Hello all
It has been a while since I made a post, so here goes.
I have collected some survey data of a rail alignment and imported this into BRT.
I have created a curvature diagram so that i can see roughly where the constant radii, transitions etc fall.
Below is the curvature diagram, as you can see the radii is all over the place.
The Yellow line on the bottom is where I think the transition starts and ends and the red lines are where I think the constant curve starts and ends. But I cannot make sense of the rest as the radii is scattered
I was wondering if someone could give me their opinion on this curvature diagram.
Any help will be greatly appreciated.
Thank You.
Hi Stewart,
It looks like the accuracy of the imported points are poor (either rather poor track-geometry or something wrong in your workflow).
Check if the x- and y-coordinates of your imported points have the proper format/result and they have the decimals imported as well (i.e. not like *******.000) . If nothing wrong there, you may want to try to import every second or third point instead (change the "Minimum Distance Toleraance" in "Advanced" tab of "Add Horizontal Regression Points").
Hope this gives you some ideas for a solution.
LiPeng (Denmark)
Answer Verified By: Stewart Souten
Stewart
In my humble opinion this is what you have (I'm not sure about the start - too little data). Two compound curves (large and smaller radius with intermediate transition) then reverse curve with two transitions to large radius. I agree with LiPeng data is rather poor. Are there any sets/turnouts in this section?
Neville du Plessis
Rail Design
LiPeng
Thank you for your input this really helped. I did check if the decimal points had been imported and they did. Your recommendation about importing only the second/third point really help clear up the noise in the data.
Thanks
Hi Neville
Thanks for your insight.
I did manage to come up with the same solution but it was nice to know that someone lese agreed. What helped was removing some of the points which cleared up the curvature diagram.
Much appreciated.
Perhaps, here is a need for a clarification: If all survey points are measured exactly the same way, it's difficult to justify which points belong to "noise". The method reading fewer points does help giving you a clear curvature diagram (due to the way the diagram is build up), and easier for you to decide the correct elements and lengthes. When you've created an initial alignment, it will be a good idea to perform a multi-elements regression (incl. the transition curves), you can here added more survey points. In this way you'll get the best over all regression and truer to the survey, regardless the "mean error" etc. become larger than initially.