Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[inkyatari]

I used to live in the Kankakee, Illinois area, and this overpass on I-57 just south of the Kankakee River always bugged me for some reason that I could never place my finger on.  Then it hit me that there's no central pier on the overpass!

This has to be a rare design.  Any other examples of this?

https://goo.gl/maps/ME4Kib2GU2C2

[roadman]

Before it was recently replaced as part of the Salem to Manchester widening project, the bridge that carries Pillsbury Road over I-93 in Derry, NH had no central pier.

[Big John]

This is common on I-39 south of I-80 in Illinois.

Example: https://www.google.com/maps/@40.7072529,-89.0292785,3a,75y,18.78h,94.39t/data=!3m7!1e1!3m5!1s3PPxiLh8r_3Y1hLsvJE3FA!2e0!6s%2F%2Fgeo3.ggpht.com%2Fcbk%3Fpanoid%3D3PPxiLh8r_3Y1hLsvJE3FA%26output%3Dthumbnail%26cb_client%3Dmaps_sv.tactile.gps%26thumb%3D2%26w%3D203%26h%3D100%26yaw%3D281.94357%26pitch%3D0%26thumbfov%3D100!7i13312!8i6656

[NE2]

[Bitmapped]

WVDOH built a single lane, single span bridge over I-68 a couple years ago. It replaced a large culvert under the highway and connects two halves of a Christmas tree farm. See https://goo.gl/maps/79AMpjWKUqp

[kphoger]

Toll road between Monterrey (NL) and Saltillo (Coah)

GSV #1
GSV #2
GSV #3

[briantroutman]

I’m not sure what you mean by “standard overpasses”, but single-span overpasses with no central support were a design hallmark of the original 1940 Pennsylvania Turnpike (and its extensions through the mid ’50s). In fact, I read somewhere that early design sketches called for a center support but that the BPR requested that the center pier be eliminated to reduce the risk of collision.

[kkt]

But if you're going to run off the left side of the road, isn't it better to run into a stationary central pier, instead of oncoming traffic?

[kphoger]

But if you're going to run off the left side of the road, isn't it better to run into a stationary central pier, instead of oncoming traffic

A central pier would be closer (and thus more likely to be hit) than oncoming traffic.

[briantroutman]

But if you're going to run off the left side of the road, isn't it better to run into a stationary central pier, instead of oncoming traffic?

Excuse the long-winded digression, but I found this to be a surprisingly intriguing question, one in which the answer isn't very clear-cut—at least in terms of this 1940 example. Certainly if you're the innocent motorist who’s on the correct side of the road, you would rather the errant car be stopped by a pier or anything else—leaving you unscathed, regardless of the cost to the other guy.

Now if you are that other guy, I’m not sure which collision would be preferable. Sure, the pier isn’t hurtling toward you at 70, but it’s essentially unforgiving, thus redirecting all of the crash energy back to you rather than deforming itself. It’s also narrower than most vehicles and will potentially concentrate the destructive energy in a smaller area. Of course much of the outcome would depend on the specifics of the collision.

But I think the bigger issue with the dilemma you pose (strike pier vs. oncoming car) is that it tends to assume that the two outcomes are equally likely. Let’s assume you’re on a trajectory that will take you into the median under an overpass—and possibly into the left lane of opposing traffic. I’m visualizing the scenario as a sort of “spaghetti plot” with all possible paths drawn on it:



(Of course there are many more possible paths, but this should give you an idea.)

If there’s a center support under the overpass, that’s the end: There’s a 100% chance that you’re going to collide with the pier—and likely injured severely or killed. All of the scenarios in which you recover your vehicle and return to the correct lane are now taken away from you.

But let’s assume the worst and say that you haven’t recovered your vehicle in the median, so now you’re in the left lane of opposing traffic. There are still a few more things that have to go wrong before you have a catastrophic head-on collision. There has to be a vehicle in the opposing left lane in the same place and at the same time as you, the driver of that vehicle must fail to take evasive action to avoid hitting you, and you have to fail to take evasive action to avoid hitting it. Certainly, all of those things can happen, but I tend to conclude that they’re much less likely than the very severe outcome of a single vehicle striking the support.

Of course this could all be avoided with a wider median and/or a continuous central barrier, which is the path the PA Turnpike Commission has taken on reconstruction and expansion projects since the ’60s.

[triplemultiplex]

Sure, the pier isn’t hurtling toward you at 70, but it’s essentially unforgiving, thus redirecting all of the crash energy back to you rather than deforming itself.

Small point of physics; the speed of two objects moving toward each other cancel out in a collision between those two objects.  They are not cumulative.  This means if two vehicles collide head-on at the same speed, the end result for either vehicle is the same as if they hit a stationary object.
If you ever hear a news report about two vehicles colliding at "combined speed of XXX!!" they are full of shit.

I know you didn't specifically make that error, but I feel like it was implied.
Its a fun physics lesson that demonstrably goes against one's intuition.  Surely, driving into oncoming traffic is going to be worse than driving into a parked car!  But the reality is not necessarily so.

[kkt]

Small point of physics; the speed of two objects moving toward each other cancel out in a collision between those two objects.  They are not cumulative.  This means if two vehicles collide head-on at the same speed, the end result for either vehicle is the same as if they hit a stationary object.
If you ever hear a news report about two vehicles colliding at "combined speed of XXX!!" they are full of shit.

I know you didn't specifically make that error, but I feel like it was implied.
Its a fun physics lesson that demonstrably goes against one's intuition.  Surely, driving into oncoming traffic is going to be worse than driving into a parked car!  But the reality is not necessarily so.

Really?  How does that work?

[D-Dey65]

The new I-75 bridges over US 98/FL 50 are going to have that. The one for the northbound lanes already do.

[1]

Small point of physics; the speed of two objects moving toward each other cancel out in a collision between those two objects.  They are not cumulative.  This means if two vehicles collide head-on at the same speed, the end result for either vehicle is the same as if they hit a stationary object.
If you ever hear a news report about two vehicles colliding at "combined speed of XXX!!" they are full of shit.

I know you didn't specifically make that error, but I feel like it was implied.
Its a fun physics lesson that demonstrably goes against one's intuition.  Surely, driving into oncoming traffic is going to be worse than driving into a parked car!  But the reality is not necessarily so.

Really?  How does that work?

If you hit a stationary object large enough that it won't move when you hit it while going 60 mph, you will stop just as fast as if you hit an object with the same mass as your car while going 60 mph. And you will remain stopped in both cases.

If it actually was doubled to a combined 120 mph, the end result would be that you are going backward 60 mph, not being stopped. This would happen in an elastic collision with a stationary object, but car crashes are designed to be inelastic.

[NE2]

http://gregladen.com/blog/2017/10/11/mythbusters-on-head-on-collisions/
"So it is like a 100 mph collision, shared evenly by two cars (so each gets 50 mph worth!)"

[CtrlAltDel]

If you hit a stationary object large enough that it won't move when you hit it while going 60 mph, you will stop just as fast as if you hit an object with the same mass as your car while going 60 mph. And you will remain stopped in both cases.

So, a car going 120 mph hitting a parked car is a different phenomenon than two 60 mph cars hitting each other? I wouldn't have expected that. How do the cars "know" how fast they are going in this situation?

What if (and I do mean this somewhat seriously), you had two cars heading toward each other each at 60 mph, but the whole road is a treadmill heading 60 mph in one direction? Does that count as the first situation above or the second?

[1]

If you hit a stationary object large enough that it won't move when you hit it while going 60 mph, you will stop just as fast as if you hit an object with the same mass as your car while going 60 mph. And you will remain stopped in both cases.

So, a car going 120 mph hitting a parked car is a different phenomenon than two 60 mph cars hitting each other? I wouldn't have expected that. How do the cars "know" how fast they are going in this situation?

If you hit a parked car at 120 mph, then both cars will be going forward at 60 mph immediately after the collision (and then friction will slow them down). It's the same, as you lose 60 mph (60 to 0 in my scenario and 120 to 60 in your scenario), and the other car gains 60 mph (-60 to 0 in my scenario and 0 to 60 in your scenario).

[kphoger]

but the whole road is a treadmill heading 60 mph in one direction?

That's basically what it is to drive on planet Earth:  a giant 1000-mph treadmill.

[Beltway]

If there’s a center support under the overpass, that’s the end: There’s a 100% chance that you’re going to collide with the pier—and likely injured severely or killed.

Not if the pier is protected with a guardrail, which it almost certainly will be if the pier is less than 30 feet from the nearest roadway.

[briantroutman]

Not if the pier is protected with a guardrail, which it almost certainly will be if the pier is less than 30 feet from the nearest roadway.

Of course it would today—or within the past half-century, at least—but the entire scenario I described was in reference to the original 1940 Pennsylvania Turnpike. As I understand it, original plans in the late ’30s called for an unprotected center support under overpasses, but the BPR advised against it citing the collision hazard, so the PTC standardized the single-span concrete design pictured above. I think I read that in Dan Cupper’s book—I don’t recall the source exactly.

[Beltway]

Not if the pier is protected with a guardrail, which it almost certainly will be if the pier is less than 30 feet from the nearest roadway.

Of course it would today —or within the past half-century, at least —but the entire scenario I described was in reference to the original 1940 Pennsylvania Turnpike. As I understand it, original plans in the late ’30s called for an unprotected center support under overpasses, but the BPR advised against it citing the collision hazard, so the PTC standardized the single-span concrete design pictured above. I think I read that in Dan Cupper’s book—I don’t recall the source exactly.

The PA Turnpike has a narrow cross-section, given the 10-foot-wide median.  Only 78 feet from edge of shoulder to edge of shoulder, given 12-foot lanes and 10-foot shoulders. 

A single span is really all that is needed to span that length.

Even so, they could have protected a center pier with guardrail and appropriate end treatments.  Hitting the guardrail would be far less severe than hitting a bridge pier.

A typical rural Interstate cross-section in a 1970 design would have that length spanning one directional roadway, given a 30-foot clear roadside on the right  and 20 feet on the left.

[jeffandnicole]

And 30 feet really isn't that far.  This is NJ 55, built around 1985. https://goo.gl/maps/JfBJEwKGB2Q2  As shown, there's no guardrails between the roadway and piers, which are no more than 30 or so feet from the pavement.  And except for grass, there isn't much to slow a car down before hitting them.

[Beltway]

And 30 feet really isn't that far.  This is NJ 55, built around 1985. https://goo.gl/maps/JfBJEwKGB2Q2  As shown, there's no guardrails between the roadway and piers, which are no more than 30 or so feet from the pavement.  And except for grass, there isn't much to slow a car down before hitting them.

I have seen that on some other highways with 30 foot clear roadsides.

30 feet is far enough that traffic engineering studies have shown that very rarely will they be hit by an errant vehicle, and that protecting it with a couple hundred feet of guardrail provides enough of a fixed object hazard of its own that it is a 'wash' as to the need of whether or not to install the guardrail.

[DJStephens]

Placement of sand barrels surely has been done in many locales as sort of an afterthought to the potential of impact of unprotected bridge piers by an errant vehicle.   

[sparker]

Lack of a central pier is exceptionally rare here in CA, particularly on overpasses built since about 1958-59, where the concrete superstructure is relatively massive compared to the girders utilized prior to that period.  Some unenclosed girders continued to be part of freeway overpass structure after that period (principally in rural areas), but almost all of those also had center support.