Diverging Diamond Interchanges

[tradephoric]

Florida is planning a DDI at I-4 and Martin Luther King Jr. Blvd:


The state hasn't abandoned the Parblo B4 though.  This Parclo B4 at I-75 & Sheridan Street was just recently reconstructed: 


It looks like a Parclo B4 would fit in the footprint of the proposed DDI at I-4 & MLK Blvd.   At this point I'm just glad that Florida didn't replace the Parclo B4 at I-75 & Sheridan Street with a DDI. 

[tradephoric]

Florida likes to go big with their DDIs.  Here's the first one they built along University Parkway:

[jakeroot]

All that room....they could at least put in a shared use path. And that applies for all three diagrams above this post. I love big interchanges as much as the next AARoads user, but they are just not friendly to the neighborhood at all. And agencies really don't seem to do enough to rectify this situation.

What really bothers me is when states put in a DDI to, maybe, give the appearance that they are ped-friendly, but they are perhaps less friendly than even a large Parclo interchange.

I would really like states to start installing grade-separated shared use paths through interchanges. In the Parclo interchange above (75 @ Sheridan), a wide (20ft+) shared use path could have been built below the slip ramps (and then above the freeway), making the two sides of the freeway much better connected. The sidewalks that have been used look to be painfully narrow. I can't imagine the noise!

In that Sheridan interchange, I also noticed what is a big no-no in my area: crosswalks divided into two crossing sections. The intersection where the loops meet Sheridan have two pedestrian heads installed for each slip ramp crossing. This is fine if they are synced together (as in, no need to push a second button mid-way across), but if they are divided for purposes of improving minimum signal phasing time? Not cool.

[Rothman]

O.o

Grade-separated shared-use paths?  Through intersections?

Totally infeasible due to the benefits gained from the immense costs.

[jakeroot]

O.o

Grade-separated shared-use paths?  Through intersections?

Totally infeasible due to the benefits gained from the immense costs.

*Interchanges. I don't expect it at every intersection.

"Immense costs"? It's not much compared to the rest of the interchange.

Note this interchange below. I don't know where it's from, but it has a third grade for peds. Brilliant!

[Rothman]

Even through interchanges it is an incredible expense.  Hope away, but it will never be standard practice.

[jakeroot]

Even through interchanges it is an incredible expense.  Hope away, but it will never be standard practice.

I don't expect them to be used in the middle of nowhere, but any time you have a freeway running through an urban or suburban area, agencies need to do their part to keep all network types tied together. This means cars, transit, bikes, and pedestrians. All are accounted for in all modern interchanges, but the first two (primarily the first) clearly gets top priority. Why? Because we do so little to accommodate for pedestrians, that nobody wants to walk anywhere. That leads to less pedestrians, more drivers, etc. If that's our end goal, fine. Just build the minimum requirements. But if we want less cars, we need to make it less of a pain to get around on foot. Florida sucks to walk around in during the summer months, but the rest of the year, it's quite pleasant!

By the way, it's becoming more common in the Netherlands, even in suburban areas.

[tradephoric]

What really bothers me is when states put in a DDI to, maybe, give the appearance that they are ped-friendly, but they are perhaps less friendly than even a large Parclo interchange.

Yeah, pedestrians along the University Parkway DDI have to cross 14 lanes of traffic — and 10 of those lanes cross the main-arterial.  The pedestrian exposure is worse at the University Parkway DDI compared to the Parclo example (pedestrians at the Sheridan Parclo only cross 11 lanes of traffic — and none of those lanes cross the main-arterial).   There does seem to be a disconnect to what engineers view as "pedestrian friendly" .  Charles Marohn, who runs Strong Towns, posted a tour of the "pedestrian friendly"  diverging diamond rebuttal video which is pretty good.   The original video featured an overly enthusiastic engineer talking about a newly constructed "pedestrian friendly"  DDI. 

https://www.youtube.com/watch?v=zWG49xlZ_eQ

[jakeroot]

What really bothers me is when states put in a DDI to, maybe, give the appearance that they are ped-friendly, but they are perhaps less friendly than even a large Parclo interchange.

Yeah, pedestrians along the University Parkway DDI have to cross 14 lanes of traffic — and 10 of those lanes cross the main-arterial.  The pedestrian exposure is worse at the University Parkway DDI compared to the Parclo example (pedestrians at the Sheridan Parclo only cross 11 lanes of traffic — and none of those lanes cross the main-arterial).   There does seem to be a disconnect to what engineers view as "pedestrian friendly" .  Charles Marohn, who runs Strong Towns, posted a tour of the "pedestrian friendly"  diverging diamond rebuttal video which is pretty good.   The original video featured an overly enthusiastic engineer talking about a newly constructed "pedestrian friendly"  DDI. 

https://www.youtube.com/watch?v=zWG49xlZ_eQ

He definitely summed up a lot of my issues with many DDIs. Chiefly, the requirement to cross into the middle of the interchange to get across. AFAIK, this is because old bridges are utilized in the conversion, and the crossover requires the roadways be pushed to edge of the bridge, displacing the crosswalk to the middle.

Of course, that's an excuse. A properly designed interchange should consider all users equally, even if certain users aren't in the majority. You can design a high capacity interchange with excellent pedestrian and cycling facilities (the roundabout above is just one example), but most engineers seem to like using the whole budget to improve things for cars, leaving barely any of the budget for other modes of access.

Folded diamonds are actually really nice for pedestrians, because they limit the number of crossings to two T-intersections. That may be why they are so popular in the Netherlands. You could even add in an extra slip ramp for traffic, so that the loops would only be used for one direction of traffic, and you'd still only have two roadways to cross. Hell of a lot better than the DDI.

The only proper DDI design, IMO, are the American Fork DDI's in Utah. New bridges were built, and pedestrian access was kept to the exterior of the roadways. Four lanes to cross at the lower DDI, six at the upper. At least double that if the sidewalk were in the inside. I am staggered that the University Pkwy DDI wasn't designed this way, since it was built from the ground-up.

[johndoe]

The only proper DDI design, IMO, are the American Fork DDI's in Utah. New bridges were built, and pedestrian access was kept to the exterior of the roadways. Four lanes to cross at the lower DDI, six at the upper. At least double that if the sidewalk were in the inside.

Nah.  Most built now have the pedestrians in the middle, and for good reason.  Here's an excerpt from Utah's DDI guide:
"UDOT prefers that pedestrians cross DDIs using an island
located in the middle of the interchange. While the first
of Utah's DDIs crossed pedestrians on the outside of the
interchange, this placed them in the position of crossing
free movements at locations with limited visibility. Recent
designs have shifted pedestrians to the median. This improves
lines of sight for both pedestrians and vehicles to
and from the crossings and allows pedestrians to cross
through traffic at a signalized location to clarify pedestrian/vehicle
rights of way. Left turns to the entrance ramps
can also run freely without conflict with pedestrian crossings."

While you guys are right that pedestrians cross more lanes, they're slower lanes coming to a signal.  Even in the case of University Pkwy with ~100' to cross, you'll get plenty of time since the crossing is tied to one of the "through" phases (which will get enough green time).  It's not like other designs that would be thrown out of coordination for many cycles if a pedestrian crossed the road.

[jakeroot]

So signalize the turns to and from the ramps. Dumb reasoning.

Alternatively, design the crosswalks so that pedestrians are more visible.

While you guys are right that pedestrians cross more lanes, they're slower lanes coming to a signal.  Even in the case of University Pkwy with ~100' to cross, you'll get plenty of time since the crossing is tied to one of the "through" phases (which will get enough green time).  It's not like other designs that would be thrown out of coordination for many cycles if a pedestrian crossed the road.

You're not getting it. All modern interchanges assist pedestrians getting to and from each side. That's not the issue here. The consideration needs to be: do pedestrians want to cross here? What can we do to make this area less frightening? Not just "does this work?". The American Fork DDIs allow pedestrians to interact with the roadways as little as possible, making them much more inviting for pedestrians.

FWIW, it's because of signal timing requirements that I prefer for crossings to be grade separated where feasible. The 75 & Sheridan interchange above is an example of a node where the roadway plans could have been scaled back slightly to allow for an exclusive, grade-separated pedestrian path. Remember why everyone drives: because no one wants to walk. Cars are their own worst enemy.

[johndoe]

So signalize the turns to and from the ramps. Dumb reasoning.

Alternatively, design the crosswalks so that pedestrians are more visible.

You're not getting it. All modern interchanges assist pedestrians getting to and from each side. That's not the issue here. The consideration needs to be: do pedestrians want to cross here? What can we do to make this area less frightening? Not just "does this work?". The American Fork DDIs allow pedestrians to interact with the roadways as little as possible, making them much more inviting for pedestrians.

FWIW, it's because of signal timing requirements that I prefer for crossings to be grade separated where feasible. The 75 & Sheridan interchange above is an example of a node where the roadway plans could have been scaled back slightly to allow for an exclusive, grade-separated pedestrian path. Remember why everyone drives: because no one wants to walk. Cars are their own worst enemy.

Haha I don't want to be a jerk, but I do this work as my career, I think I "get  it".  I love this forum because I'm a roadgeek too ... but it's a lot different to "design things on paper" with a general interest than it is to lay in vertical and horizontal curves, design the signing/striping, and have your professional signature on a plan set.  One of us has done those things...I think my opinion is at least worth hearing.  If nothing else people can know why engineers do some of the things we do.

Regarding grade separations... Sure that'd be nice and if there were enough pedestrians at a particular spot (or money) you'd probably see it more often.  Just as Rothman said, don't hold your breath.  Bridges are expensive, and most decision-makers care about the majority of users (and interchanges could have at least 500x as many cars as peds).  I don't always agree, but that's the world we live in.  Also keep in mind crossing an interchange is probably no worse than the signals all the way down the corridor.  You'd probably need bridges at those, too.  Also keep in mind that really long pedestrian structures have a reputation (haven't seen data) for crime.  And have fun with the ADA requirements.

On convincing people to walk, absolutely that FL DDI looks daunting.  But take a look at both signals adjacent to it.  Long crossings and a lack of refuge islands mean the pedestrian simultaneously crosses vehicles going different directions.  With the DDI you cross one direction of traffic at a time.  You'll get plenty of time without messing up signal timing for cars.  It also gives you the chance to cross from north to south side.  That's something many interchanges don't do.

As far as designing the crossings on outside of DDI for visibility... Good luck.  Especially if the freeway goes over (many bridge piers and shadows hiding peds).  Again this is why I posted the Utah guidance... I think they know what they're talking about 

As far as signalizing the lefts onto the freeway...that completely removes one of the benefits of the DDI (free lefts) and adds the complexity of an additional signal.  Avoiding this is certainly not "dumb reasoning".  It's also tough to place the signal heads in "crossroad under" situations.  This has been built though, I think Missouri has a couple like that.

Again I don't want to be condescending, but hopefully this describes why peds in middle is (and will be) more common.

[inkyatari]

Speaking of pedestrian access to a DDI, here's the preferred alignment document for the Weber Rd. I-55 interchange...

http://www.idot.illinois.gov/Assets/uploads/IDOT-Projects/District-1/I55-at-Weber-Rd-Study/files/3_Preferred%20Alternative.pdf

http://www.idot.illinois.gov/Assets/uploads/IDOT-Projects/District-1/I55-at-Weber-Rd-Study/files/6_Aerial%20Interchange.pdf

This is going to be a huge project as it includes a couple miles of Weber Rd to the south, and portions of Normantown Rd.

Full project site:

http://www.idot.illinois.gov/projects/i55-at-weber-road

[tradephoric]

From my understanding DDI's often don't coordinate to adjacent signals because they need to run shorter cycles to prevent on/off ramp traffic from queuing up between DDI signals.  But in the Florida example, the massive capacity of that University Parkway DDI has me questioning if they are attempting to coordinate with the adjacent signals.  The DDI at University Parkway isn't the first DDI built along a 6-lane arterial yet it is by far the highest capacity DDI built in America to date.   IMHO, if a 6-lane arterial needs to widen out to 12-lanes at the interchange to achieve an acceptable level of service, there is something wrong with that design (sure, it may move vehicles through it... maybe even efficiently... but football fields worth of asphalt shouldn't be needed to achieve that goal). 

As far as signalizing the lefts onto the freeway...that completely removes one of the benefits of the DDI (free lefts) and adds the complexity of an additional signal.  Avoiding this is certainly not "dumb reasoning".  It's also tough to place the signal heads in "crossroad under" situations.  This has been built though, I think Missouri has a couple like that.

I agree it's not ideal signalizing the lefts onto the freeway.  But the alternative is to direct pedestrians to the center median and force them to cross main-arterial traffic (in the case of the University Parkway pedestrians must cross 10-lanes of arterial traffic).  The fact that the typical DDI design favors moving traffic efficiently onto the freeway over reducing pedestrian exposure kind of proves that the DDI is not really designed to be pedestrian friendly (which isn't completely surprising considering how few pedestrians are going to cross a DDI to begin with).  I'm not expecting a suburban freeway interchange to be pedestrian friendly, but when you have that video of an overly enthusiastic engineer talking about how "pedestrian friendly"  the DDI is... there just seems to be a disconnect there.  The tick boxes may have been checked to adhere to ADA standards, but many DDI's can still be miserable places to walk (even worse than other suburban freeway interchanges).  And even after pedestrians are redirected to the center median and forced to cross main arterial traffic, you still have unsignalized crosswalks where drivers are turning right onto the freeway (and this seems to be the norm at most DDI's).

[jakeroot]

Regarding grade separations... Sure that'd be nice and if there were enough pedestrians at a particular spot (or money) you'd probably see it more often.  Just as Rothman said, don't hold your breath.  Bridges are expensive, and most decision-makers care about the majority of users (and interchanges could have at least 500x as many cars as peds).  I don't always agree, but that's the world we live in.  Also keep in mind crossing an interchange is probably no worse than the signals all the way down the corridor.  You'd probably need bridges at those, too.  Also keep in mind that really long pedestrian structures have a reputation (haven't seen data) for crime.  And have fun with the ADA requirements.

It's a fact of life because we've designed our roads to be the way they are, and facilitate movements the way they do. Today's engineers are brilliant. There's no denying that. They come up with some very clever stuff, but it's all invariably aimed at improving traffic flow, and then assisting other users through the junction. I just wish that engineers designed infrastructure with priority from most vulnerable to least vulnerable. You can still have a high capacity junction that is enjoyable for pedestrians to use; the junction just needs to be redesigned from a different perspective. The Dutch are very good at this sort of thing. I'll be damned if the constant protected signals wouldn't drive me crazy as a driver, but I can't fault them -- it works, because it's well designed (unlike protected turns here, which seem to be placed nonsensically).

I understand that "vulnerable" is one of those words that make engineers cringe, but it's a term that's here to stay because it promotes the safe design of an area for users without crumple zones.

All pedestrian corridors can have crime issues, if they aren't well-trafficked. Which is why we need to do more to improve overall pedestrian infrastructure. If we get more people to use the ped paths, there's a far smaller chance that someone will be the victim of a crime. The alternative is to get rid of all pedestrian infrastructure, reducing the chance of ped-on-ped crime to zero since you're not allowed to walk. Though that's a little ridiculous.

On convincing people to walk, absolutely that FL DDI looks daunting.  But take a look at both signals adjacent to it.  Long crossings and a lack of refuge islands mean the pedestrian simultaneously crosses vehicles going different directions.  With the DDI you cross one direction of traffic at a time.  You'll get plenty of time without messing up signal timing for cars.  It also gives you the chance to cross from north to south side.  That's something many interchanges don't do.

So, the DDI above is acceptable because it's marginally better than the garbage intersections just outside of its perimeter? Good lord, we need to reconsider what we consider "good design".

I can't immediately think of ways to improve the intersections themselves (a lot less room to work with), but we can start at junctions with lots of ROW to at least prove what's possible.

As far as designing the crossings on outside of DDI for visibility... Good luck.  Especially if the freeway goes over (many bridge piers and shadows hiding peds).  Again this is why I posted the Utah guidance... I think they know what they're talking about

I don't doubt that they know what they're talking about, but I get the feeling that their infrastructure (like so many agencies out there) is designed for cars first, everyone else second. So the question they're asking becomes, "how can we best help peds through this intersection?" In my opinion, that's the wrong question. What might a better question be? My professors would have to help me there, but one might be, "first, how can we best facilitate vulnerable movements across this freeway?"

For underpasses, the key is for them to be wide (maximum light), and well lit. Thusly, the support piers will have to be designed with a wide ROW in mind.

As far as signalizing the lefts onto the freeway...that completely removes one of the benefits of the DDI (free lefts) and adds the complexity of an additional signal.  Avoiding this is certainly not "dumb reasoning".  It's also tough to place the signal heads in "crossroad under" situations.  This has been built though, I think Missouri has a couple like that.

To be honest, I'm not particularly worried about a 15-second pedestrian interval at a left turn. It's a minor annoyance for drivers, meanwhile massively improving things for pedestrians and cyclists. If need-be, the on-ramp signals can be tied to the off-ramp signals. The on-ramp would be red at the beginning of the off-ramp phase, and then go green after the crossing period finished. Since off-ramp traffic is not getting back on, the demand for the left turn would be minimal during this time. It might be wise to design DDI's with less option-lane left turns if this were to become more popular.

IMHO, if a 6-lane arterial needs to widen out to 12-lanes at the interchange to achieve an acceptable level of service, there is something wrong with that design (sure, it may move vehicles through it... maybe even efficiently... but football fields worth of asphalt shouldn't be needed to achieve that goal). 

I agree 100%. So, in summary, the DDI is a terrible experience for pedestrians, and needs 12+ lanes of traffic to work correctly. I wouldn't call that a pass.

[jakeroot]

Coming back to this after a day and a half, I realize that I might have gone off the rails just a bit. I am still not a fan of central crossings at DDI's, and I think it's wise to consider designing corridors that pedestrians would want to use, not just what they can use. But, given how much more geometric restriction there is on roadway design, agencies really have no choice but to design the roads first, and then work in a pedestrian corridor from there. It should then be up to the agency to properly redesign the roadway to best accommodate the pedestrian corridor.

For example, re-aligning slip lanes so that they point at the roadway closer to a 70- or 80-degree angle, to improve pedestrian visibility. Or, installing a signalized crossing where a yield crossing currently exists. Small changes that don't necessarily hurt vehicle access, but make areas far more inviting to pedestrians.

After a week of solid reading about public spaces, it's amazing how picky humans are about our space. In a car, we seem to be a bit more forgiving. But on foot, we notice everything, and very quickly realize an area's problems.

[NoGoodNamesAvailable]

But, given how much more geometric restriction there is on roadway design, agencies really have no choice but to design the roads first, and then work in a pedestrian corridor from there.

I think you're getting into the heart of the issue there. Motorist level of service trumps all other factors in street design in the US, often at the expense of delaying other road users by minutes or making certain trips effectively impossible without a car.

If you listened to some traffic engineers (especially the oldtimers), you'd think the earth would implode if we reduced traffic lanes on city streets from 11'—12' to 9'—10', or prohibited RTOR by default in urban areas.

But basic experience can confirm how ridiculous that is–the Bronx River Parkway, for example, frequently squeezes down to 8 (!) feet per lane in some places, and passenger cars manage to rip 70 down that road. Have you ever looked at a street and gone, "damn, these 10-foot lanes are really delaying the bus and truck traffic on this road"? Of course not, because buses and trucks manage 10-foot lanes fine on plenty of city streets.

NYC and European cities ban RTOR by default and there aren't protests in the street about it like engineers act like there would be, because people realize the delay to motorists is in exchange for significantly lower pedestrian crash rates.

There's no reason why our most pressing concern should be saving seconds for motorists when the same decisions made to that end cause minutes of inconvenience to other users.

[SkyPesos]

This seems like the general thread for DDIs from a quick search, but if there's a better thread, let me know.

Found a DDI (that's supposedly temporary) in the Netherlands, at A44 and N206 at Leiden. The configuration itself is interesting, with two right turn lanes from WB to NB, one from a frontage road and one from the through road. It seems like the final (and permanent) configuration for this interchange is a B4 parclo, from recent Google Maps markings.

[tolbs17]

Is a diverging diamond interchange better than a dumbbell or dogbone?

[ran4sh]

Is a diverging diamond interchange better than a dumbbell or dogbone?

Yes, but I don't think they have similar use cases. In most of America, drivers can't use multi-lane roundabouts correctly, so a dumbbell or dogbone would only be helpful for 2-lane (1 lane each direction) roads. DDIs are mostly for multi-lane roads.

[jeffandnicole]

Is a diverging diamond interchange better than a dumbbell or dogbone?

Totally dependent on the situation. DDIs are good where a lot of traffic enters or exits the highway. If the traffic is mainly heavy thru traffic, DDIs can cause significant congestion.

[MCRoads]

Not sure if it has already been mentioned, but Oklahoma's first DDI is open!

[froggie]

O.o

Grade-separated shared-use paths?  Through intersections?

Totally infeasible due to the benefits gained from the immense costs.

Responding a few years late here, but MnDOT says hi...

Does it work in every situation?  No.  But it also isn't the budget buster you claim it to be.  In this case, all it took was a little extra grading and a couple of oversized culverts.

[tolbs17]

I just know that there is not a single diverging diamond interchange in New Jersey. And none are considered there.

They are pretty rare in Ohio, There's a decent amount in Missouri and a lot more being considered here in North Carolina.

[Scott5114]

The DDI seems to me to be a solution in search of a problem. Having
two "cross-over" locations is just asking for trouble...you're bound
to get Grandpa Artie all confused and driving into incoming traffic.
I'm not sure if the additional thought required of drivers is too much
for the safety improvement of not having to turn left. It is inferior
to the SPUI because it could still require waits at two signals,
whereas the SPUI requires only one (though left-turning traffic from
the ramps will only have to have one, they will have a yield to
contend with, which can potentially be as long of a wait as a
stoplight if traffic is heavy). About the only good thing about the
DDI is that it's cheaper than a SPUI.

Nineteen-year-old me was an idiot.