These five shock tube tests are from Toro's book (Toro, 1999, p.225)
and use the Euler Equations with dimensionless variables, and are
listed in Table 2.1. They have varying levels of
difficulty: test 1 is a modified version of Sod's test; test 2 is a
strong double rarefaction; test 3 has a very strong shock; test 4 has
three strong discontinuities; and test 5 is the same as test 3 except
that the contact discontinuity is stationary. Results are shown in
Fig. 2.1; the solid black line is the solution with
10,000 cells, and the lines with points are the solution obtained with
200 cells. All runs use a CFL number of and the artificial
diffusion prescription of Falle (1998) which includes some
shear and bulk viscosity to the momentum fluxes, using a viscosity
parameter of
. This only noticeably affects
Test 2 by helping the calculation at ; without added viscosity
there is a spike in the internal energy due to small errors in the
(very small) density and pressure. It also fixes any entropy glitches
in rarefactions.
It is interesting to note that the shock and contact are much better
resolved for Test 5 than for Test 3 because they are almost stationary
with respect to the grid. This is a feature of shock-capturing
schemes; adding extra diffusion would mitigate this.
Table 2.1:
Shock Tube Test Problems. For these hydro tests, the three
state variables are density, pressure, and normal velocity,
, and all tests use an adiabatic equation of state
with
. Results are obtained using a second order
accurate (in space and time) integration with a Courant (CFL)
number of 0.7 and an artificial viscosity parameter
.
Test
Left State
Right State
Toro 1
0.2
0.3
Toro 2
0.14
0.5
Toro 3
0.012
0.5
Toro 4
0.05
0.3
Toro 5
0.012
0.8
Figure 2.1:
1D shock tube results for adiabatic hydrodynamics. The
solid line shows the results using 10,000 grid cells, and the
points show results using 200 cells. The figures show results for
Toro's tests 1-5 from top to bottom (see Table 2.1).