we give the details of the interpolation

2x

j1

x

j

x

j1

x

j

(a

j1

a

j

)

x

j

2x

j1

x

j1

x

j

(a

j

a

j1

)

.

procedure used in our relativistic version of PPM. Although

this procedure is identical to that in the original PPM formula-

tion, we will repeat the formulae here for completeness.

Using

m

a

j

, instead of a

j

, in Eq. (60) guarantees that a

j1/2

lies

in the range of values dened by a

j

and a

j1

. This calculation Step 1. First, interpolated values of a (where a stands for

any of the quantities p, , v) are calculated at all zone interfaces yields a value of a

j1/2

which is third-order accurate for nonequi-

distant grids, even where the zone size changes discontinu- j '. These interpolated interface values a

j1/2

are obtained

using the quartic polynomial uniquely determined by the ve ously [20].

PPM METHOD FOR 1D RELATIVISTIC HYDRODYNAMICS 11

In smooth parts of the ow, away from extrema, the limiting Step 3. Near strong shocks the order of the method is re-

duced locally to avoid spurious postshock oscillations. This is values of a at the left and right interface, a

L , j

lim

xx

j1/2

a(x)

and a

R , j

lim

xx

j1/2

a(x) are then given by the relation acheived by attening the distribution inside the corresponding

zones. In these zones the quantities a

L , j

and a

R , j

dened above

are substituted by a

L , j1

a

R , j

a

j1/2

. (62)

a

flat

L, j

a

j

f

j

a

L , j

(1 f

j

),

(70)

The values of a

L , j

and a

R , j

have to be modied later so that

the unique parabola dened by a

L , j

, a

R , j

, and a

j

is a monotone a

flat

R, j

a

j

f

j

a

R , j

(1 f

j

).

function in each cell, thereby introducing discontinuities at zone

interfaces (see Step 4).

The weight function f

j

is given by the maximum of f

j

and

f

js

j

, where

Step 2. The interpolation procedure described in Step 1 has

to be slightly modied to produce narrower proles in the

vicinity of a contact discontinuity. This process is called contact

f

j

min

1, w

j

max

0,

p

j1

p

j1

p

j2

p

j2

(1)

(2)

. (71)

steepening. As suggested in [20], we consider that a jump is

predominantly a contact discontinuity, if the condition

The index s

j

of f

js

j

is either 1 or 1 depending on whether

the difference p

j1

p

j1

is positive or negative. This way zone

K

0

j1

j1

min(

j1

,

j1

)

p

j1

p

j1

min( p

j1

, p

j1

)

(63)

j s

j

is always the next zone upstream of zone j, if the latter

is in a shock.

In Eq. (71), the quotient ( p

j1

p

j1

)/( p

j2

p

j2

) is a holds, where K

0

is a constant. In all zones j satisfying (63) the

measure of the steepness of the pressure jump across the zone density distribution is steepened by modifying the values a

L , j

j,

(1)

and

(2)

are constants, and w

j

is equal to 1, if the zone and a

R , j

according to

is inside a shock and zero otherwise. The criterion for a zone

being inside a shock is a

L , j

a

L , j

(1

j

) a

d

L, j

j

,

(64)

a

R , j

a

R , j

(1

j

) a

d

R, j

j

p

j1

p

j1

min ( p

j1

, p

j1

)

(2)

, v

j1

v

j1

. (72)

with

Step 4. Now we are ready to describe the monotonization

a

d

L, j

a

j1

m

a

j1

2

, a

d

R, j

a

j1

m

a

j1

2

, (65)

step (see Eq. (1.10) of [20]). In smooth parts of the ow, away

from extrema, a

L , j1

a

R , j

a

j1/2

, so that the interpolation

is continuous at x

j1/2

. However, near discontinuities the values

and

of a

L , j

and a

R , j

obtained in Step 1 to 3 are modied such that

in each zone j the interpolation parabola is a monotone function,

j

max[0, min(

(1)

(

j

(2)

), 1)]. (66)

which takes on only values between a

L,j

and a

R,j

. According to

[20] the following modications are necessary:

In this last expression,

(1)

and

(2)

are free parameters, while

j

is dened as

a

L , j

a

j

, a

R , j

a

j

if (a

R , j

a

j

)(a

j

a

L , j

) 0 (73)

a

L , j

3a

j

2a

R , j

if (a

R , j

a

L , j

)

2

a

j1

2

a

j1

x

j1

x

j1

(x

j

x

j1

)

3

(x

j1

x

j

)

3

a

j1

a

j1

a

j

a

L , j

a

R , j

2

(a

R , j

a

L , j

)

2

6

(74)

if

2

a

j1

2

a

j1

0, a

j1

a

j1

(67)

(1)

min(a

j1

, a

j1

) 0

(68)

a

R , j

3a

j

2a

L , j

if (a

R , j

a

L , j

)

0, otherwise,

a

j

a

L , j

a

R , j

2

(a

R , j

a

L , j

)

2

6

. (75)

where

Note that in RHD the monotonic character of the reconstruction

algorithmm ensures that the interpolated interface velocities

2

a

j

1

x

j1

x

j

x

j1

(69)

are always smaller than the speed of light, if this holds for the

zone averaged values, too.

a

j1

a

j

x

j1

x

j

a

j

a

j1

x

j

x

j1

.

The parameters K

0

,

(1)

,

(2)

, and

(1)

, introduced in Step 2,

12 MARTI

AND MU

LLER

TABLE V

Values of the Reconstruction Parameters Used in the Calculations

K

0

(1)

(2)

(1)

(1)

(2)

(2)

1.0 5.0 0.05 0.1 0.52 10.0 1.0

and

(1)

,

(2)

, and

(2)

introduced in Step 3 are, in principle,

problem dependent; i.e., their values have to be xed for every

calculation. Although the quality of the results depends on the

values of these parameters, a set of values can be found, how-

ever, which is well suited for a wide range of problems. The

parameter values used in all our test calculations are given in

Table V.

FIG. 7. Flow pattern of the colliding relativistic blast wave problem before

the interaction of the waves. The values of the hydrodynamical quantities in

the regions R1 to R9, which characterize the ow, are given in Table VI. They APPENDIX II: EXACT SOLUTION FOR THE COLLISION

have been computed using the formulae of Appendix II.

OF TWO RELATIVISTIC BLAST WAVES

In this appendix we describe the exact solution to the problem

and x

2

are the positions of the head and tail of the rarefaction of the collision of relativistic blast waves for the initial data

wave in region R2. They move according to given in Section 4.3. The problem of the collision of two blast

waves was introduced in classical hydrodynamics by Wood-

x

1

(t) 0.1 0.6324t (76) ward and Colella [22] to test the accuracy and performance

of various nite difference methods in case of a challenging

x

2

(t) 0.1 0.8222t. (77)

numerical problem. The multiple interactions of discontinuities

and rarefactions together with the resulting narrow ow struc-

Inside R2, i.e., x

1

(t) x x

2

(t), the distribution of the hydrody-

tures make this problem an extremely difcult test case for any

namical quantities can be obtained in two steps. First, one

Eulerian method. In its relativistic version, we have changed

solves the algebraic system of equations for the sound and ow

the boundary conditions, however (from reecting to outow),

velocity given by

to avoid the reection and subsequent interaction of rarefaction

waves. Thus, the ow structure is less complicated than in the

c

2

(x, t)

v

2

(x, t) (x 0.1)/t

1 v

2

(x, t)(x 0.1)/t

(78) Newtonian case. However, due to relativistic effects narrower

structures and larger jumps occur in the ow.

Early on, the ow consists of two blast waves created by

v

2

(x, t)

(1 v

1

)A

(x, t) (1 v

1

)

(1 v

1

)A

(x, t) (1 v

1

)

, (79)

the decay of the initial discontinuities at x 0.1 and x 0.9.

The two waves propagate towards each other and collide at

t 0.4200. We have used the procedure described in [17] to

solve the Riemann problems at x 0.1 and x 0.9 which

determine the solution before the collision. Up to this stage,

TABLE VI

the solution consists of nine different regions (R1 to R9; see

Constant States of the Relativistic Blast Wave Collision Problem

Fig. 7) linked at points x

1

to x

8

, where x

i

is the position of the

interface between regions Ri and R(i 1). Regions R1, R5,

Region p v c

and R9 correspond to the initial states, whereas regions R3 and

R1 1.000E 03 1.000E 00 0.00E 00 6.323E 01

R4 as well as regions R6 and R7 are the intermediate states of

R3 1.471E 01 4.910E 02 9.57E 01 6.321E 01

the Riemann problem dened by the initial discontinuity at x

R4 1.471E 01 1.439E 01 9.57E 01 5.591E 01

0.1 and x 0.9, respectively. The values of the hydrodynamical

R5 1.000E 02 1.000E 00 0.00E 00 1.163E 01

quantities in these constant states are given in Table VI. Finally,

R6 4.639E 00 9.720E 00 8.82E 01 5.002E 01

R7 4.639E 00 1.120E 01 8.82E 01 6.303E 01 regions R2 and R8 are rarefaction waves.

R9 1.000E 02 1.000E 00 0.00E 00 6.316E 01

Besides the values of the hydrodynamical quantities in the

constant states, the complete analytical solution must also give

C1 3.698E 02 1.044E 02 4.56E 01 6.084E 01

the position of the points x

i

(i 1, ... , 8) as a function of time,

C2 3.698E 02 1.173E 02 4.56E 01 6.056E 01

and the ow quantities inside the rarefaction waves. Points x

1

PPM METHOD FOR 1D RELATIVISTIC HYDRODYNAMICS 13

FIG. 8. (1) Sequence of snapshots showing the evolution of the density prole of the colliding relativistic blast wave problem up to the moment, when

the interaction of the waves occurs. The proles have been computed using the formulae of Appendix II. (2) Same as Fig. 8(1) but showing the ow velocity.

where x

7

and x

8

. The positions of all these points as a function of time

are given by

A

(x, t)

1 c

2

(x, t) 1 c

1

1 c

2

(x, t) 1 c

1

2/1

. (80)

x

5

(t) 0.9 0.9274t (85)

x

6

(t) 0.9 0.8820t (86)

Then, the isentropic character of the ow inside the rarefaction

can be used to get the density and pressure proles through

x

7

(t) 0.9 0.5668t (87)

x

8

(t) 0.9 0.6315t. (88)

2

(x, t)

c

2

2

(x, t)(c

2

1

1)

c

2

1

(c

2

2

(x, t) 1)

1/( 1)

1

(81)

Inside the rarefaction on the right, i.e., for x

7

(t) x x

8

(t),

the solution of the following algebraic system of equations and

p

2

(x, t)

2

(x, t)

p

1

. (82)

c

8

(x, t)

v

8

(x, t) (x 0.9)/t

1 v

8

(x, t)(x 0.9)/t

(89)

Point x

3

is the locus of a contact discontinuity. Hence, it moves

v

8

(x, t)

(1 v

9

)A

(x, t) (1 v

9

)

(1 v

9

)A

(x, t) (1 v

9

)

(90)

with the velocity of the uid in regions R3 and R4, i.e.,

(83) x

3

(t) 0.1 0.9570t.

with

Point x

4

, nally, gives the position of the shock heading the

left blast wave propagating towards the right; i.e., its motion

A

(x, t)

1 c

8

(x, t)

1 c

8

(x, t)

1 c

9

1 c

9

2/1

(91)

is governed by

(84) x

4

(t) 0.1 0.9776t,

gives the sound and uid velocity. Then, Eqs. (81) and (82)

allow one to calculate

8

(x, t) and p

8

(x, t), if the indices 1 and where the shock speed has been calculated from the Rankine

Hugoniot conditions for the jump between states R4 and R5. 2 are substituted by 9 and 8, respectively.

Figure 8 shows four snapshots of the evolution of the ow, The blast wave on the right has a similar structure, with a

heading shock at x

5

propagating into the central initial state, a including the moment of the collision of the blast waves at

t 0.4200. At this moment, the collision (of regions R4 and contact discontinuity at x

6

separating regions R6 and R7, and

a rarefaction wave propagating to the right bounded by points R6) occurs at x 0.5106, giving rise to two new states C1

14 MARTI

AND MU

LLER

REFERENCES

1. S. Phinney, in Superluminal Radio Sources, edited by J. A. Zensus and

T. J. Pearson (Cambridge Univ. Press, Cambridge, UK, 1987).

2. D. Strottman, in Relativistic Fluid Dynamics, edited by A. Anile and Y.

Choquet-Bruhat (Springer-Verlag, New York/Berlin, 1989).

3. J. R. Wilson, Astrophys. J. 173, 431 (1972).

4. J. R. Wilson, in Sources of Gravitational Radiation, edited by L. L. Smarr

(Cambridge Univ. Press, Cambridge, UK, 1979).

5. J. Centrella and J. R. Wilson, Astrophys. J. Suppl. Ser. 54, 229 (1984).

6. J. F. Hawley, L. L. Smarr, and J. R. Wilson, Astrophys. J. Suppl. Ser. 55,

211 (1984).

7. J. von Neumann and R. D. Richtmyer, J. Appl. Phys. 21, 232 (1950).

8. M. L. Norman and K.-H. A. Winkler, in Astrophysical Radiation Hydrody-

namics, edited by M. L. Norman and K.-H. A. Winkler (Reidel, Dor-

drecht, 1986).

9. J. M

a

. Mart , J. M

a

. Ibanez, J. A. Miralles, Phys. Rev. D 43, 3794 (1991).

FIG. 9. Details of the density prole of the colliding relativistic blast wave

10. M. R. Dubal, Comput. Phys. Commun. 64, 221 (1991).

problem showing the new states (regions C1 and C2; see Table VI) produced

11. P. J. Mann, Comput. Phys. Commun. 67, 245 (1991).

by the interaction of the two waves. Note the change in scale on both axes

12. A. Marquina, J. M

a

. Mart , J. M

a

. Ibanez, J. A. Miralles, and R. Donat,

with respect to Fig. 8(1). The prole has been computed using the formulae

Astron. Astrophys. 258, 566 (1992).

of Appendix II.

13. M. H. P. M. van Putten, J. Comput. Phys. 105, 339 (1993).

14. V. Schneider, U. Katscher, D. H. Rischke, B. Waldhauser, J. A. Maruhn,

and C.-D. Munz, J. Comput. Phys. 105, 92 (1993).

and C2 (see Table VI and Fig. 9). The solution of the Riemann

15. F. Eulderink, Ph.D. thesis, Rijksuniversiteit te Leiden, 1993 (unpublished).

problem dened by the states R4 (left) and R6 (right) allows

16. R. J. LeVeque, Numerical Methods for Conservation Laws (Birkhauser,

us to determine these new states, as well as the positions of Basel, 1992).

the two shock waves and the contact discontinuity (x

41

, x

42

, and

17. J. M

a

. Mart and E. Muller, J. Fluid Mech. 258, 317 (1994).

x

43

, respectively) which separates the newly created regions

18. K. W. Thompson, J. Fluid Mech. 171, 365 (1986).

from each other and from the former states R4 and R6,

19. B. J. Plohr and D. H. Sharp, in The VIIIth International Conference on

Mathematical Physics, edited by M. Mebkhout and R. Seneor (World

Scientic, Singapore, 1987).

x

41

(t) 0.5105 0.088(t 0.4200) (92)

20. P. Colella and P. R. Woodward, J. Comput. Phys. 54, 174 (1984).

x

42

(t) 0.5105 0.456(t 0.4200) (93)

21. S. K. Godunov, Mat. Sb. 47, 271 (1959).

22. P. R. Woodward and P. Colella, J. Comput. Phys. 54, 115 (1984).

x

43

(t) 0.5105 0.703(t 0.4200). (94)

23. P. R. Woodward, in Parallel Computation, edited by G. Rodrigue, (Aca-

demic Press, New York/London, 1982).

The solution described above applies until t 0.4300, when

24. R. Courant and K. O. Friedrichs, Supersonic Flow and Shock Waves

the next interaction (between states R4 and C1) takes place.

(Springer-Verlag, New York/Berlin, 1976).

25. G. E. Forsythe, M. A. Malcolm, and C. B. Moler, Computer Methods for

ACKNOWLEDGMENTS

Mathematical Computations (PrenticeHall, Englewood Cliffs, NJ, 1977).

26. P. Colella and H. M. Glaz, J. Comput. Phys. 59, 264 (1985).

This work has been supported by the Human Capital and Mobility Program

27. T. L. McAbee, J. R. Wilson, J. A. Zingman, and C. T. Alonso, Mod. Phys.

of the Commission of the European Communities (Contract ERBCHBICT

Lett. A 4, 983 (1989).

930496). Additional nancial support is acknowledged from the Spanish

DGYCIT (Ref. PB91-0648). One of the authors (J.M

a

.M.) would like to thank 28. J. M

a

. Mart , E. Muller, and J. M

a

. Ibanez, Astron. Astrophys. 281, L9

(1994). the Max-Planck-Institut fur Astrophysik for the kind hospitality during his stay.

- General Theory of RelativityEnviado porneo0000
- ME1303-GAS DYNAMICS AND JET PROPULSION.pdfEnviado porNarayana Samy
- 1106Enviado porDaniel Molano
- BOOK - Cut-off_Grades.pdfEnviado pordrina_pd
- Amath301-syllabusEnviado porjikixis
- HEAT TRANSFEREnviado pormanish_dce09
- A Note on Shock Waves and Sedov's Explosion ProblemEnviado porrasheedannou
- Numerical Analysis.pdfEnviado poraauppal
- Diffusion Processes on Nonholonomic Manifolds and Stochastic Solutions for Einstein SpacesEnviado porSergiu Vacaru
- Evaluation of Dynamic Response of a Structure When Subjected To Blast LoadEnviado porInternational Journal for Scientific Research and Development - IJSRD
- Presplitting study.pdfEnviado pornavarro_cvl
- A New Breakup Regime of Liquid Drops Identified in a Continuous and Uniform Air Jet Flow 2007Enviado por李海鹰
- 14v150gEnviado porJherson Nieto
- bibliografiaEnviado porIgnacio Lopez
- AERODYNAMICS OF AIRCRAFTEnviado porsreetejlakkam
- 14 Experimentgdsgdal Tests of Gravitational TheoryEnviado porwexler
- R. Paul Drake- Radiative Shocks in Astrophysics and the LaboratoryEnviado porGhoree23456
- Numerical Methods Assignment 2Enviado poraqstax
- C. Michaut et al- Jump conditions in hypersonic shocks: Quantitative effects of ionic excitation and radiationEnviado porGhoree23456
- Operator SplittingEnviado porJuanito Dada
- Ada 417311Enviado porUmut Cırık
- Lecture FourEnviado porMohamed Maher
- Detection of Black HolesEnviado porgeovanna prado
- C. Pfrommer et al- Cosmological structure formation shocks and cosmic rays in hydrodynamical simulationsEnviado porPogz1995
- 11344Enviado porArnalt Stalin
- Diekhoff Et Al 77Enviado porErre Rodríguez
- 1-s2.0-0377042789901532-mainEnviado porSuntood
- Andrew K. Henrick, Tariq D. Aslam and Joseph M. Powers- Simulations of pulsating one-dimensional detonations with true fifth order accuracyEnviado porSteemWheel
- Lecture 1Enviado porSyuhadah Noordin
- Paige Saunders2Enviado porHAYME THARRIS

- Analytical Methods in Conduction Heat TransferEnviado porMayank Sharma
- EdpEnviado porAndres Acosta
- An Improved Unsteady, Unstructured Artificial Compressible Finite Volume Scheme for Viscous Incompressible FlowsEnviado porilyesingenieur
- separation variableEnviado porilyesingenieur
- Fundamentals of CombustionEnviado poradityanarang147
- marine hydrodynamicsEnviado porAnu Parameswaran
- Separation Partial Derivation Fiffusion Parallel PlatesEnviado porilyesingenieur
- Numerical Methods in Heat, Mass, And Momentum TransferEnviado porPallabi Mitra
- fluid mechanicsEnviado porAlireza Zare'zade
- Origins of EntropyEnviado porilyesingenieur
- Popov Ppm GazEnviado porilyesingenieur
- CFD03 Flux FunctionsEnviado porilyesingenieur
- Riemann Solvers for Solving the Incompressible Navier-Stokes Equations Using the Artificial Compressibility MethodEnviado porilyesingenieur
- Tait Pure WaterEnviado porilyesingenieur
- Cranfield Univ CFDEnviado porilyesingenieur
- Non Conservative Conservative Incompressible Flows 2006Enviado porilyesingenieur
- A User Guide to Amr 1dEnviado porilyesingenieur
- Numerical Methods for Conservation Laws_-_LeVequeEnviado porEng. Mahmoud Al-Zahrani
- Num Meth for Comp Flows IMPOEnviado porgopalpaul
- Conservation_laws FVM Riemann Astro J LevequeEnviado porilyesingenieur
- Flow_control Blasius Eq Lam TurbEnviado porilyesingenieur
- Methode d Elements FrontieresEnviado porilyesingenieur
- pred-corr-tEnviado porilyesingenieur
- Comsol Tutorial Heat Conduction in a SlabEnviado porjohnnytulip

- 5-Guidelines for the Content of Rig Move ProceduresEnviado porRidamrut
- Chap09 Privacy Crime and Security HoEnviado porAprila Saputra
- Three County Fairgrounds zoning enforcement letter Atty Michael Pill to Northampton MA Builiding Commissioner Louis HasbrouckEnviado porMCSNews Archive
- ASHAA-81-00001-0001 Rev 2Enviado porGokulGokul
- inBIO480 Installation Guide V2 1-20120105 (3).pdfEnviado porchilenocl
- Econ161A Spring14 Assignment3 KeyEnviado porMatthew Kim
- Judicial AffidavitEnviado porJane Sue
- canonet_giii camera manualEnviado porDavid M231412312
- Assertiveness EssentialsEnviado porSrinathvr
- furrirEnviado porbnm007
- RMP2G 0007 SP 001 Structural Design Criteria Rev a 20110412Enviado porAndrea Rusconi
- 2014 Mobile Telephone Reimbursement PolicyEnviado porDebajyoti Ghosh Hazra
- Band OnEnviado porSyed Waqas Ahmed
- Tabel BinomEnviado porDaniel Boenfie
- Kargil ConflictEnviado porParrot
- Ip Video Security GuideEnviado porRubenMonarca
- có keyEnviado porKiệt Huỳnh Anh
- Shinohara Hedenquist 1997 Constraints on Magma Degassing Beneath the FSE Porphyry Cu-Au Deposit, Philippines.Enviado porEdwin Ayala
- 6Enviado portotochakraborty
- SRM Paper Final-- 8-20-10 (3)Enviado porgjhpe7779
- As 2550.3-2002 Cranes Hoists and Winches - Safe Use Bridge Gantry Portal (Including Container Cranes) Jib AndEnviado porSAI Global - APAC
- bikram cervoEnviado porVipul Gupta
- Robert Dudley v. John Stubbs, 489 U.S. 1034 (1989)Enviado porScribd Government Docs
- Fireforge Games Bankruptcy FilingEnviado porMichael Futter
- R/C Soaring Digest - Apr 2010Enviado porAviation/Space History Library
- r12 Frp FlowEnviado porshameem_fics
- Vortex SheddingEnviado poraishahrahman
- Act JusticeEnviado porKevin Scott
- Empirical Progressionof Lean Manufacturing: Literature ReviewEnviado porInnovative Research Publications
- Weld and Base Metal DiscontinuitiesEnviado porHari Hk