hi

i cannot find in the poor documentation which is the best scemes and solver to use for transonic calculation, where there is supersonic region, sonic region and subsonic region( with possible recirculation)

numerically speaking the only good way to deal with that problem is to use a time marching technique because equation are always parabolic following time and whatever the flow (sub, sonic or supersonic).

in addition, information propagates trough the domain by means of waves at sound velocity. in subsonic low velocity region this velocity is very high, requiring a prohibitive time step compared to supersonic region. the solver should rescale this sound waves depending on the region (if low (quasi-incompressible) or high (highly compressible) velocity): one means to achieve that is to use matrix preconditionning.

secondly, some special care should be brought to the differenciation of the convective term: more upwinding in the supersonic region(hyperbolic) with a flux splitting method such as Roe allowing enough dissipation to capture shock waves.

Finally inlet and outlet boundary condition should take into accont the flow caracteristic: at outlet, if it is supersonic the variable should be extrapolated from the interior, while if it is subsonic, pressure should be prescribed...

SO , does PHOENICS embodies these capabilities??In this case, which solver , scheme options to take because there is no theory/help about these problems in the *"documentation"

thanks for your replies

There are a large number of library cases that exemplify the use of PHOENICS for compressible flows. These may be perused and selected from the pre-processor, and such cases include:

N111, N112, B522 (transonic flow in a nozzle) N113 (transonic underexpanded jet) B523 (supersonic wedge cascade) 911 (supersonic rocket exhaust plume) 159 (sonic underexpanded jet) B156 (inclined supersonic flow in a duct)

References on simulations of relevant flows using PHOENICS can be found on-line at

http://www.cham.co.uk/website/new/support/publish.htm.

These include:

Malin MR & Sanchez L One-dimensional steady transonic shocked flow in a nozzle. PHOENICS Journal of Computational Fluid Dynamics and Its Applications Vol. 1, No 2, pp 214-246. Published by CHAM 1988 88/22

Palacio A & Malin MR Simulation of supersonic flow past a double-wedge profile PHOENICS Journal of Computational Fluid Dynamics and its Applications Vol. 1, No 3, pp 371-408 1989. Published by CHAM 1989 89/4

Palacio A, Malin MR, Proumen N & Sanchez L Numerical computations of steady transonic and supersonic flow fields International Journal of Heat Mass Transfer, Vol. 33, No 6, pp 1193-1204, 1990. Published by Pergamon Journals 1990 90/25

Palacio A, Malin MR, Proumen N & Sanchez L Flowfield predictions of transonic and supersonic problems Proceedings of the 3rd PHOENICS User Conference-Dubrovnk. The PHOENICS Journal of Computational Fluid Dynamics and Its Applications Vol. 2, No 2, pp 202-218 1989. Published by CHAM 1989 89/

Mace ACH, Rogerson JS & Smith AG. *** Axisymmetic jet expansion into a cylindrical tube. Presented at the 1st International PHOENICS User Conference, Dartford, Kent 1985. Proceedings published Lecture Notes in Engineering 18,pp 227-238, 1986. 85/63

Smith AG, Ayris JN & Beasley J The modelling of blast wave propagation using PHOENICS. Published in the Proceedings of the 5th International PHOENICS Conference, Nice, 1992 and in The PHOENICS Journal of Computational Fluid Dynamics and its Applications, Vol. 6, No 3, pp 325-341 & Hi-Hx, 1993. 1992 92/28

Lee D-B Numerical simulation of oblique shock wave/turbulent boundary layer with wall heat transfer Proceedings, 'Seventh International PHOENICS User Conference', Seville, May 1997. The PHOENICS Journal of Computational Fluid Dynamics and its Applications, Vol. 10, No. 2, pp 179-193, 1997. 97/22

Fang MTC, Kwan S & Hall W Arc-shock interaction inside a supersonic nozzle IEEE Transactions on Plasma Science, Vol. 24, No. 1, February 1996 96/24

Bolot R, Coddet C & Imbert M The use of the PHOENICS code for plasma jet modelling Proceedings, 'Seventh International PHOENICS User Conference', Seville, May 1997. The PHOENICS Journal of Computational Fluid Dynamics and its Applications, Vol. 10, No. 3, pp335-352, 1997. 97/40

Some of the foregoing papers provide descriptions of the mathematical model and solution procedure. The following paper documents the use of higher-order schemes for convection discretisation, and includes some examples for transonic flows, and these cases are in the input library of examples.

43. Malin, M.R. and Waterson, N.P, 'Schemes for convection discretisation in PHOENICS', PHOENICS J., Vol.12, No.2, pp173-201, (1999).

thanks for the reply... i gonna investigate about the exemples and the different articles...

just wanted a quick view if phoenics could deal with a flow containing in the same time, supersonic region with shock -boundary layer interaction, and subsonic region with boundary layer separation (need elliptic function because parabolic solver cannot track recirculation)

thanks again. yann

This seems like the most relevant reference in terms of application area, and since it is published in the PHOENICS Journal, the Q1 input file ought to be included in the journal with the paper.

Lee D-B Numerical simulation of oblique shock wave/turbulent boundary layer with wall heat transfer Proceedings, 'Seventh International PHOENICS User Conference', Seville, May 1997. The PHOENICS Journal of Computational Fluid Dynamics and its Applications, Vol. 10, No. 2, pp 179-193, 1997. 97/22