Synchronized Interchange (w/contraflow off-ramps)

[tradephoric]

This interchange design only requires one bridge deck, doesn't have any loop ramps, and the corridor is capable of good signal progression in both directions (ie. no traffic signal stops both directions of through traffic along the corridor).  The contraflow off-ramp lanes would increase the required length of the bridge deck, but there are plenty of interchange designs with lots of lanes directly under the bridge deck that people seem to praise (IE. the recently constructed DDI at University Parkway has 12 lanes under the bridge deck).). 

https://www.youtube.com/watch?v=ihpgxEzNtz0

[tradephoric]

I edited the SYNCHRO to create a "tight diamond".  The geometry of the interchange is such that drivers exiting the freeway turning left onto the arterial will nearly always hit a green light at the upstream traffic signal.  Of course it's a little unorthodox to have a turnaround crossing two directions of traffic, and this is where this design could get a little confusing.

https://www.youtube.com/watch?v=9lT_vGJWFjs

[tradephoric]

Here is a design where crossover traffic doesn't cross two directions of traffic.  The downside is you are more likely to get poor signal progression for traffic exiting the freeway turning left onto the arterial (but no different than standard DDI operation where left-turning exiting freeway traffic potential stops at two interchange traffic signals).  The upside is traffic can more easily merge onto the arterial in the lane they want to be (ie. easier to enter that gas station drive that's just past the interchange). 

https://www.youtube.com/watch?v=DQdcdNz_n-0

[kphoger]

requires one bridge deck

From the looks of it, I imagine that would be a massive bridge deck, such that it might actually end up being constructed as two decks anyway.

[tradephoric]

requires one bridge deck

From the looks of it, I imagine that would be a massive bridge deck, such that it might actually end up being constructed as two decks anyway.

The University Parkway DDI has a 280 foot long bridge deck with 12 lanes of arterial traffic and a 60 foot wide pedestrian central island underneath it.  That 60 foot pedestrian central island alone could fit the 4 contraflow lanes in the models above and then you are just left with 6 lanes of arterial traffic as opposed to 12 lanes in that Florida DDI.  Also, these SYNCRHO models are more conceptual than anything, so the contraflow lanes are spread out much farther apart than they would really need to be. 

[jakeroot]

Good lord, the amount of ROW being taken up on the first and last example is insane. The last example has to have at least 200 feet of ROW curb-to-curb, never mind the legal minimum 10 feet of sidewalk (5 on either side).

[kphoger]

The University Parkway DDI ... 12 lanes of arterial traffic

Yes, good point.

[tradephoric]

Here's a Synchronized Interchange where traffic entering the freeway uses the contraflow movements as opposed to the exiting freeway traffic.  Of course the traffic signals controlling the turnaround traffic limits the effectiveness of the contraflow lanes, as traffic wishing to enter those contraflow lanes to get on the freeway may be stuck at a red light just before being able to enter them.  It would be a better design if somehow those turnaround lights didn't need to be signalized (and maybe they wouldn't need to be signalized if traffic exiting the freeway was light.. but often not the case!). 

https://youtu.be/W01GG3l58vg

[jakeroot]

The University Parkway DDI ... 12 lanes of arterial traffic

Yes, good point.

More lanes ≠ greater capacity.

[tradephoric]

Here's a Synchronized Interchange where traffic entering the freeway uses the contraflow movements as opposed to the exiting freeway traffic.  Of course the traffic signals controlling the turnaround traffic limits the effectiveness of the contraflow lanes, as traffic wishing to enter those contraflow lanes to get on the freeway may be stuck at a red light just before being able to enter them.  It would be a better design if somehow those turnaround lights didn't need to be signalized (and maybe they wouldn't need to be signalized if traffic exiting the freeway was light.. but often not the case!). 

https://youtu.be/W01GG3l58vg

Maybe this would be a better variation of the Synchronized Interchange...

https://www.youtube.com/watch?v=HPHyz4qI7k4

[roadfro]

Here's a Synchronized Interchange where traffic entering the freeway uses the contraflow movements as opposed to the exiting freeway traffic.  Of course the traffic signals controlling the turnaround traffic limits the effectiveness of the contraflow lanes, as traffic wishing to enter those contraflow lanes to get on the freeway may be stuck at a red light just before being able to enter them.  It would be a better design if somehow those turnaround lights didn't need to be signalized (and maybe they wouldn't need to be signalized if traffic exiting the freeway was light.. but often not the case!). 

https://youtu.be/W01GG3l58vg

Maybe this would be a better variation of the Synchronized Interchange...

https://www.youtube.com/watch?v=HPHyz4qI7k4

You might need to check your origin-destination parameters or turn restrictions in these simulations, as there are various elements not operating realistically.

For example, in the first video, I observed several westbound arterial vehicles that used the turnaround, then proceeded to clear eastbound (effectively completing a U-turn) or use the crossover to enter the freeway northbound (essentially completing an inside clockwise loop through multiple signals instead of just turning right on the freeway). In the second video, take a look at every vehicle colored dark yellow/brown that exits the freeway, as they immediately cross to the left turn to re-enter the freeway... It's been a while since I've used Synchro/SimTraffic, but I seem to recall that there are some parameters in there somewhere that can help control the simulated vehicles so that they don't make unrealistic movements.

I'm not sure if you're looking at the analysis numbers closely or not, but this is bound to affect the analytics.

[hotdogPi]

Another option: two signals very close together. For example, if they're 5 seconds apart, and the green time is 30 seconds, only ⅙ of cars will get stopped by the second light (less if previous signals are timed correctly). This would require much less ROW.