Designing a Surfboard Fin

[NormandC]

Hi TK,

The problem was with the "blade edge" of the fin meeting exactly with the bottom face of the box. Moving the box 0.01mm in the Z direction makes the boolean cut successful in FreeCAD v0.13.1830 with OCC 6.5.0.

It's a good thing that OCC 6.7.0 solved it, but it is still best to avoid situations like that.

[woodbuddha]

Thanks guys. Your knowledge of free ad is awesome.

TK

[triplus]

I am guessing this one did it:

Tolerance of checking distance between vertex and edge in BOPInt_Context::ComputeVE() increased by additional Precision::Confusion() to allow for interference to be detected in this particular case.

http://git.dev.opencascade.org/gitweb/? ... 80abbb51e4

Anyway it is nice to see demonstrations like this in real use cases on how improvement and bug fixes done to FreeCAD kernel affects FreeCAD itself!

Before:



After:

[NormandC]

Yes, it's encouraging to see that OCC does improve and fixes annoying bugs and limitations.

[woodbuddha]

Hi Guys,

Got another strange one for you all. I import a NACA symmetrical airfoil (e.g. 0012) from airfoil tools using Mark's macro as a Dwire. All is good. When I convert it to a bspline I get a flat spot on the bottom side of the airfoil. The strange thing is, that if I do a bspline conversion to a NACA0012 foil or above (0013, 0014, 0015, etc) I don't get a flat spot on the bottom edge. Interesting.

Not sure whether it'sa bug or not.

nacaairfoilafterbsplineconversion.png (147.55 KiB) Viewed 1450 times

nacaAirfoilsTest.fcstd

(38.63 KiB) Downloaded 51 times

TK

OS: Ubuntu 12.04.4 LTS
Platform: 64-bit
Version: 0.14.3342 (Git)
Branch: master
Hash: ca6d37202593752c0eeeb1ce56cb4d8c3baf27c0
Python version: 2.7.3
Qt version: 4.8.1
Coin version: 3.1.3
SoQt version: 1.5.0
OCC version: 6.7.0

[woodbuddha]

Here are the NACA files.

TK

nacaairfoils.tar.gz

(1.47 KiB) Downloaded 63 times

[woodbuddha]

Hi Guys,

Any luck with this one?

TK

[bejant]

Macro v1.5 seemed to work fine, but this is what I got when lofting probably due to my inexperience.

20140416a_woodbuddha.jpg (14.27 KiB) Viewed 1376 times

[bejant]

OS: Windows XP
Platform: 32-bit
Version: 0.14.3252
Python version: 2.6.2
Qt version: 4.5.2
Coin version: 3.1.0
SoQt version: 1.4.1
OCC version: 6.5.1

[triplus]

I get a flat spot on the bottom side of the airfoil.

I tested both versions of the macro on Ubuntu 13.10 using latest Development Release and can confirm that flat spot. It looks OK as Wire and converting to sketch gives predictable results. Converting to BSpline adds that flat spot. Editing created BSpline shows points are in the expected/correct space but the curve is a bit off. I do not know if this is related but using the tool to convert from Wire to BSpline does produce this:

Code: Select all

Coin error in SoGroup::removeChild(): tried to remove non-existent child 0x3ab2150 (Separator).

viewtopic.php?f=19&t=4057
viewtopic.php?f=20&t=4618

About the reason why the curve in resulting BSpline is a bit off and does not flow directly trough the points maybe Yorik or somebody else will have more insight on that. To explain what is going on.