Difference between revisions of "FAQ/Solution procedure"

From OpenFOAMWiki
< FAQ
m (removed line break near the initial FAQ description)
(FAQ Section faqnumber::4: faqname::Solution procedure: Added " Why doesn't my solver have any residuals for the velocity equations? ")
Line 24: Line 24:
  
 
For Version 1.3: Recompile OpenFOAM with ''single-precision'' enabled (for details see [[Howto_compile_OpenFOAM]])
 
For Version 1.3: Recompile OpenFOAM with ''single-precision'' enabled (for details see [[Howto_compile_OpenFOAM]])
 +
 +
 +
=== Why doesn't my solver have any residuals for the velocity equations? ===
 +
# Edit the file {{tt|system/fvSolution}}.
 +
# Look inside the {{tt|SIMPLE}}, {{tt|PISO}} or {{tt|PIMPLE}} block that you have configured for the solver you're using.
 +
# You will find that you probably have {{tt|momentumPredictor}} set to {{tt|no}}.

Revision as of 13:47, 10 August 2015


1 FAQ Section 4: Solution procedure

Manipulating the solution procedure by non-programming means

1.1 How does one turn off the dimensional checking?

Dimensional checking is one of the cool features in OpenFOAM that helps to develop solvers that do something physicaly meaningfull. Therefore: don't do this if not absolutely necessary.

To switch it off at run-time set

dimensionSet 0;

in OpenFOAM-1.1/.OpenFOAM-1.1/controlDict.

A more controlled way is to use the dimensions()-method to set the dimension of the volXXXField in question to the desired dimension.

(source: [1])

1.2 Is there a way to switch between single and double precision?

For Version 1.2 and older: Only by editing src/OpenFOAM/primitives/scalar/scalar.H and recompiling OpenFOAM. But why would you want to do that (except for memory issues)? (source: [2])

For Version 1.3: Recompile OpenFOAM with single-precision enabled (for details see Howto_compile_OpenFOAM)


1.3 Why doesn't my solver have any residuals for the velocity equations?

  1. Edit the file system/fvSolution.
  2. Look inside the SIMPLE, PISO or PIMPLE block that you have configured for the solver you're using.
  3. You will find that you probably have momentumPredictor set to no.
Facts about "FAQ/Solution procedure"RDF feed
FaqdescriptionManipulating the solution procedure by non-programming means +
FaqnameSolution procedure +
Faqnumber4 +