CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Phoenics (http://www.cfd-online.com/Forums/phoenics/)
-   -   NAMGRD=CONV. compressible calculations (http://www.cfd-online.com/Forums/phoenics/51657-namgrd-conv-compressible-calculations.html)

yann August 22, 2003 10:19

NAMGRD=CONV. compressible calculations
 
hi,

i have to make compressible calculation in BFC, with shocks, transonic flow...

in several case library, the option NAMGRD=CONV is used in conjunction with UCONV=T ...

i understand UCONV=T is used to say that higher order scheme will be used , but what is the effect of NAMGRD=CONV , and should these 2 options work together and should be used together??

also, it is not clear, in documentation and examples, what is the energy formulation to be used for such calculation, mainly the problem is i cannot see the exact partial equation for eaach case... do u know what is the best formulation and why?

finally, i never find an example with correct boudary conditions: normally they should change following iterations: for a nozzle, for instance, the mass flow rate is not know a priori and the shoking or non-shoking of the nozzle should be computed. so total values should be prescribed (generally known) and mass flow rate should be computed from solution in fly: it can be done by giving a mass flow rate condition (with scalar P1) but C and V calculated in ground:

C= (2*gamma)/(W1*(gamma-1)) W1 at current iteration V= (Po*RHO1)/(RHO_o) with RHO1 at current iteration

the condition for W1 is COVAL(in,W1,ONLYMS,SAME)

for those who are interested, it works well.

but problems arise when energy is added. i thought to solve static enthalpy and prescribe for it a fixed value (in ground):

h1 = CP*To -0.5(W1)**2 with W1 at current iteration and in conjunction with the perfect gas law

in this way stagnation values are all given:Po, To, RHO_0

but the coupling involve some problems , do you know how to deal with the energy equation

cham technical support September 22, 2003 10:50

Re: NAMGRD=CONV. compressible calculations
 
1) NAMGRD=CONV does not activate a higher-order scheme as such, but rather it activates a different formulation of the algebraic momentum convection fluxes on staggered meshes. The default treatment is unsuitable for shock waves, as was discussed in: One-dimensional Steady Transonic Shocked Flow in a Nozzle M.R. Malin and L. Sanchez, PHOENICS Journal of CFD & its applications, volume 1, No2, pp 214-246, (1988).

2)Non-linear or linear higher-order schemes can be activated from the VR Editor Menu, and these schemes are documented in: Schemes for Convection Discretisation in PHOENICS ,PHOENICS Journal of CFD & its applications, volume 12, No2, M.R Malin and N.P Waterson pp 173-201, (1999).

3.)The energy equation may be solved in terms of the static enthalpy (specific) or the total enthalpy (specific), but under certain conditions the total enthalpy is constant, and in these cases it need not be solved.

4.) Your comment on the updating of inflow conditions for nozzle flows is very well known in CFD, and advice and/or demonstration examples have been given to maintained users in the past. I also believe that some advice has already been posted on this forum in the past. The library examples date back to the late 1980s and originally the nozzle cases employed such a boundary condition via GROUND coding. However, this was removed from the standard PHOENICS installation, because at that time there was no mechanism for running library cases with GROUND coding as part of the built-in installation. In any event, the simplified inlet condition is not really incorrect but rather it uses a deliberate approximation and thereby suffers some inaccuracy due to the non-adjustment of the inlet velocity to satisfy the specified total conditions. The original cases used to state that this approximation was being made, but I guess it has been lost with the passage of time. Please note that your use of SAME is approximate and therefore not strictly correct, as was mentioned on a previous posting on this forum. Incidentally, it would be a straightforward matter nowadays to update the relevant library cases by employing PLANT coding for such inlet conditions, but at the present time this is not deemed a high priority for the deployment of manpower.

5) Your last question makes little sense to us because we would have to examine your q1 and GROUND files to ascertain precisely what you are doing and then offer advice. If you are a maintained user, you can send your files and questions to the techncial support centre that issued your support contract, i.e. your local agent or CHAM UK.



All times are GMT -4. The time now is 16:30.