Tiered wall assembly

[Neilw_05]

I am wanting to model a tiered retaining wall with 5' high tiers, 10' apart at 2% grade between tiers. My strategy is to use a corridor with an assembly consisting of a retaining wall subassembly, a conditional cut/fill subassembly and either a daylight or a bench and another tier depending on the cut/fill condition (see attached image).

The problem I am having is the wall subassembly seems to require a target for the height. The help file indicates the height target is optional, so I expect that if I set the target to none in the corridor target settings, the wall height parameter in the subassembly will be used. When I do this the wall does not build. The only way to get it to build is to assign a height target. The only target I have is the existing ground surface and if I use that then the wall projects to EG at the first tier and thus no benching.

Can this approach be made to work without having to create some sort of height target? I'm atttaching the assembly (2012 version) for evaluation.

[jmayo-EE]

You can set up linkwidthandslope subs to form a uniform hgt wall. A 4 foot high wall with a 12" top would need a sub for the face of wall with a width of .1, a slope of 400% and a sub for the top of wall, width .9 and zero slope.

You can make a sub from a pline and there is also the subassembly composer at autodesk labs.

[Neilw_05]

Thanks for the suggestion John. Before going that route I'd like to know if the wall subassembly can work, partly because it would be easier to tweak the parameters to fit varying conditions. Did you get a chance to look at the assembly? It may be that it is just not set up properly.

[peterfunkautodesk · Accepted solution]

Neil,

The wall that ships was designed to go from the attachment point to a surface, with the profile for the height as an over rule. If you don't have the surface set, you won't get a wall.

If the walls are all the same, you might just go with a polyline SA and then the conditionals. Another way would be to make a wall in Subassembly Composer that could also do the conditional tests. I might construct one that does a wall and bench/daylight and then stack a bunch of them together (we don't allow loops in SAC because of the problems that those could cause).

Cheers,

Peter Funk

Autodesk, Inc.

[Neilw_05]

Thanks for clarifying that Peter. I'll give up on the wall SA. It would have been nice to have all the codes and links if it worked but I don't know that I'll need them. I might get by with just a benched daylight SA.

[Neilw_05]

I am re-visiting this post after putting this little task on the back burner, as I've not found a solution.

I've attempted to solve it using both gradings and a corridor, neither of which give satisfactory results.

My objective is to model a tiered wall with 5' maximum tier height, 10' separation between tiers and 2% bench slope. The teirs are to repeat until the depth from the base of wall to exsisting ground is less than the wall height (i.e. <5'). When that condition is met I want to slope to EG at 2:1. The wall is to be in cut only, Any portions that are in fill will daylight to EG at 2:1.

These are initial parameters based on a preliminary assesment of the site conditions by the architect (see attached sketch). Accordingly I want to be able to adjust all the parameters should the intitial parameters do not give a satisfactory solution. Also I will need to determine the square footage of the resulting face of the walls to provide a cost estimate to the client.

Here are the solutions I've tried along with the pros and cons

1. Corridor using a benched daylight subassembly

Pros: Assmblies can Intelligently determine whether to add a bench and tier based on depth of cut

Con's: Benches overlap at concave angle points along the wall (see screen capture). Adding fillets at the angle points is not an option.

2. Gradings

Pros: Benches automatically clean up where they over lap vs. corridor.

By using locked criteria I can chenge the entire wall by adjusting the locked grading criteria parameters

Cons: gradings lack intelligence to automate benching and daylighting. It becomes a manual process to adjustthe benching and daylighting when the critieria changes. Also gradings cannot miter corners of thebenches.

I've also considered creating offset alignments for each tier and applying the wall assembly to each bench independently but the downside is I would need to create a profile for each bench and I would have to manually adjust the regions for each bench on each tier.

So I'll throw this one out there for the gurus who like a challenge.

In the attached zipped file

EG.XML = existing terrain

Base C3D.dwg=Wall corridor (2012 format)

wall.jpg=Architects wall sketch

screenhunter=corridor bench overlap

[BrianHailey]

I haven't looked at the files but I think I may have a solution for you. It consists of a corridor with several baselines. You mentioned you could create alignments for each level of the wall but that would mean you would have to create profiles for each one, which is true however, you don't have to MANUALLY create the profiles. Create a corridor that steps back from the centerline or first level of the wall (or wherever) and have it create a surface. Use this surface with the next wall alignment and create a surface profile. Use the surface profile in the next baseline of the corridor and continue until you get to the top. You could use a set of conditional subassemblies to test if you are still in cut from the existing ground, if you are put in the wall, if not, grade back to existing ground.

Hope this helps.

[Neilw_05]

I did consider that idea Brian. The problem is the benches cross over each other when I use a corridor, so the same problem will also appear in the profiles of the benches.

I've been thinking about a hybrid approach by using gradings for the benches since they clean up the overlaps, but I figured with gradings I would lose the intelligence of the subassembly to determine where to apply the benches. However an idea just came to mind of using a conditional assembly to apply the first tier of the wall if the assembly is in cut and extract the resulting featureline to use for grading the benches. I would not apply a bench at all with the assembly. The one concern I have in that scenario is the top of wall featurlelines will apppear and disappear when I adjust the wall parameters resulting in corrupt gradings.

Actually I now recall why I figured that approach won't work. The reason being the corridor would drop to elevation zero wherever the baseline extends beyond the ends of the surface profile. It would be a manual process to adjust the regions for each tier.

[sboon]

My version uses a series of featurelines built using the Stepped Offset command, with a grading object surrounding the top level to tie into EG.  The problem I had was at the ends of your alignment, where the walls are either higher or lower than EG.  On one end I tried to run a slope at 1:1 up to ground.  On the other end I extended the feature lines until they intersected with a contour line.

[Anonymous]

I love trying to model stuff like this.  In this case I think it is possible but maybe not that easy and probably a little time consuming.  Also I don't have c3d on this computer so I could only look at your pictures. This shouldn't be very hard if you are familiar with subassembly composer. 

I think to solve this problem without using gradings you need to be able to define the breakline that you are essentially crossing in the overlap areas.  It is the line where the two 2% grades will meet, the "V" line...This line should bisect the angle between your walls.  For the assembly, first you need to figure the cut / fill part then go into subassembly composer and make an assembly, it will have several conditions, a surface target (o.g.), and an offset target ("V" lines). 

First it looks to see if it is 5' or more below the grade.  If it is than start the wall, else 2:1 to og and exit.

Assuming you are doing the wall, first draw the top part of the wall. 

Calc. the distance to the "V" line (aux. point), if it is less than 10 feet, do a 2% slope to the line and exit. <- This prevents you from overlapping.  Else, do a 2% slope for 10 feet. This point should be the last point that is showing in layout mode before you save the assembly.

Now just copy and paste a few of those into your assembly similar to what Peter said (because you can't loop) and I think you will get the result you are looking for.  You should be able to just select all of the "V" lines when you set the target and tell it to use nearest.

If you haven't used subassembly composer before, take this opportunity to try it.  This isn't really a hard one to make.  Subassembly composer opens up a whole new world of possibilities.

There will be a small area where this won't work, it is where the thickness of the wall crosses over the "V" line.  You could modify the code to account for this.  Also if you have an area where the 2:1 cut merges into a 2% cut between walls that will probably require grading.  Dang architects...

Anyway, that is my initial thought.  I hope it helps you.  I usually change my mind once or twice while making a model.

- Mark

[Neilw_05]

It is certainly a nice looking wall but with featurelines alone there is no intelligence to determine where a bench is needed. How do you know?

[Neilw_05]

I've never been in subassembly composer though I have seen a couple of videos at AU Online. It may do the job but it is beyond me. Since you have the skills could you put it together and show how it works? There have been a couple of other posts like this, so if you get it to work your assembly would be nice to have.

This is one of those cases where I wish we had support from an expert. Better yet, I wish the software didn't require such a complicated solution for a relatively simple task.

[BrianHailey]

Can you post up your alignment, profile, and surface as well? Otherwise we'll just be making stabs in the dark without being able to see exactly what you are running into. You can e-mail it to me if you would rather first name dot last name at cad-1.com

[Neilw_05]

The files were attached in my post on 4/8.

Thanks!

[sboon]

For those who are looking at a corridor solution I think that you might be missing a major issue at the corners.  It may be possible to deal with the inside corners but I don't see any way to build the outside corner where the baseline alignment is not perpendicular to the upper walls.  The pic below is from Neil's drawing with the frequency turned up to show what happens.

[Neilw_05]

You are dead right on that Steve. It was a concern to me but I felt that issue could be dealt with using infills, though not an ideal solution. Perhaps the offset alignment approach will give a better solution if it can be done efficiently.

The bottom line is corridors or the grading tools need some improvements.

[sboon]

IMHO this type of problem is more suited to grading than a corridor solution.  We've both commented previously on the lack of mitered corners for grading objects.  I spent some time playing around with how mitered and chamfered corners might be implemented in a future release, and found that it's more difficult that I would have imagined.

[Anonymous]

I'll see if I can put something together after work today, I have a few crazy dealines.  I'm anchorage so it might be pretty late at night your time.

[Anonymous]

I wouldn't worry about the outside corners, just throw in a very tiny like 0.005' curve and have it sample at a user defined station.

[Neilw_05]

It seems like a good idea but as the curves get projected to each tier the fillets become more pronounced. It might be workable though.