The video below demonstrates how to define a new Coordinate System in Civil 3D.
This particular demonstration shows how to set up a Low Distortion Projection
for a small town in Colorado, using a Lambert Conformal Conic Single Parallel
projection, and the NAD83 datum as the basis for our new coordinate system.
After watching this video, you may also wish to watch the next part on how to
transform data between a State Plane Grid and
our new Low Distortion Projection. Also note that, at the bottom of this
web page, there is further information on how the scale factor was determined
for this Site.
The tasks illustrated in this demonstration may be performed in any version
of Civil 3D, and do not require the Sincpac C3D.
(Run time: 6 min 30 sec)
A Note on the Scale Factor
This section is an addendum for those who are confused about the choice of scale
factor in this video.
Basically, my goal was to scale the NAD83 datum up, so that the ellipsoid for my
new projection is running through ground level in my project. We then
use this "scaled-up ellipsoid" as the basis for a new grid, with its origin in the center of our "area of interest," aka
our "Site". (Note: in more-complicated usages, we may wish to have an
origin that is some distance from our Site. I won't go into that usage; I
just wanted to mention it.)
In order to determine how much to scale up my ellipsoid, I first need to know the
average ellipsoidal height for my project. To get this value, we take the
average NAVD88 elevation for our Site, and then add the average geoidal height
for our Site. We then come up with a scale factor using the following
equation, where R is the radius of ellipsoid at our Site, and P is
the average ellipsoidal height for our project:
scale factor = (R + P) / R
In this equation, it is almost always good enough to use an average curvature
radius for R, such as R = 20,906,000 ft. In this particular
project, the average ellipsoidal height for our Site is 6000 ft, so that's the
value we'll use for P. Using these values
to solve for our scale factor, we get 1.00028699989 and change.
However, when we did this, we scaled the ellipsoid all the way up to our project
elevation. This means that our grid scale factor is exactly 1 right at our
grid origin, and it gets greater than 1 as we get further from the origin.
Instead, we may wish to drop the projection a little bit further, so that it is
actually slightly below the surface of the Earth at the origin. This makes
our grid scale factor less than 1 at our project origin. As long as we
don't lower the projection too much, we can extend the range of our projection
without harming its accuracy.
In this particular case, I decided to multiply the scale factor by 0.999999.
This value results in an "effective zone" of roughly 16 total miles north and south
(8 miles in each direction) of
the standard parallel for my projection. This means that, anywhere within
the "effective zone", we can completely ignore the grid scale factor, and we
will introduce no more than 1 part-per-million of error. We could relax
this further. For example, if we multiply by 0.99999 instead, we drop the
surface even further. In this case, we would have an "effective zone" of
more than 70 miles, but our accuracy is down to 10ppm over that zone. We
also introduce another 1ppm for approximately every 21 feet of vertical
difference from the Project elevation we identified earlier. But as long
as we choose appropriate values, we should have a rather large area where our
Grid distances are essentially the same as our ground distances.
For this particular Site, we chose to multiply our calculated scale factor by
0.999999, which gives us 1.00028599866.
To make it a bit easier to deal with, I then rounded this number to 1.000286
exactly, which is the value you see me use in the video.
Note that if any of this is confusing to you, you may find it worthwhile to read
the papers
"Working with Grid Coordinates" (pdf, 0.5mb), "Solving Grid to Ground Problem" (pdf, 0.30mb)
and Sinc's AU2010 PowerPoint presentation "Custom Coordinate Systems" (pptx, 6.20mb)