Topic: Corridor driving me crazy

[KATHYC]

Thank you to all that are helping me.  I have been assign another project that has become more pressing, so I won't be back on the one I posted until early next week.  I'll post the final product when I'm done.  Jeff_M  I don't need your critism  everyone overlooks things every now and then.  I thought this was a place to ask question when your stumped.  Mike you were right on at the intersection  I didn't explain myself very well.  Scout I did miss the side bar, I knew I read it somewhere (book, webpage or somewhere but I overlooked in my attempt).

[Jeff_M]

Kathy, I apologize for sounding like I was criticizing you, that was not what I had intended. Yes, it did come out that way, I was more peeved at Michael for finding yet another way to jump on the text & authors of, what I find to be a very good teaching aide. I shouldn't have even posted in this thread, and even deleted a post I had composed earlier than the one I did post.

This IS a great place to ask questions and, usually, get quality feedback. Again, my sincerest apologies for making you think otherwise.

FWIW, I also had worked up an example of the final corridor based on your drawing, initial post, and the text from the book (as I understood it) and had planned to post that. But since Dana (scout) has chimed in, she will be better suited to that since she wrote it in the first place.

[John Mayo]

I though about staying out of this but then remembered why I don't come here much anymore.


The real issue is Michaels NEED to drag others into his own battle.

Appologies should go to the OP who most likely has no idea what is really going on here.

It's a good book that has help MANY.

[Jeff_M]

<deleted stuff>

Now I'm off to have a beer and relax. Have a good weekend everyone!

Oops, not so fast. I feel compelled to resond once more....

<deleted some more stuff>

And NOW I 'm off

[Edited to delete some useless stuff - Jeff]

[mjfarrell]

Casual observation as to how the book is ineffective; to which neither of you needed to respond, thus prompting me to repeat my opinion. 

There may not be a reason as you see it for my first post; that being the case let me appear unreasoned then.


And you will both take note that none of this impacts the speed nor care that I will affect in answering your questions at any future time.  If I were unreasonable, I would hold a grudge, and stop helping you out when in need. However that is not in my nature. As always I stand ready to be of assistance. I guess just don't ask me for a book review, or recommendation for books or movies because obviously we disagree there..

[mjfarrell]

To the concerned gentlemen; as you both felt so strongly that my only intent is or was to besmirch another I have removed those comments.

From this point forward however; if you take offense to an opinion expressed by this member, please remember that it is only an opinion and you need not comment any further on same.


So come on back and and enjoy the posts and information that you find helpful. And those post in which I express something you do not appreciate; I trust you will do the same thing I do from time time when I see others post information of comments not to my liking, I often do not hit the reply button and simply move on.  Much like a buffet take the stuff that you like and is good for you, and just walk quickly and quietly past the boiled chicken feet, hog nostrils, entrails, etc that you do not like, without pointing and saying " Ewwwww Gross Pig Lips" and don't upset the serving line.

[Jeff_M]

Hi Kathy, Scout suggested I go ahead and post my solution to your query. Here's a pic of what my result looks like.

As MF & scout had asked, I, too, thought that  the crown would likely curve from the point where the cul-de-sac begins to it's radius point. To achieve this, I created a new alignment "FG-crown" that follows the FG(1) alignment until the point of curvature, then curves the the left to the RP. I then created a profile, superimposed the FG(1) Finish grade profile to it, and created a new FG-Crown FG profile. For the actual crown profile through the new curve I just held the elevations of the original FG profile at the BC & RP for a straight grade between the two. This is easily altered if desired.

I then added BasicLaneTransition subassemblies the 2 partial assemblies. The width & slope do not matter since I set the transition toggle to "Change ofset and Elevation". Next up I added an alignment, using the TIN lines of the surface, along the assumed EOP, and created a profile of the EG from the surface (I didn't need to see this so I didn't add a ProfileView). I also made sure all of the Daylight slopes used the same params.

Finally I edited the Corridor properties. First I edited the Bulb region to set the targets for the LaneTransition to the FG-Crown alignment & Profile. For the 2 curb returns I Split them, using the intersection of the EOP alignment with the FG(1) alignment as the split point (you cannot snap to this, so a zoomed in eyeball pick works for this). Now I set the targets for first region of each return to be the EOP alignment & EG profile, for the second region of each I used my FG(1) alignment and FG profile as the targets. Lastly I used the Set All Targets to set the EG as the target for the daylights. Note that the ditches are funky at the curb returns, this comes from those portions being in a Fill condition,m whereas the remainder of the corridor is in Cut. ALso note that the daylighs fail at the end of the cul-de-sac due to the surface not completely covering this area.

There are other ways, I'm sure, to do this, but this is pretty much what I've used for some time now. One thing I would like to mention, and others may have something to say about this, too, is that ALL of my curb returns I define in the same direction. IOW, all curves curve to the right. The begin & end stations do not matter, but it allows me to use just 1 Assembly (aptly named Base Curb Return) for all of my returns.

I hope this helps to clarify things. If you still get stuck & you'd like to use the drawing, just say so & I will post it.

[mjfarrell]

I suggest that users not create ALL curb return alignments in a single direction. The method I employ is to create those alignments in what ever is the natural direction of the baseline alignment. As the process of adding baselines and regions, etc is still a fundamental process that will be required in other areas of various corridors there is no sound reason to avoid doing so for these features. Yes, this requires a user to create a Right and Left Assembly for the purpose, this too is a fundamental skill the user will need to acquire and to some degree easier to learn and use than to 'remember' to draw all returns in a single direction, and the names of the assemblies let the user know to use the Return Left, on the anticlockwise returns, and the Right Return assembly when the returns, knuckles, or bulbs turn to the right. (Notice: even autodesk placed a right and left return assembly on the default tool palette in 2009 for this very reason)

You might have also pointed out to the user that you increased the sample frequencies in all those regions to some short distance to get all those projection lines. And the reason that you DID NOT use the 'superimposed' profile as your target in the corridor model is that quite often it FAILS to work correctly; thus you essentially 'traced' over it to create a profile by layout to prevent this failure in C3D from being an issue. Additionally you did this to cover the fact that the superimposed profile is NOT and exact representation of the profile you used to create it from, due to it being created as just a bunch of segments depending on the accuracy settings one employs when creating it.


For grins Jeff, when one creates this 'fictional' crown alignment in this manner the corridor will work nearly as well using an assembly with offsets within that Bulb, and at the curb returns. Give that method a try and see how you like it. (Yes I know there are a couple additional targets to set that way) and right at the very end of the bulb the model gets a little weird using this method. This is due to autodesk NOT allowing us to actually use the offset alignment stations to control the offset group stationing within the corridor, a function sorely lacking. IMO

[Jeff_M]

I suggest that users not create ALL curb return alignments in a single direction. The method I employ is to create those alignments in what ever is the natural direction of the baseline alignment.

As I said, there are other ways to do this. I've tried numerous methods and found I prefer the "all CR's in the same direction with 1 Assembly" method.

You might have also pointed out to the user that you increased the sample frequencies in all those regions to some short distance to get all those projection lines.

Why would I say this when I didn't change any Frequency settings. Those you see are the same as the drawing Kathy posted.....

And the reason that you DID NOT use the 'superimposed' profile as your target in the corridor model is that quite often it FAILS to work correctly; thus you essentially 'traced' over it to create a profile by layout to prevent this failure in C3D from being an issue. Additionally you did this to cover the fact that the superimposed profile is NOT and exact representation of the profile you used to create it from, due to it being created as just a bunch of segments depending on the accuracy settings one employs when creating it.

Actually, the only purpose for the superimposed profile was to easily get the elevations at the BC & RP from the original profile. I could have gotten these a number of other ways, too. The crown in the bulb is the only place I used this crown FG as a target...I kept the original FG as the target elsewhere.

For grins Jeff, when one creates this 'fictional' crown alignment in this manner the corridor will work nearly as well using an assembly with offsets within that Bulb, and at the curb returns. Give that method a try and see how you like it. (Yes I know there are a couple additional targets to set that way) and right at the very end of the bulb the model gets a little weird using this method. This is due to autodesk NOT allowing us to actually use the offset alignment stations to control the offset group stationing within the corridor, a function sorely lacking. IMO

I use offsets for knuckles. I once tried them for a bulb and did not like the results what-so-ever. I WILL admit that I haven't done too much with c-d-s's (5 or 6) due to the local gov't agency determining that they are banned in any new development. They tend to "impede the flow of pedestrian traffic which leads to more vehicular traffic". 

[mjfarrell]

Just explaining the loginc behind doing this another way, as you say there are many, only in reality there are not that many ways with C3D's tool set.

It appeared that you had, as I did not really make a note of the sampling frequency, only the source of the users confusion within that particular corridor.

Merely pointing out a weakness of the superimposed profile tool. You will also discover that at times it does not project correctly into the destination profile views, often coming up short and being of little use.


As stated your bulb results were poor (or not to your liking) at the ends due to autodesk NOT allowing the user to actually use the offset alignment's stationing withing the corridor.

You need not defend your methods, as you said there are many, and I was offering an alternate. And in no way implied that yours was in error, flawed, or otherwise inadequate.   

[KATHYC]

Thank you for all your help.  I have worked through the intersection successfully.  As for the bulb, I have tried two alignments and it gave me a bowtie at the curve to the RP.  I then made 3 alignments around the bulb  ( I know overkill)  and it gave me a gap at the curve to the RP.  I know have tried one alignment and there is places that the asphalt is running over to the face of curb.  There is also extra triangulation at the curve going to the RP and an area at the lip of the gutter.  Can someone explain why this could be happening?  Thanks

[mjfarrell]

It appears some of your targeting, and or assembly construction assignments (left, for right) are twisted, so check your targeting carefully.

It is highly suggested you name them Target Me to this alignment and that profile, and to name your groups within the assembly so you know if you are Right or Left. And if you are left when you need to be right; well that would be wrong. Also some times I name stuff with no target, Target me Not, so I know to leave them alone.  Some might suggest this is over kill, however I am a human and can only really keep so much in my brain and the more complicated your corridor assembly usage becomes the more relevant good component naming plays in good corridor modeling practices.

[mjfarrell]

I have seen the file in question.  So very close!  After the repairs are made me thinks this kat will be well on her way to making some mighty fine corridor models!