Troubleshooting gradings

[Anonymous]

I have a surface built from gradings. I turned off automatic rebuild to make
some edits to the gradings. Now when I rebuild the surface, C3D crashes
every time. I tried turning the automatic surface creation in the grading
settings to off and then on in hopes it would recreate a valid surface, but
each new surface crashes as before. There is something in the gradings that
is causing the problem. How can I determine what is wrong? Is there any
diagnostic tool I can use? All I did was stretch some feature lines for a
parking stall to eliminate a stall. I've tried making adjustments, making
sure nothing crossed, etc. but it doesn't help.

[Anonymous]

Your assesement about the temp grading surface is correct. Good job.

Now to explain:

My intent was to control the grading of the parking by creating a baseline
alignment, profile it, create a feature line from the alignment and profile
and use it to create a grading for a drape surface for the pavement and pad.
By draping the pavement and pad edge onto this temp surface I can control
the grading entirely by the profile and temp grading object. Any changes to
the grading would simply entail re-draping the pavement and pad lines onto
the temp surface. To pick up the fine details of the profile such as high
and low points, I need to drape the feature lines with intermediate
vertices. If I omit the intermdiate vertices when draping, I will have to
manually apply the key grading points to the pavement and pad lines which in
effect negates the main advantage to using the temp surface. This technique
for grading can be very efficient. It is a shame that the software doesn't
handle it well.

Secondly, the reason the pavement lines are in several pieces is to
accomodate different types of curb. In some places high back/vertical curb
is required and ribbon curb in others. I felt it would be most efficient to
use seperate feature lines and rapidly apply the appropriate curb grading to
them. To try to model those variations from a single feature line would be
tedious. Again the software does not accomodate this scenario all that well.

Your suggestion to work from back of curb inward is a good one. It would
provide precise elevations at the key points for staking and minimize the
filleting effects of the gradings. However, to be manageable it would also
require breaking the feature line into pieces to accomodate the various curb
types. Still it may prevent some of the overlap issues as well since the
gradings would tend to contract rather than expand and encroach on each
other. I might give it a try.

Thank you for your assesment on this. It demonstrates the value of
experience over book knowlege.



__________ Information from ESET NOD32 Antivirus, version of virus signature database 4371 (20090826) __________

The message was checked by ESET NOD32 Antivirus.

http://www.eset.com

[Anonymous]

Intermediate vertices  are placed at each intersecting TIN triangle
line.  By having a temporary surface with the tessellation setting you
have yours at produces a good number of unnecessary points.  I would
suggest adding the weed vertices to your tools chest. 

In fact, just switch the order of operation around to add intermediate
elevations AFTER you have applied your gradings.  The image above is
adding the points after the fact and then weeding the unnecessary
points out.  The vast majority don't make a difference and only add to
the complexity of the model.



In my opinion only, I would prefer to manually add intermediate
vertices manually at locations that I know will need to be made, places
like the noses of the islands.



I don't think what I accomplished in a very short amounts of time was
tedious.  It was a matter of selecting one featureline and snapping to
stations start and end for each segment of curb needed.  This allows
the computer to understand the relationship each curb has to the other
and would allow you to place transitions between one type of curb to
the other.  By keeping the featurelines separate, you have allowed a
grading object to run along your defined design featureline, overlaping
and present the computer something very difficult to triangulate and
prioritize which featureline elevation will win. Thats why moving those
parts to separate sites also allows you to complete the grading model.



Its really a matter of battles.  I wish the software could handle
ungraceful exits better, but it doesn't.  I still wish to complete a
job and I am not yet ready to just push an EZ button yet and nor have I
dropped back to LDD in years.   If you want separate featurelines for
different curbs, move them to a different site.  Avoid the overlap
conflicts.



All in all, I think you have a good grasp on the grading tools and
would only suggest looking at adding a slight re-arrangement of your
workflow and to add some Roundup to your toolbox as well. 😉



Experience is only learned in the trenches and getting your hands
dirty. 

Matthew Anderson, PE CFM




neilw wrote:

Your assesement about the temp grading surface is correct. Good job.

Now to explain:

My intent was to control the grading of the parking by creating a baseline
alignment, profile it, create a feature line from the alignment and profile
and use it to create a grading for a drape surface for the pavement and pad.
By draping the pavement and pad edge onto this temp surface I can control
the grading entirely by the profile and temp grading object. Any changes to
the grading would simply entail re-draping the pavement and pad lines onto
the temp surface. To pick up the fine details of the profile such as high
and low points, I need to drape the feature lines with intermediate
vertices. If I omit the intermdiate vertices when draping, I will have to
manually apply the key grading points to the pavement and pad lines which in
effect negates the main advantage to using the temp surface. This technique
for grading can be very efficient. It is a shame that the software doesn't
handle it well.

Secondly, the reason the pavement lines are in several pieces is to
accomodate different types of curb. In some places high back/vertical curb
is required and ribbon curb in others. I felt it would be most efficient to
use seperate feature lines and rapidly apply the appropriate curb grading to
them. To try to model those variations from a single feature line would be
tedious. Again the software does not accomodate this scenario all that well.

Your suggestion to work from back of curb inward is a good one. It would
provide precise elevations at the key points for staking and minimize the
filleting effects of the gradings. However, to be manageable it would also
require breaking the feature line into pieces to accomodate the various curb
types. Still it may prevent some of the overlap issues as well since the
gradings would tend to contract rather than expand and encroach on each
other. I might give it a try.

Thank you for your assesment on this. It demonstrates the value of
experience over book knowlege.



__________ Information from ESET NOD32 Antivirus, version of virus signature database 4371 (20090826) __________

The message was checked by ESET NOD32 Antivirus.

http://www.eset.com

[Anonymous]

I finally got a chance to look at your drawing
Matt. Everything seems to be working now thanks to your cleanup so the
drawing is salvaged.

I tried redraping the pavement lines
with intermediate vertices and everything updated without crashing. However
the drawing became agonizingly slow whenever I tried to do anything. Even
with the weeding it is doggy. I guess that is what we have to live with when
using gradings. For now I'll have to use a combination of draping
and manually setting elevation points to get by.

Also, in regards to my comment about the multiple
curbs types from one feature line being tedious, you are right, it was not
a big deal on this site. What I was thinking when I made the comment was that as
a workflow it is more tedious than using individual feature lines. On
a larger site it could turn out to be a lot of extra work.

All in all this has been a good learning
experience. Thanks for all the work you put into this Matt.

By the way, if you would like to see what I was
trying to achieve, here is a screen recording of Power Civil changing the
grading of the entire site simply by applying different profiles to the
baseline. It is VERY fast.

href="http://screencast.com/t/B6m1FvTficJH">http://screencast.com/t/B6m1FvTficJH

Take care.

[Civil3DReminders]

I've addressed it before, detach the surface from the grading and then paste it into the design surface: http://blog.civil3dreminders.com/2009/04/stop-exploding-grading-objects.html

Christopher

[Anonymous]

I'm happy to report that I managed to get the
grading to perform as I wanted.

The technique is to create an alignment for the
edge of pavement (EOP). Next create a surface profile for that alignment,
sampling the temporary draping surface created from the baseline. Next, create a
feature line from the EOP alignnment and link it to the surface profile for the
drape surface. The result is a feature line that responds to and updates to any
changes made to the drape surface, which is controled by the baseline profile.
It works nicely in testing.

From that edge of pavement line I can create curbs
using gradings, adding target regions for variations in curb type. I'm pleased
with the results so far.

__________ Information from ESET NOD32 Antivirus, version of virus signature database 4381 (20090830) __________



The message was checked by ESET NOD32 Antivirus.



http://www.eset.com