3D Modelling

asheshouse asheshouse at btinternet.com
Sun Nov 20 10:13:26 GMT 2005

Just a couple of thoughts arising from my tinkering with Daren data.

Two additional data styles would probably cover all eventualities.

"pitch"  with three variables defined lx, ly, angle

those being long axis dimension, short axis dimension  and bearing  of 
long axis
This would be like "passage" but would give a correctly modelled 
vertical tube.

"chamber"  three variables left, up, down

a series of points in a loop arranged clockwise.

If the chamber was surveyed as radials from a central point then this 
could still be modelled using "chamber" but entering the data as two loops.

modelling of sharp corners may need some thought?


I've noticed other comments about the 3D model looking better if 
modelled as oval tubes rather than prisms. This would not necessarily be 
more accurate. Due to lack of detail data the 3D model is likely to 
always just be a representation rather than an accurate model. Compass 
gives three options, tubes, octagonal prisms, rectangular prisms. Which 
looks best depends on the form of the original cave. I would think it 
would be most useful to develop a fully featured solid modelling system 
based on a simple rect. prism. Once all the visualisation bugs are 
resolved in the rectangular form it _may be_ relatively simple to 
convert it to other cross section types.


Dave Ramsay

More information about the Survex mailing list