72 lines
3.3 KiB
Text
72 lines
3.3 KiB
Text
|
Setup for computing buoyancy forces and moments of a 3d shape using the
|
||
|
|
Gerris Flow Solver, http://gfs.sourceforge.net/wiki/index.php/Main_Page
|
||
|
|
|
||
|
|
1. The hull of the vessel needs to be described by a 3d model.
|
||
|
|
The hull model must be a manifold
|
||
|
|
It is preferable if the origin of the 3d model is the same as the
|
||
|
|
origin of the FDM structural frame.
|
||
|
|
|
||
|
|
2. Convert the hull 3d model to GTS format for use with Gerris.
|
||
|
|
Example with the model originally in blender:
|
||
|
|
1. Export the model in STL format.
|
||
|
|
2. Convert the STL model to GTS format with the stl2gts tool (from the gts
|
||
|
|
library). For unknown reasons the normals have to be reverted in this
|
||
|
|
step.
|
||
|
|
% gts2stl --revert < floats.stl > floats.gts.base
|
||
|
|
3. Check the model with gtscheck
|
||
|
|
% gtscheck -v < floats.gts.base
|
||
|
|
If there are any problems, such as incompatible faces or non-manifold
|
||
|
|
edges, the model is very likely not to work in the latter stages.
|
||
|
|
Make a cleaned up version for the buoyancy computation.
|
||
|
|
|
||
|
|
The model.gts.base will be the original for the various cases/configurations
|
||
|
|
described below.
|
||
|
|
|
||
|
|
3. Edit the reference length and solid model scaling in the buoyancy3D.gfs
|
||
|
|
file.
|
||
|
|
|
||
|
|
4. Prepare the cases/configurations of interest. A case/configuration is one
|
||
|
|
position and orientation of the 3d model with respect to the water surface.
|
||
|
|
Each case becomes a separate gerris3D run.
|
||
|
|
The create_experiment.sh script can be used to create a set of cases.
|
||
|
|
The script needs to be edited to set the location of the hydrodynamic
|
||
|
|
reference point, HRP, (which should be close to the CG), the desired set
|
||
|
|
of data points in roll and pitch, and finally the distance between the HRP
|
||
|
|
and the water surface. The last parameter currently has to be manually
|
||
|
|
entered in the script file for each set of cases.
|
||
|
|
Example:
|
||
|
|
(Uncomment the distance line "HAGL=1.2192 # 4ft")
|
||
|
|
% ./create_experiment.sh 20150529 p4ft
|
||
|
|
(Uncomment the distance line "HAGL=HAGL=1.524 # 5ft")
|
||
|
|
% ./create_experiment.sh 20150529 p5ft
|
||
|
|
...
|
||
|
|
This creates a base directory 20150529 with a multitude of p4ft_* and
|
||
|
|
p5ft_* etc. subdirectories, each one being a case.
|
||
|
|
|
||
|
|
5. Run the cases.
|
||
|
|
% ./run_experiment.sh <base dir> <name base>
|
||
|
|
Example:
|
||
|
|
% ./run_experiment.sh 20150529 p4ft
|
||
|
|
You can also leave out the second argument to run all cases in 20150529.
|
||
|
|
Do consider running a suitable number of cases in parallel
|
||
|
|
./run_experiment.sh calls to speed up this step. E.g. as many as you have
|
||
|
|
CPU cores if you don't run out of memory.
|
||
|
|
|
||
|
|
6. Summarize the case results into tables that can be copied into a JSBSim
|
||
|
|
hydrodynamics configuration.
|
||
|
|
% ./summarize_experiment.sh <base dir>/<name base>
|
||
|
|
Will summarize the cases (roll and pitch) for one distance between the
|
||
|
|
HRP and the water surface.
|
||
|
|
This script must be edited to set the roll and pitch range and the
|
||
|
|
scaling factors used.
|
||
|
|
The resulting "coefficients" are scaled for imperial units and expressed
|
||
|
|
as F/(Rho*G) and M/(Rho*G).
|
||
|
|
|
||
|
|
Example:
|
||
|
|
% ./summarize_experiment.sh 20150529/p4ft
|
||
|
|
|
||
|
|
See Systems/c172p-hydrodynamics.xml to see where the tables for buoyancy
|
||
|
|
lift force and pitch and roll moment "coefficients" go and how the
|
||
|
|
"coefficients" are transformed back into forces and moments.
|
||
|
|
(Search for 'breakPoint="4.0"'.)
|