Dumbbell vs Dogbone Roundabout Interchanges

[peterj920]

Roundabout interchanges are becoming more common, but the designs vary.  In my home state of Wisconsin, only Dumbbell Interchanges are being built were 2 full roundabouts are built at a diamond interchange.  Here is an example:
https://www.google.com/maps/@44.5237363,-88.0813119,768m/data=!3m1!1e3?hl=en

Next door in Minnesota and the Keystone Parkway in Carmel, Indiana, a Dogbone Interchange is used where the roundabout is a 3 leg roundabout where the circle does not connect at the near side of the freeway, since it would only be used to make a U turn.  If someone would want to make a U turn, a motorist could continue over or under the freeway and make a U turn at the following roundabout. Here is an example:
https://www.google.com/maps/@44.0584498,-93.2764442,649m/data=!3m1!1e3?hl=en

I think the Dogbone Interchange makes more sense than the Dumbbell Interchange, thoughts? 

[jakeroot]

As I see it:

Dogbone advantages:

- less conflict points, so potentially safer
- Less pavement = lower overall cost (dogbone = cheaper)

Dumbbell advantages:

- Consistent design with regards to all roundabouts (rather than having a missing piece)
- Connecting road can be upgraded with a hard median without having to worry about the lack of a U-turn location that isn't a few hundred meters away (potentially).

I'm not sure there's one good choice here.

[Bruce]

Both designs at least allow for vehicles to re-enter the freeway immediately after exiting off, which is useful for express bus service (unlike certain other designs...).

[triplemultiplex]

Winter snow reveals how most diamond interchanges should probably just have the dogbone style.  There's only ever like 1 set of tire tracks in the snow on the 'inside' quadrant of the roundabout.  What little traffic that would use it as a turnaround can always cross the freeway and turn around there.

The full roundabout at the ramp terminal only seems useful if the surface street has a bunch of RIRO accesses leading up to the interchange.  It would be more convenient for those vehicles wanting to go "left" out of those right only accesses.

[2Co5_14]

With the dumbbell style, since there are so few vehicles making that U-turn, drivers will get used to never seeing vehicles there and potentially ignore the yield sign.  This becomes a safety hazard, and makes the dogbone style a better choice.  As stated before, drivers can still make that U-turn; they just have to travel a bit farther to do it.

[JCinSummerfield]

They installed a dogbone at US-23 & US-223 in Michigan.  More advantageous, IMO, because you do not have to yield right-of-way entering the second half of the dogbone. 

[tradephoric]

Here's a dumbbell style roundabout at 26 Mile & M 53 in Macomb County, Michigan.  The trailer park drive is 700' from the roundabout which happens to be the standard distance for a Michigan left turnaround.  If this was a dogbone, drivers coming out of the trailer park would have to drive 1800' to make their turnaround (drivers wouldn't be able to see the turnaround 1800' away and would just risk making a direct left turn).  It's easy to say drivers would just have to drive a little bit farther down at a dogbone but very few would actually do it. 


https://www.google.com/maps/@42.7139603,-83.0234248,290m/data=!3m1!1e3

[peterj920]

Here's a dumbbell style roundabout at 26 Mile & M 53 in Macomb County, Michigan.  The trailer park drive is 700' from the roundabout which happens to be the standard distance for a Michigan left turnaround.  If this was a dogbone, drivers coming out of the trailer park would have to drive 1800' to make their turnaround (drivers wouldn't be able to see the turnaround 1800' away and would just risk making a direct left turn).  It's easy to say drivers would just have to drive a little bit farther down at a dogbone but very few would actually do it. 


https://www.google.com/maps/@42.7139603,-83.0234248,290m/data=!3m1!1e3

This interchange design makes no sense at all.  There shouldn't be a roundabout there at all because there's 2 ramps that send traffic 2 different directions.  The only cross traffic is people turning left onto M-53.  According to streetview, the trailer park has full access at that interchange.  I looked at the other side of the interchange, and see that there's a lot of right turn only access and people would use the inside curve of the roundabout.  Then I noticed that there is a U turn lane right before the roundabout for people that need to make a U turn.  This interchange looks like it was designed very poorly, not because it's a dumbbell interchange, but because there's 2 ramps that be free flow, but are stuck into the roundabout.  Here's an example of a roundabout interchange where 2 of the ramps are not sent through the roundabout because they are free flowing ramps.

https://www.google.com/maps/@42.9509841,-88.108945,929m/data=!3m1!1e3

[tradephoric]

Here's the other side of the interchange and how it has evolved over the years.  Direct left turns along 26 Mile have been replaced with RIRO designs:

2007:


2010:


2015:

[johndoe]

Not that I've looked that hard at many RCUT, but it's interesting to have dual unsignalized turns.  Or are Michiganders so used to median-left-turns they can handle it?

[tradephoric]

Not that I've looked that hard at many RCUT, but it's interesting to have dual unsignalized turns.  Or are Michiganders so used to median-left-turns they can handle it?

It's uncommon to have dual unsignalized median-left-turns in Michigan.  I like roadway designs that don't require traffic signals so if it works great!  That said, I'm not that familiar with this interchange so I'll reserve judgement to this RCUT's effectiveness.

[english si]

- Consistent design with regards to all roundabouts (rather than having a missing piece)

You mean rather than being a 'longabout' (think of a dogbone as a small roundabout that has been stretched across a freeway, or a big roundabout that has been squeezed so that the interchange needs one bridge, rather than 2 incomplete roundabouts).

Which one to use is relatively site dependant, though most of the time the difference between the two is negliable.

this junction uses both a roundabout and a 'teardrop' (half a dogbone)

[jakeroot]

- Consistent design with regards to all roundabouts (rather than having a missing piece)

You mean rather than being a 'longabout' (think of a dogbone as a small roundabout that has been stretched across a freeway, or a big roundabout that has been squeezed so that the interchange needs one bridge, rather than 2 incomplete roundabouts).

Pretty much. Having each "node" signed as a roundabout, if though they're both only 3/4 of a roundabout, does seem a bit misleading. But I suppose it's not really that big of a deal since you'll always be able to U-turn at some point. The problem is that the US signs each "node" as a roundabout, even though you only need one "roundabout ahead" sign, upon entry to the interchange (a repeater halfway across is not necessary since you are technically in the roundabout).

[english si]

^^ Examples A, B, C and D

[jakeroot]

^^ Examples A, B, C and D

This reminds me...many US transport agencies have been moving away from diagrams and towards lane-use assignments. See the diagram below:

The top right two are becoming more common than the bottom two (at least in Washington), even though the bottom two are way easier to understand.



British Columbia does it right, in more than one case. First of all, they use the more unique dumbbell diagram if it applies (see below). Second, all their roundabouts are signed with a diagram beforehand so you know what to expect (and they do have a very British taste, which is hardly surprising seeing as how British BC is), followed by (if necessary) over-head lane use signs to tell you exactly what each lane does. So in short, you know the shape of the intersection beforehand, as well as the correct lane to be in. Perfect.

Street view of what I'm getting at.

[DaBigE]

The top right two are becoming more common than the bottom two (at least in Washington), even though the bottom two are way easier to understand.

Not that I'm endorsing either style: It's usually because the top right two are smaller, therefore cheaper and can fit easier in more locations. They also closely match the destination signs typically used at "traditional" intersections. I've seen other agencies use plain route markers assemblies as well.

[jakeroot]

The top right two are becoming more common than the bottom two (at least in Washington), even though the bottom two are way easier to understand.

Not that I'm endorsing either style: It's usually because the top right two are smaller, therefore cheaper and can fit easier in more locations. They also closely match the destination signs typically used at "traditional" intersections. I've seen other agencies use plain route markers assemblies as well.

I think that's my problem. Roundabouts aren't traditional intersections. If anything, a diagram helps re-enforce the difference between a four-way junction and a roundabout.

I think at a glance, both types of signs can be understood, but the bottom two are understood quicker.

Seeing as you are a sign-maker, and considering all obstacles such as a budget, what's your opinion (or rather, your favorite)?

[DaBigE]

The top right two are becoming more common than the bottom two (at least in Washington), even though the bottom two are way easier to understand.

Not that I'm endorsing either style: It's usually because the top right two are smaller, therefore cheaper and can fit easier in more locations. They also closely match the destination signs typically used at "traditional" intersections. I've seen other agencies use plain route markers assemblies as well.

I think that's my problem. Roundabouts aren't traditional intersections. If anything, a diagram helps re-enforce the difference between a four-way junction and a roundabout.

I think at a glance, both types of signs can be understood, but the bottom two are understood quicker.

Seeing as you are a sign-maker, and considering all obstacles such as a budget, what's your opinion (or rather, your favorite)?

I prefer the D1-5 series (diagrammatic) version. And if you can't fit it in the ROW, mount it overhead as VDOT is proposing for I-81 at Exit 150. If budget is an issue, the roundabout design and location/context determines what other style I would use. There's no one-size-fits-all when it comes to approach signing. Bias will show a bit here, but I do like the guide sign options WisDOT uses (except 3c), which depends on the site context and amount of necessary way-finding.

[tolbs17]

I like the dumbbell interchange better.

[SkyPesos]

With the discussion about roundabount signage, what's the general opinion on this one? Personally, I like it, though the through movement road could be added to the top (in this case, OH 161 West).

[tolbs17]

With the discussion about roundabount signage, what's the general opinion on this one? Personally, I like it, though the through movement road could be added to the top (in this case, OH 161 West).

It seems okay, the signage has details of where the entrances/exits and through movements go.

Now it seems like NCDOT is ramping up with the dumbbell/dogbone interchanges in rural areas.  There has been a LOT more popping up on highways such as the Monroe Expressway, I-95 widening, etc.

[Scott5114]

I like the dumbbell interchange better.

This is a bad reason to bump a 6-year-old thread.

[tolbs17]

I like the dumbbell interchange better.

This is a bad reason to bump a 6-year-old thread.

Ah. Next time I will add more context before bumping old threads like this.

[roadfro]

With the discussion about roundabount signage, what's the general opinion on this one? Personally, I like it, though the through movement road could be added to the top (in this case, OH 161 West).

I like. I'm surprised we don't see more roundabout guide signage similar to this.

[johndoe]

I thought the intent of overhead lane use signs was to be directly over the lanes, so for that reason I'm not a fan of that sign.  I think that sort of sign could be smaller and "side mounted"