Here are my notes on the Vespa CFD pipeline I am using to generate results automatically. This is run on on
www.massive.org.au
Pipeline Overview
Here are the steps in the pipeline:
-
Assemble Vespa
-
Todo: Automatically generate variations of Vespa streamline parts
-
Select a number of parts from $PIPELINE_HOME/scooter_bits and combine them into a single STL model file
-
Place the model file in $PIPELINE_HOME/common/stls for reference
-
Calculate Frontal Area
-
Process the single STL module file with
CalculateFrontalArea
-
Generate OpenFOAM Case
-
Create a stub case in $CASE_HOME (based on Motorbike tutorial for now)
-
This may include automatic updates to things like snappyHexMeshDict for working out optimal mesh sizes
-
Set frontal area
-
cp -f $PIPELINE_HOME/forceCoeffs $CASE_HOME/system/
-
Update "Aref" to correct vaule (sed command probably)
-
Create PBS batch script for launching the job
-
cp -f $PIPELINE_HOME/common/run_openfoam ./$CASE_HOME/
-
Gather Results
-
Run OpenFOAM case and calculate power requirements from Cd and Aref
-
Todo: Based on results submit new jobs using optimisation techniques
Where:
$PIPELINE_HOME= my working dir
$CASE_HOME = the specific test dir
Pipeline Detail
These are working notes, so beware of mistakes and vagueness ;)
 |