Topic: Why are Civil plans 17,000 miles from the orgin?

[mjfarrell]

Civil/Survey = REAL WORLD

Architects = Fantasy

Ahh lets put that to the test.  Lets take a dozen civilians off the street that know nothing about our trades and explain to them our layering name conventions.  See which convention gets understood first.

That would depend entirely on the organization, and or if they had bought in to the NOT NCS for their works.

[Dent Cermak]

It's more like:
Civil/Survey = Model Space
Architects = Paper Space.
And it's a REAL long streatch to laud Architectural layer name conventions. X-DES-GRD-FLR-BEM-REV1-LATTADOODOO?

[sinc]

All this banter is fun and all, but to get back to the original topic, here's something that may provide some useful information about the whole theory around using Grid Projections like State Plane or UTM:

http://www.ejsurveying.com/Articles/Working_with_Grid_Coordinates.pdf

The basic idea is that effects of "curvature of the Earth" and "elevation differential" get automatically corrected for, by using a Grid System with Floating Scale Factor.  This lets us pretend that we're in a straight Cartesian grid, even though we're Surveying on the surface of the Earth (a sphere where the ground level potentially changes by a lot, especially in mountainous areas).  Net result is that we can use simple trig in our calculations, and all the other details are abstracted out into the Grid Projection + Geoid Model, where the computer takes care of the grungy details.  (There are no Surveyors who are members of the Flat Earth Society - except maybe as a joke - because we can easily see the effects of Earth curvature in our every-day work...)

[Jeff H]

Thanks Sinc,
 
That article helps to make sense of it all.

[Birdy]

This thread raises an issue that I have been pondering.  That is, AutoCAD's ability to accurately calculate points.  My understanding is that ACAD can accurately calculate to something like 14 significant digits. 

If your civil origin is at some predetermined municipal point on the map, and the clients house sits on a 5000 acre site, and is say, 45,000 square feet.  Said 'origin' may be rather far from the actual architects geometry.  Maybe a few miles, for arguments sake.  Now I come along to fabricate all the millwork in that house and I need accuracy to 0.1mm.  (Our CNC machines can achieve this quite easily I believe.)

If we all follow the origin started by the civil drawings, am I likely to run into problems with my drawings because I am relatively far away from (0,0,0)?

My question is, isn't it good cad practice to not draw so far away from the origin?  I'm not looking for an argument here, but an education. 

[mjfarrell]

Being distant form the origin typically only aggrieves autocad in trying to figure out hatch patterns.
This is due in part to how they are calculated.

Your mill work, and profiles for the Door Frames, and Chair Rails...would most likely be drawn as separate details.
These details would most likely be drawn from some arbitrary start point in a new file, or even at or near 0,0.

So the origin of the Civil coordinate system would have no impact on your millwork detail drawings.

[danallen]

My question is, isn't it good cad practice to not draw so far away from the origin?  I'm not looking for an argument here, but an education. 

The accuracy issue is important, that is why I think AutoCAD has insertion base INSBASE. As an architect I usually take the civic survey or base plan, save a copy to my working file, purge, color bylayer, etc, and set INSBASE to the lower left corner of the property line. Then I xref the plan with insertion point at 0,0,0. If I get an update from civil, I only have to fix INSBASE.

My building often has the same insertion point, or something logical like a grid line intersection, center of buiding, etc., though in the early design phase we may not know all those details and the propertly line is safer. The key is to keep a copy of the property line in my plan file so that there are no mistakes aligning the building on the site plan.

Dan

[mjfarrell]

My question is, isn't it good cad practice to not draw so far away from the origin?  I'm not looking for an argument here, but an education. 

The .......
Dan

All of the above is great, now tell him where you draw the details for the custom millwork.

[Kerry]

17,000 miles is about 3/4 the way around the world isn't it ?
... but if you've far enough North it could be next door.

... perhaps that demonstrates perfectly my ignorance of civil stuff.

 

[danallen]

All of the above is great, now tell him where you draw the details for the custom millwork.

That depends on how you set up your drawings. I draw all my details for a sheet in the same file, with viewports for each scale. We have a template with titleblock grids for all scales from FULL to 1/8", and the 1:1 scale has the left corner at 0,0

For elevations, I have 0,0 equal ELEV. 0'-0", so that I can just do an ID to check the height of something. The x direction usually is a grid, but not so important, just not 17,000 miles away...

Dan

[mjfarrell]

All of the above is great, now tell him where you draw the details for the custom millwork.

That depends on how you set up your drawings. I draw all my details for a sheet in the same file, with viewports for each scale. We have a template with titleblock grids for all scales from FULL to 1/8", and the 1:1 scale has the left corner at 0,0

For elevations, I have 0,0 equal ELEV. 0'-0", so that I can just do an ID to check the height of something. The x direction usually is a grid, but not so important, just not 17,000 miles away...

Dan

As you stated earlier you MOVE the Civil Plans  to suit your practices.
So how I set up my drawings has no impact on where you draw your millwork details.
As you said, you start at 0,0.
Which is pretty much what I said happens.

So the answer to Jim H; is that the accuracy of autocad is generally not going to be an issue in regards his millwork.

[danallen]

Which is pretty much what I said happens.

No disagreement, I was posting in response to Jim H's request for education/information.

Being distant form the origin typically only aggrieves autocad in trying to figure out hatch patterns.

This is easily fixed by setting SNAPBASE to a point near the hatch,  and updating the hatch geometry by HATCHEDIT. The hatch now has an internal relative origin near it and resolves correctly.

We do this all the time for custom tile patterns and controlling the start point.

Code - Auto/Visual Lisp: [Select]

1. ;;; HatchTouch - Update selected hatch pattern start to current SNAPBASE point
2. ;;;   (For realigning ceiling grids or brick modules, for example)
3. (defun-q c:HATCHTOUCH ( / hn hs ha)                                                     ;
4.   (princ "\nSelect hatches to TOUCH to update snapbase origin")
5.   (if (setq ss1 (ssget ":L" '((0 . "HATCH"))))  ;only allow inserts on unlocked layers
6.     (progn
7.       (foreach e1 (xyz_sel2lst ss1)
8.         (command "-hatchedit" e1 "properties" (XYZ_CMDACTIVE ""))
9.       ) ;foreach
10.     ) ;progn
11.     (princ "\nNo hatches selected")
12.   ) ;end if
13.   (command "select" ss1 "")
14.   (princ)
15. )
16. ;;;  (C.) 1996 by Vladimir Nesterovsky <vnestr@netvision.net.il>
17. ;;; sel2lst is also on Reini Urban's site (see URL in the quotings).
18. ;;; convert SELection set to LiST of e-names
19. (defun-q xyz_sel2lst ( sel / l len )
20.   (if (= 'PICKSET (type sel))
21.     (repeat (setq len (sslength sel))
22.       (setq len (1- len)
23.             l   (cons (ssname sel len) l)
24.       ) ;setq
25.     ) ;repeat
26.   ) ;if
27. )
28. ;==========================================================
29. ; Continue pausing until exited command mode
30. ; nil = pause
31. ; otherwise pass string to use
32. ;==========================================================
33. (defun-q XYZ_CMDACTIVE ( passcmd / )
34.   (if (null passcmd) (setq passcmd pause))
35.   (while (not (= 0 (getvar "cmdactive")))
36.     (command passcmd)
37.   ) ;end while
38. )
39.  
40.  

[mjfarrell]

I'm aware of the problem (hatch Patterns) and the fix for same (SNAPBASE)

Someone might find that of use.

[Birdy]

So, being 17,000 miles away from the origin has little or no affect on my drawings accuracy, except for maybe hatches, which can easily be remedied?  Is that correct?

Also, what benefit is it to me to be 17,000 miles away just because the civil drawings are?

[danallen]

For non-civil/survey drawings, I recommend that you do not put your drawings 17,000 miles away, but locate the original intentionally close. To deal with background drawings from others, use XREFs and change the INSBASE, or just move them after attaching.