ORD - Changing Curve Radii is Moving Curve PIs/Geometry Edit Challenge

Thanks so much for the update and the carefully worded process of what you did to fix your issue. Much appreciated!

Best Regards,

Mark

Thanks to all those who responded! This one proved tricky… but all the shared insight was a big help. Thanks to everyone for the information; I really appreciate it.

Using that information, I determined because the alignment end was on a curve, it could not have been defined as a fillet like the other curves. Even if I was able to extend the complex alignment on the correct tangent, the curve was still defined differently and thus difficult to edit. 

What I wound up doing was deleting the curve and then redefining it. This was my solution process to modify the ending curve of the alignment:

1. Backup the alignment DGN.
2. Draw a (dumb) line - Line 1 - representing a tangent off the end of the curve. This is to have a tangent to refer to, since if the curve gets flatter the alignment must lengthen. It’s not exact but is a best guess.
3. Mark the current curve PI by drawing another (dumb) line - Line 2 - from the current PI.
4. Under Modify --> Geometry Builder Edit, use the table editor to delete the end curve.
5. Deleting the ending curve moves the remaining geometry to the same start point. So, use the Transform command to move the remaining geometry back to the same location (usually by snapping to the *other* end point of the geometry).
6. Click on the complex alignment to get handles to appear. Then use the handle at the end of the tangent to move the end of that tangent to the end of Line 1.
7. Use the Modify --> Insert Fillet command on the newly extended tangent. For the vertex, I snapped to the end of Line 2. I specified the desired radius and completed the edit.
8. Line 1 was drawn as a best guess and was inevitably a little too long, so under Modify --> Geometry Builder Edit, use the table editor to delete the remaining tangent off the end of the new curve.
9. Similar to Step 5, deleting that ending tangent moves the remaining geometry to the same start point. So, use the Transform command once more to move the remaining geometry back to the intended location. I used the other end of the geometry this time too.
10. Under Modify --> start station, I reset my stationing so the stationing past the edit to the first curve remined the same.

It’s a bit of a workaround, but it works. Thanks again for all the help and tips!

A few notes to add here from what both Marks stated.  In some cases, the Table Editor will simplify geometry and clean up rules.  Once an alignment is designed, one way to quickly "rebuild" or correct a quirky alignment for easier manipulation is using the Import Horizontal Geometry from Ascii tool using the format below. (initial design report to help grab XY e.g. HorizontalElementsTable.xsl).  This has worked to clean up PRC, PCC, etc. in rather complex designs that wouldn't open in the Table Editor or modify easily.  I've had people combine so many random methods and various tools in building an alignment that the Table Editor didn't recognize!  

jcallicott,

I have had a lot of success using the "Geometry Builder Edit" command to modify any existing geometry.  I have used it on complex geometry edits such as adding a compound curve to at exiting simple curve.  I would recommend drawing the geometry with Microstation lines and arcs if possible which makes the edits a  lot simplier.  The interface takes a little to get use to but this command is my go to for any complex geometry edits.  The rules are basically ignored and lines and arcs can be inserted or removed easily.  It is worth a try.

I have use it to move alignments and extend alignments for long distances.  The command treats the alignment as a container which is similar to InRoads. Elements can be removed and replaced at will.

Hi jcallicott,

In my opinion, ORDs biggest weakness on the design side of things is exactly what you are highlighting here - specifically the inability to redefine the way Geometric elements are connected to each other, hence your pain. Looking at the data, the reason curve 2 is so easy to edit is because it is defined as a fillet, which uses the connecting straights to define its incoming/outgoing bearing so adjustment is easy (and predicable).

The first curve is a bit less predicable to edit, something all users, including myself have issues with because it may have been defined many different ways - as an Arc Between Points or Arc to/from Element, and how was that Arc defined, by bearing/direction, center etc. This is what all of us spend a lot of time trying to figure out during editing.

Looking in the Alignment properties, the first arc is an Arc defined by Through Point, Sweep Angle and some other points. This being the case, the Tangent Points of the Arc will not move on radius change by default. To change this behaviour, select the PI manipulator, change the Placement Method to "Start/Radius", enter the radius you want and lock it and when the command is accepted it will retain the PI point and update the arc to the new radius.

Next, to extend the start, select the "Move Point" manipulator at the start of the arc and you will see it wont let you move it (for some reason). Change the placement method from "Center/Radius" to "Start/Radius" and it will let you extend it without changing the radius as long its value is still locked.

In most circumstances we just give up and recreate the curve and use the "Complex Redefine" command. Definitely some work needed here to avoid us having to be detectives to figure out how to manipulate geometry!