Topic: (C3D '09) Best Practices: Curb Creation

[Mark]

I know we've been over this before but ... what is the best way to create a curb line in C3D that will display correctly and be accurate in terms of the collected field points. Below is an image w/ notes. Hopefully that will help.

thanks

[mjfarrell]

that would depend on level of accuracy you want from your terrain model/figures..

in my classes we practice collecting these items in this manner:

Top Back of Curb
Top Face of Curb
Flow Line
Then Lip of Gutter if there is one....

and you use the point on curve command while collecting them...as explained in the Curve Codes in the Help file.

if I understand your question...

[mjfarrell]

if you aren't going to do it in the manner above, then one would try to use the Horizontal and Vertical offset commands and 'Template' that curb around....I do not like this function as it assumes ALL those offset distances and elevation changes are CONSTANTS until you tell it to STOP



in short you may need to refine you field practices to collect the data such that it draws what is there...using figures, and to do that most accurately you will need to evolve your process to work with C3D....

[Mark]

Our field crew doesn't collect the top face of curb so I have to create that line/points.

[Mark]

in short you may need to refine you field practices to collect the data such that it draws what is there...using figures, and to do that most accurately you will need to evolve your process to work with C3D....

"when Hell freezes over" would most likely be the answer.

[mjfarrell]

Our field crew NOW collects the top face of curb so I [color=red]DO NOT [/color] have to create that line/points.

They SHOULD be....and that is the method to getting the most out of the Field to Finish Functions...


otherwise one is reduced to MANUALLY creating a 'stepped' offset of that feature line to get ones model  accurate right from the collect data...it's faster and far more accurate than you 'faking' them in no matter how you decide to perform the function....

They (the surveyors) could even include an 'offset' note in the data, however this isn't a true offset, however it would draw that line for you.

have the guys take that shot.....

[mjfarrell]

in short you may need to refine you field practices to collect the data such that it draws what is there...using figures, and to do that most accurately you will need to evolve your process to work with C3D....

"when Hell freezes over" would most likely be the answer.

I would take them outside and let them collect the data in both fashions...and then process the line work to show them WHY they need to adapt...

in a few of my classes at South East Tech...there is a small concrete drainage feature....basically a flat bottom channel with curbs....in every class they would first shot top back of curb, center, and then top back of curb...this when added to the model creates a Vee Section....
then we go back and collect
Top back, top Face, Flow line, Center if defined,  Flow Line, Top Face, and then Top Back of curb, then we process that and See that the model actually develops a flat channel section with out any MANAUL functions at all....then they ALL change right there that day...because they see why.....
the model is more accurate
it is FASTER in the long run...
you are more productive, and they can anticipate raises and bonuses...

[Mark]

Our field crew NOW collects the top face of curb so I [color=red]DO NOT [/color] have to create that line/points.

They SHOULD be....and that is the method to getting the most out of the Field to Finish Functions...


otherwise one is reduced to MANUALLY creating a 'stepped' offset of that feature line to get ones model  accurate right from the collect data...it's faster and far more accurate than you 'faking' them in no matter how you decide to perform the function....

They (the surveyors) could even include an 'offset' note in the data, however this isn't a true offset, however it would draw that line for you.

have the guys take that shot.....

Without getting into a huge argument over field practices. I cannot tell the crew to take an extra shot. So I have to do this in C3D. Can the software do that? Can I 'fit' a feature line through a series of points to create a curve?

[mjfarrell]

compromise....have them shoot the Top Face, and NOT the back....

Mark trust me on this one....I've suggested the Best Practice really....

otherwise...one might try this process

build the surface,
add ONLY the Top Back Break figures as Break Lines...
Extract those Objects from the Surface, and then offset them...

I think that would work...as long as we both acknowledge that the ELEVATIONS for them is NOT exact....
I'd rather see you folks perform the survey as accurately as possible instead...given there is no reason to not get the data, and allow the softwae to do what you paid for it to do...

[Mark]

compromise....have them shoot the Top Face, and NOT the back....

Mark trust me on this one....I've suggested the Best Practice really....

I believe you Michael, but our field process doesn't revolve around C3D and that ain't gonna happen anytime soon.

otherwise...one might try this process

build the surface,
add ONLY the Top Back Break figures as Break Lines...
Extract those Objects from the Surface, and then offset them...

I think that would work...as long as we both acknowledge that the ELEVATIONS for them is NOT exact....
I'd rather see you folks perform the survey as accurately as possible instead...given there is no reason to not get the data, and allow the softwae to do what you paid for it to do...

I'll give that a try, if I can figure out what you're talking about. *grin*

When I said 'accurate' I meant in terms of the curve not so much the elevation.

[sinc]

Michael, your recommended workflow seems idealized to me...

The problems arise along curves.  True, we have things like Multicurves, but (unless Autodesk changed things while I wasn't looking) we can't collect multiple multicurves at once.  And that is further complicated by the nastiness that arises when breaklines cross.

So your recommended workflow of shooting TBC, TFC, FL, and Lip at every station would really start to lead to problems.  Unless you also recommend that field surveyors only collect linework in straight tangent segments, instead of using curve linework commands...?

[mjfarrell]

No Sinc; curves being the 'special' cases that they are do involve the requisite tedium...of collecting them more in sequence than linear features...however...they can and should be collected in the field


and I'm trying to recall what command, and exact sequence we used to take shots along all the features, and still get C3D to draw the arcs correctly....and I don't have (cant find) any of the students data, nor a collector here to go play with....
I'll try to follow up on this




For Mark,

uh..nevermind...

create the figures...then Export them as AUTOCAD objects.....
import and offset as desired....

[sinc]

Basically, it sounds like you are telling me what I already know...  Namely, that despite the improvements in C3D 2010, field linework in C3D is still pretty Mickey Mouse...   

[mjfarrell]

yes, and no...we did something a little different with one of the curve and feature code and the curves sorted themselves out....



...trust me I am looking for those files, and notes...and or brain cells...

[reno]

Am I missing something?  Import these points using the new linework commands in 2010:
79,2399.5198,1488.1455,0.0000,B FL
80,2405.2308,1485.2692,0.0000,B BOC
81,2433.3675,1541.1358,0.0000,BC BOC
82,2427.6564,1544.0122,0.0000,BC FL
83,2431.0347,1554.8388,0.0000,FL
84,2437.5252,1555.2301,0.0000,BOC
85,2435.9611,1571.6579,0.0000,BOC
86,2429.8579,1569.7494,0.0000,FL
87,2425.5674,1578.8360,0.0000,FL
88,2430.4727,1582.9957,0.0000,BOC
89,2419.5738,1593.7520,0.0000,EC BOC
90,2416.0083,1588.4438,0.0000,EC FL
91,2373.3361,1617.1059,0.0000,FL
92,2376.9015,1622.4141,0.0000,BOC

And you get curves with multiple points not taken in sequence.

[mjfarrell]

Thanks RENO...that's exactly what we did...I just reviewed (found) one of the student files....and what a great day to survey that was...December in Sioux Falls, SD...and just for fun t snowed on us for about 30 minutes of the the field lab.

[sinc]

That's good news.

We'll be moving up to 2010 pretty soon, so we'll be looking forward to that.  Multicurves in 2009 are pretty worthless.

[mjfarrell]

That's good news.

We'll be moving up to 2010 pretty soon, so we'll be looking forward to that.  Multicurves in 2009 are pretty worthless.

will there be snow when I visit you???      

[sinc]

Visit...?  Why, are you planning on heading out this way?

We have a little bit of snow at the moment, but it's melting fast.  It's supposed to get up to almost 50° today.  But I think it's supposed to snow again tomorrow.

[reno]

I think he wants you to pay him to come out and train you.   

[sourdough]

Mark,
I would like to know if turning your linework including curves, into feature lines for your surface.
I realize that you are trying to automate. If you are, have you found that even in 2010 that feature line curves
don't act properly. I have found that working with curves and using them as breaklines, just don't work. So, I
have been working the the tangent lines (after making them feature lines from fieldwork) and then inserting nodes along
the feature lines and making short tangent lines to show a smoother curve that can be used to create a surface.
I like the idea you are presenting here about automating. I have had better luck working with feature lines because
of the offset abilities to mold and hold the grade between points. Your thoughts?

mjp

[Mark]

Mark,
I would like to know if turning your linework including curves, into feature lines for your surface.

I only get points from the field, so I do all the line work. Yes I'm using feature lines but no curves, just lines from point to point. Working with curves and a surface is to much of a pain. fork-it!!

[sourdough]

Mark,
I would like to know if turning your linework including curves, into feature lines for your surface.

I only get points from the field, so I do all the line work. Yes I'm using feature lines but no curves, just lines from point to point. Working with curves and a surface is to much of a pain. fork-it!!

Mark,
  I haven't done any automated line work yet. The field crews I have run into so far haven't got the automation thing
in use yet. My question is, can the automated part you are working with connect the lines as feature lines?
 I have found that when going out in the field with the crew, that to many things interrupt the workflow for the automation to work
easily. I have seen some great data collectors that can create linework in the field for download, but money to reequip
is the biggest hurdle, not the willingness to get the features you are trying to use.

MJP

[sinc]

If you are using the F2F capabilities of C3D, then your linework is automatically created as Survey Figures.  These Survey Figures can be added to your surface as breaklines.  When you add the breaklines, you can specify a midordinate distance (I frequently use 0.1'), and the surface triangulation is created appropriately around curves.

And yes, there are times when this doesn't work so hot.  A big one is large buildings, where the field surveyor may shoot the building in pieces, from multiple setups.  Another problem is when the field surveyor can't see a shot or series of shots, such as around funky building facades, so the field surveyor hand-tapes measurements and records them in the field notes, rather than shooting each corner.  In such cases, it is often possible for the field surveyor to still get everything done with linework codes, but at a significant cost in time and complexity, and it's often better to do cleanup in the office.

But in general, F2F functionality is a significant time-saver.  It also eliminates much of the guesswork involved when an office person is trying to "connect the dots" for a survey done by a field surveyor.

[mjfarrell]

Mark,
I would like to know if turning your linework including curves, into feature lines for your surface.

I only get points from the field, so I do all the line work. Yes I'm using feature lines but no curves, just lines from point to point. Working with curves and a surface is to much of a pain. fork-it!!

Mark,
  I haven't done any automated line work yet. The field crews I have run into so far haven't got the automation thing
in use yet. My question is, can the automated part you are working with connect the lines as feature lines?
 I have found that when going out in the field with the crew, that to many things interrupt the workflow for the automation to work
easily. I have seen some great data collectors that can create linework in the field for download, but money to reequip
is the biggest hurdle, not the willingness to get the features you are trying to use.

MJP

This is where the demonstration/field application portions of my survey instruction are invaluable.
Because we work through the processes required to deal with 'that complexity' in the end it does make sense and it does get better through practicing it.  And I'm fair certain there would be no need to buy any new equipment to adopt the practices required.  ( I have worked for an been involved with collecting line work field codes for over 12 years)  So if your equipment isn't too much older than that you should still be good to go.