Removing kinks at the start/end of horizontal curves

[Anonymous]

Hi,

I have been developing a project for students of my transport engineering course that involves re-designing roads around the university. In doing so I have been attempting to generate a solution myself. Part of this has been drawing horizontal curves into the route; however I have found an issue when transitioning out of the curves back to straight. When I click to represent the change in direction of the road and create the curve, once the curve is finished i click again to continue the road and create the new straight; however this creates a "kink" in the road which I cannot seem to smooth out or remove. The program does not consider these to be curves and when I try to adjust it using the on-map PVIs it only changes the geometry of the curve but does not smooth out the kink.

I have attached photos to show this.

How do I remove these kinks, or how do I ensure when drawing my horizontal curves that during transition these kinks don't occur.

Thanks,

Anthony Ferraro

[John_DeLeeuw]

@Anonymous, I think you are using the standard Planning Roads instead of the newer Design or Component Roads. Standard Planninng Roads don't have intelligent Alignment parts like the others have.

When you select your Alignment, do you notice any (blue) curve objects in your layout like below?

[Anonymous]

@John_DeLeeuw I am using the "two-lane" component road so I could also do superelevation and cut & fill analysis which is why I found it a bit odd. There are distinct curve objects which you can see in the attached photos; however they are spiral-curve-spiral instead of just curve objects.

Update: As an update - I removed some of those points designated by black squares and it appears to have smoothed out the curve substantially. Odd as I tried doing this initially and it gave me an error saying they could not be removed. It appears the issue is fixed now!

Nevertheless, for curiosity's sake, do you suggest that the curve object should be used instead of spiral-curve-spiral? In this case is there still an easy way to see the distinct transition curves to main curve and back again? Ideally we would like sutdents to be able to identify all features of horizonal curves.

[John_DeLeeuw]

@Anonymous, okay, that clears things up! I think something went wrong while attaching the new parts.

The correct procedure is starting near the endpoint of the already drawn linear part and then InfraWorks 360 automatically reqognizes that part and uses the already assigned Design Speed to create the new curve or spiral-curve solution. But take note of your Design Criteria which don't always alow for a certain radius according to your chosen Design Speed. 

You can also use a Right-Mouse-Button (RMB) while drawing to choose which solution you would like next:

[gchakri]

@Anonymous fromyour images it looks like that you have your alignment PI's so lose to each ther making "the rule driven" geometry a challenge and couldn't fit a curve based on standards. In such cases it fills a very small radius to fillet. 

Can you change, in the road properties UI on the right, its design speed to 40 kmph and road type to local? and then select a PI and right click to see "convert geometry" flyout and then choose "curve"?

If it doesn't work, could you please tell me which version of InfraWorks you are using?

Thanks

[gchakri]

@Anonymous Actually @John_DeLeeuw suggestion as you are laying out is right approach. Also make sure your road type is change to right type with right speed.

[Anonymous]

@John_DeLeeuw @gchakri - essentially I tried to draw the road as a continunous component because when I noticed that when i segmented the road according to straights, curves and back to straights, I wasn't able to get the vertical profile for the entire road, but rather only the sections. Would I be wrong to design it in this way?

The design speed for the entire road is 50km/hr and I designated it as a local road so the minimum curve radius should be quite low. I ended up removing a few of the PVIs and have been left with the following, which appears to have removed the kinks nicely. I left the first as a curve and second as a S-C-S. Overall I'm not sure if I am taking a short-cut in doing this, I don't want to have the students doing the wrong method. In the end we will be getting the students to verify the minimum curve radius for the design speed, as well as identify the different elements of horizontal and vertical curves based on the output of their designs using the program.