Are diverging diamonds a fad?

[tradephoric]

Here's a diverging diamond interchange where traffic seems to be needlessly backing up through the adjacent intersection.  Would a SPUI or even roundabouts be more effective than existing diverging diamonds?  I don't see many advantages that diverging diamond interchanges have over other designs.


https://maps.google.com/maps?ll=37.25132,-93.31049&z=17&t=h&output=classic&dg=brw

[thenetwork]

IMHO, DDI's are most effective if the majority of the surface road traffic changes direction onto or off of the freeway. If the majority of surface road traffic is only thru traffic and does not turn onto the freeway, then the DDI is not the best solution as the thru traffic may commonly have longer waits to clear the interchange vs. a traditional diamond interchange.  At that point, I would opt for a SPUI.

[vdeane]

A DDI has two and only light phases.  A SPUI needs at least three, and I've seen the local one have as many as six at times, seemingly at random.  Wouldn't that make it take longer to get through a SPUI?

[dfwmapper]

SPUIs require expensive new bridgework. A standard diamond can be converted into a DDI with no changes to the existing bridge, just some concrete work on either side of it. Major capacity boost for a small fraction of what any other solution would cost. Very useful in a state like Missouri that has some of the worst highway funding problems in the nation.

[Thing 342]

Maybe it's my lack of a background in civil engineering, but I fail to see how the problem with stacked thru traffic at the demonstrated intersection would be improved using a SPUI or normal diamond.

[NE2]

Maybe it's my lack of a background in civil engineering, but I fail to see how the problem with stacked thru traffic at the demonstrated intersection would be improved using a SPUI or normal diamond.

In most interchange designs, thru traffic on the surface road can proceed simultaneously in both directions. But the DDI prioritizes turning traffic over thru traffic, meaning that only half the total time can be assigned to each direction.

[DaBigE]

DDIs are also easier for larger trucks (read: OSOW) to navigate. There have been a couple DDIs selected over roundabouts in Wisconsin with OSOW accommodations being one of the biggest reasons.

[tradephoric]

Here's a roundabout interchange outside of Milwaukee on I-43.  This appears to have similar if not less ROW requirements than the DDI example from Missouri.  Overall, would two roundabouts be cheaper than a DDI (since there are no traffic signals to maintain, same ROW requirements)?

https://www.google.com/maps/@42.9327037,-88.160052,577m/data=!3m1!1e3

[dfwtbear]

What about doing something like this interchange in Texas? Thru traffic on the street tunnels under the access roads so the only traffic at the lights are the access road and turning traffic from the street.

https://www.google.com/maps/place/Richardson,+TX/@32.940039,-96.745678,346m/data=!3m1!1e3!4m2!3m1!1s0x864c1ee979bea767:0x2cdb29c046270495

[mhh]

What about doing something like this interchange in Texas? Thru traffic on the street tunnels under the access roads so the only traffic at the lights are the access road and turning traffic from the street.

https://www.google.com/maps/place/Richardson,+TX/@32.940039,-96.745678,346m/data=!3m1!1e3!4m2!3m1!1s0x864c1ee979bea767:0x2cdb29c046270495

Interesting. I had assumed that this similar interchange in Royal Oak, Michigan was unique:

https://www.google.com/maps/place/Royal+Oak,+MI/@42.4744205,-83.1454238,587m/data=!3m1!1e3!4m2!3m1!1s0x8824c5f96ccd2bf7:0x70fd341efca3310

[kphoger]

The above link just searches for the town of Richardson for me.

[Big John]

What about doing something like this interchange in Texas? Thru traffic on the street tunnels under the access roads so the only traffic at the lights are the access road and turning traffic from the street.

https://www.google.com/maps/place/Richardson,+TX/@32.940039,-96.745678,346m/data=!3m1!1e3!4m2!3m1!1s0x864c1ee979bea767:0x2cdb29c046270495

The Texas U-turn there.  Helps traffic at intersections with one-way access roads as traffic wanting to turn around avoid 2 sets of traffic signals.  Without knowing traffic patterns there it may be the optimal solution to a busy intersection with access roads.

[Ned Weasel]

What about doing something like this interchange in Texas? Thru traffic on the street tunnels under the access roads so the only traffic at the lights are the access road and turning traffic from the street.

https://www.google.com/maps/place/Richardson,+TX/@32.940039,-96.745678,346m/data=!3m1!1e3!4m2!3m1!1s0x864c1ee979bea767:0x2cdb29c046270495

Interesting. I had assumed that this similar interchange in Royal Oak, Michigan was unique:

https://www.google.com/maps/place/Royal+Oak,+MI/@42.4744205,-83.1454238,587m/data=!3m1!1e3!4m2!3m1!1s0x8824c5f96ccd2bf7:0x70fd341efca3310

Those are technically volleyballs.  Sure, they help thru traffic flow on the cross street, but they're probably significantly more expensive to build than a DDI, a roundabout interchange, or a SPUI, since you're looking at a three-level interchange rather than a two-level interchange.

[dfwmapper]

IIRC, throughput on roundabouts is much lower than on a DDI. That Wisconsin example isn't really a similar situation, since that's basically a 2 lane rural road that happens to widen to 4 lanes for a mile surrounding the freeway, and travels through empty land and low-density residential areas, while the Missouri one has multiple large retailers plus the fairgrounds nearby, not to mention being the primary route between Springfield and the KC metro. Suggesting what was done at the US 75/Spring Valley interchange in Richardson is just silly, tunneling under an existing road without completely shutting it down and doing a cut and cover costs a fortune.

[tradephoric]

Looking through aerials on Google Earth of roughly 150 multi-lane interchange roundabouts and 30 DDI's,  none of the interchange roundabouts seem to be experiencing any significant backups while several of the DDI's have significant backups on at least one leg (and in some cases are backing up through the adjacent intersection).  Here's example of a higher capacity interchange roundabout that is also on I-43 (it's a 3-lane roundabout).  Again, no significant backups. 

https://maps.google.com/maps?ll=42.951318,-88.108882&spn=0.006471,0.006813&t=h&z=17

DDI's are simple two-phase traffic signals but the surrounding signals of a DDI are often complicated 4-phase signals which requires longer cycle lengths.  Oftentimes, a DDI doesn't coordinate with a signal that might only be 600 feet away (such as the original example posted).  When you have green to red coordination, you will inevitably see a queue of cars (that just got a green light at the signal 600 feet away) backing up along the DDI.

[tradephoric]

[DaBigE]

Looking through aerials on Google Earth of roughly 150 multi-lane interchange roundabouts and 30 DDI's,  none of the interchange roundabouts seem to be experiencing any significant backups while several of the DDI's have significant backups on at least one leg (and in some cases are backing up through the adjacent intersection).  Here's example of a higher capacity interchange roundabout that is also on I-43 (it's a 3-lane roundabout).  Again, no significant backups. 

https://maps.google.com/maps?ll=42.951318,-88.108882&spn=0.006471,0.006813&t=h&z=17

DDI's are simple two-phase traffic signals but the surrounding signals of a DDI are often complicated 4-phase signals which requires longer cycle lengths.  Oftentimes, a DDI doesn't coordinate with a signal that might only be 600 feet away (such as the original example posted).  When you have green to red coordination, you will inevitably see a queue of cars (that just got a green light at the signal 600 feet away) backing up along the DDI.

Yes, DDIs are a pain to coordinate as part of a signalized corridor. While the roundabout you linked doesn't usually have any significant backups, it has had a higher than normal crash problem. That pair of roundabouts have been a thorn in the side of the WisDOT SE Region since they opened. Those two gave the anti-roundabout crowd a lot of fuel and are part of the reason why WisDOT has put a moratorium on any more three-lane roundabouts in Wisconsin.

[tradephoric]

Looking through aerials on Google Earth of roughly 150 multi-lane interchange roundabouts and 30 DDI's,  none of the interchange roundabouts seem to be experiencing any significant backups while several of the DDI's have significant backups on at least one leg (and in some cases are backing up through the adjacent intersection).  Here's example of a higher capacity interchange roundabout that is also on I-43 (it's a 3-lane roundabout).  Again, no significant backups. 

https://maps.google.com/maps?ll=42.951318,-88.108882&spn=0.006471,0.006813&t=h&z=17

DDI's are simple two-phase traffic signals but the surrounding signals of a DDI are often complicated 4-phase signals which requires longer cycle lengths.  Oftentimes, a DDI doesn't coordinate with a signal that might only be 600 feet away (such as the original example posted).  When you have green to red coordination, you will inevitably see a queue of cars (that just got a green light at the signal 600 feet away) backing up along the DDI.

Yes, DDIs are a pain to coordinate as part of a signalized corridor. While the roundabout you linked doesn't usually have any significant backups, it has had a higher than normal crash problem. That pair of roundabouts have been a thorn in the side of the WisDOT SE Region since they opened. Those two gave the anti-roundabout crowd a lot of fuel and are part of the reason why WisDOT has put a moratorium on any more three-lane roundabouts in Wisconsin.

Good info DaBigE.  That's the first i've heard about that moratorium.  A 3-lane roundabout in SE Michigan was recently restriped as a 2-lane roundabouts.  Of course, the roundabout was originally designed with the idea that the street it services would be widened to a 4-lane boulevard, but the street is still only a two-lane road.  Therefore, the reduction in lanes hasn't had an adverse effect on throughput.

[johndoe]

I need to get on a soap box for just a minute... I think the name "DDI" has already caught on and it's too late to convince people to call it something else, but "double crossover diamond" makes so much more sense.  What about this interchange diverges more than other service interchanges?  There are actually fewer diverging conflicts in a DCD than a normal diamond... I vote we all start going with "DCD" 

Anyway, good discussion here.  I don't think they'll go away like a fad; the ability to minimize bridge area is crucial.  I think they will be built in some areas that they shouldn't be (maybe this is a result of it being a fad?).  Maybe the biggest DCD shortfall will be the frontage road coordination that someone else already touched on.  On corridors with a lot of through traffic, problems at interchanges will get moved a few hundred feet to the next intersection and construction projects will be considered failures.

I'd like to see more of the aerial images, especially ones with queuing!  In fairness, I don't think that one aerial image is indicative of the performance of an interchange.  Plenty of things could cause this, maybe there was a crash nearby, it was the day of some special event, or the engineer was testing a new signal timing plan.

As far as roundabouts, one big problem is keeping them from totally failing on a busy corridor.  Keep in mind they're typically designed for 20 years in the future so they should be fine after construction until lots of growth occurs.  However, once they hit capacity they can fail (and queue) quickly.  There is no control on arrivals if there are no signals metering the traffic; that's one reason signalized options like DCD are sometimes preferable.

[BigRedDog]

Can anyone speak to how a DDI will handle traffic in a busy shopping area with multiple traffic signals? PennDOT is converting the US 19 at I-70/79 in Washington, PA. US 19 already has 5 signals in a little over a mile stretch (one south of the interchange, four north) through an area with four separate shopping areas.

I travel this area almost daily and I recognize that the US 19/I-70/I-79 interchange is in drastic need of updating. I am no way suggesting that a DDI is wrong here - I don't have anywhere near the training or education to make that decision. But, as a driver in this area, I worry about an often congested corridor becoming even worse.

[Revive 755]

^ Since the I-70/I-79 interchange is a full cloverleaf, I would question whether a DDI would work better than removing two of the loops.  I would assume since the shopping area north of I-70/I-79 appears to be on the edge of town there would be a decent amount of through traffic.  Therefore it might not be the best place for a DDI, but the five-way intersection to the south may do a good job of metering traffic into the interchange.

[BigRedDog]

^ Since the I-70/I-79 interchange is a full cloverleaf, I would question whether a DDI would work better than removing two of the loops.  I would assume since the shopping area north of I-70/I-79 appears to be on the edge of town there would be a decent amount of through traffic.  Therefore it might not be the best place for a DDI, but the five-way intersection to the south may do a good job of metering traffic into the interchange.

Thanks for the input. PennDOT has plans to remove all four loops of the cloverleaf and just keep the four outer ramps. There is some information here on the overall project: http://www.i-70projects.com/I-70andI-79NorthJunction.html

As for the shopping areas, my experience tells me that stretch of US 19 (between I-70/I-79 and Cameron Rd.) has a decent amount of traffic from ~8 a.m. until ~9 p.m., with obvious fluctuations at lunch time and in the evenings. Saturday afternoon can also be tricky.

[Mr_Northside]

Since the I-70/I-79 interchange is a full cloverleaf, I would question whether a DDI would work better than removing two of the loops.

I think one of the reasons of the DDI over that idea is so they can use the existing bridges over US-19 while 6-laning I-70/79.  If they left some loop ramps in, They'd have to use what will be the outer I-70/79 lanes as acceleration/deceleration (depending on what loops were left) lanes for the ramps instead.

[mgk920]

Have the safety numbers on that interchange in New Berlin been improving lately?  If so, I'd chalk that up to what I call the 'drivers ed' issue - getting people up to speed on how they work and how to use them.

The roundabouts around here, including a few with three lanes, appear to me to be working well with good safety.

Mike

[kphoger]

It's true: DDIs take almost no new driving skills to master if you've never encountered one before. Normal merges and straight-through stoplight movements. Roundabouts, however, take a couple of new ways of thinking for drivers unfamiliar with them: everyone keeps moving rather than everyone stopping, priority goes to the left, occasional exits from an inner lane, etc.