G'Day from Canada,

I have a bend node located 60m upstream of an intersection, at which point a single through lane splits into two separate lanes. I am trying to measure queue length and delay, and have two questions with the resulting SimTraffic Report...

Q1: do the queue lengths in the B## column represent queue lengths from the intersection, or just from the bend node?

Q2: it appears that the corresponding delay for these vehicles only is considered for the 60 segment. How do I get the "complete" delay for vehicles on that apprach (i.e. including the queues past the bend node)??

I'd appreciate any help offered!

MArk

assume system node numbers, north is up ^^
4
1236
5

60m between 2 and 3

In your Simtraffic report, you can select "List Exits and Bends"
This will report the delay associated with node 2, and the EB (as example above) will have delays and queues recorded that are assiciated with backups from intersection at node 3.


Word of caution regarding short nodes like this at intersections.
You'll get an average and maximum queue reported from the simulation, which will appear in your static graphic images.
However, its very possible that the queue does not back up to the second link every cycle. In this case, the displayed queue calculations will not be an accurate representation of the signal.

The situation I had was a Signal at a Tight diamond interchange. The local road had several driveways for large apartment complexes with volumes and operations I didnt' want to disregard. As a result, I had about 5 links leading up to the signal. On the Static graphic report I got an average queue all the way back on the last link, but I saw from running the simulation that it only became that long late in the simulated hour. So it follows that the queue reported is the average and maximum queue that appears on that link, and to find the queue as a function of the signal I would have to resort to some manual calculations.
I wound up printing the drawing to scale, and watching the simulation and marking on the printout where physically the back of queue was each cycle. Then I could measure based on scale and come up with a more reasonable average and 95th queue. Sounds pretty primitive, and it wasn't precisely accurate, but it allowed me to proceed...