ORD 2021R1 Update 1 (10.10.01.03)
I'm having an issue with element profiles, in this case projecting to the offside edge of a diverge nose (highlighted).
The mainline channel is profiled using 0% from the corridor element, its profile is on the left, and is as expected
The nose profile is created at -2.5% from the mainline channel using Profile By Slope From Element, but gives steps and bumps as can be seen on the left profile.
These irregularities are not present in the geometry, the superelevation, the profiles of the corridor elements, or the duplicate element of the mainline channel, so where are they coming from and how can I remove them?
Hi Euan,
Put a ticket in for this. We found this yesterday and looks to be a new issue in 2021 R1 when the Reference/Source geometry does not have a profile that spans its entire length and has a different bearing to the original.
Regards,
Mark
OpenRoads Designer 2023 | Microstation 2023.2 | ProjectWise 2023
Hi Mark,
Where do I do that?
It was doing this in 2020 R2 as well, I updated on Monday and got through checking all my DGNs were OK with the update, tried deleting the profiles first and reapplying them, then reconstructing the layout fully, then using a fully fresh DGN.
Surely the "Slope relative to target" option in the profile properties would determine which element normals for projection are taken from?
Euan
You can do that here: Bentley.com - Technical Support
If it was, we were lucky to have not seen it - i'm 100% sure it doesnt happen in 2020 R3 though.
I thought that as well, but it doesn't look to work that way. I'm yet to get any official info from Bentley as to what it actually does.
Tried even just extending the mainline crossfall, and even though the two source elements are parallel, it's showing skewed projections in the extension which aren't at a normal to any of the elements
The strange thing is that the taper of the diverge doesn't have the issue, its -2.5% comes in perfectly, no jumps, wiggles, or steps, while a constant offset of the element on the transition (spiral) at the same grade gives issues.
The difference seems to be that one is located on a transition and the other isn't, so it could be the transition itself or the superelevation change that's causing the skewed projections
Continuing the work since the process needs gone through and modelled with errors is better than not modelled at all, the Quick Profile Transition also gives a result which is not as it should be, introducing a 154mm step where it should simply tie-in to the levels and gradients at both ends (1000m radius between taper and nearside channel at the nose)
As an update, carrying on through the merges and diverges, the issues persists even on spirals where there isn't superelevation turnover, so it appears to be an issue with spirals and their normals.
On every single fillet the quick profile transition has had to be redone by hand, as the tool doesn't seem understand two grades with a curve between,
Also a different diverge is on a large radius curve so the spiral issue isn't present, but when projecting out to a complex that has a single offset until the end of the taper, then a varying offset for the transition between hard shoulder widths from the mainline and slip road, the issue's back, with steps and zigzags where there's no reason for them.
Looks like this is a similar issue caused by different geometric combinations. Something fundamental to slope projection looks to have been broken in this release.
One of our guys has resolved these issues (for now), by using the "Simplify Geometry" tool on the source Geometry Profile. It will lose its dynamics, but seems to get rid of the spikes. Not an ideal workflow, but will help in the meantime.
OK, that will help for the time being.
I've also found an issue with varying profiles, the first section is meant to change from -2.5% to +2.5%, and the second section is at +2.5% constant (both from the same element.
The varying section doesn't really vary, it just seems to follow the start slope and the end slope has almost no effect, even adjusting the end gradient to 100% it only goes just above the +2.5% profile. So how is it meant to work?
Do any of the profile tools work correctly?
haha they used to work really well.
Keep an eye on the station ranges for your linear. That tool looks to be working okay for me.
The distance for the varying profile 0-18.5m, enough to turn over a 3.7m lane through 5%, the constant section is from 18.5-150m, and the element it's offset and graded from is the same overall length, so I don't know how it can be interpreted a different way or the stationing could be defined to make it work.
It's quite frustrating.