LRUD (was Re: Roadmap)
Martin Green
mjg54@cam.ac.uk
Fri, 19 Apr 2002 16:42:04 +0100
Hi,
Rather than defining planes, an alternative would be to define up and down,
then a number of horizontal readings. It has the advantages:
-Of the user not needing to work out the perpendicular vector to a plane.
Which may not be easy with out the relevant background
-Stations can be handled that are missing some of the data.
-At a multi way junction, points on all of the walls can be recorded.
-It allows chambers with splay shots to be drawn reasonably.
A routine to do this could be
A 'horizontal' line can be defined around all of the passage, by starting a
node you go around it clockwise until a leg or a horizontal measurement is
found, if it is a measurment a line is drawn from the previous location, if
it is a leg, start going clockwise around that leg.
A ceiling and a floor line can be defined, by connecting all the up and down
points, along the surveyed lines, if a measurement is missing, then it can
be skipped, and the lines join to the next measurement along the passage.
Then at each point the points on the floor and ceiling lines are joined to
any horizontally measured points.
In this way, a number of shapes are defined, these will mainly be triangles.
Where they are not triangles they can be split into triangles.
Entering data:
The station LRUD you suggested could be a method of data entry. With a tag
before hand to specify wheather we are dealing with bisectors of first two
legs mentioned for the station (with an R afterwards if resersed), of
perpendicular to the first leg on which it was mentioned. a character canbe
used to mark a measurement as not read.
With a secound method which could be quite general, with multiple data items
seperated by commas. This could be written as
location of center, up, down, list of horizontal shots seperated by commas
Where:
location of center = station_name
or station_name_1 station_name_2
fraction_along_leg
up = distance
or distance compass clino
or <blank>
down = distance
or distance compass clino
or <blank>
horizontal shot = fore_leg back_leg fraction distance
or fore_leg back_leg fraction clino distance
or compass distance
or compass clino distance.
I hope that this was usefull, and not just me getting carried away. It may
be an idea to have a simple LRUD data format for now but allow potential to
expand to what I suggested above.
Alternatively LRUD could be ignored, until the drawing up stage, when a 3d
solid model should be producable from a drawn up plan and extended
elevation. Given LRUD will never show the full picture.
Martin