|
Have you ever wondered how roads are designed? Ever questioned why the traffic signal down the road needs two sets of signal heads? Have you ever stared in disgust at a road and wondered, what kind of idiot designed this, and how does he remember to breathe? That idiot is me! I'm a traffic engineer, and I design roads for a living. Well, that’s not entirely true… I mostly fix other engineers' 50-year-old mistakes. I've gone through years of schooling in the field, and worked in both the public and private sectors in 2 countries: the USA and France. I'm quite willing to share my insights and industry insider knowledge with those of you bold enough to ask. I'm rather knowledgeable in the fields of traffic (including signs, stripes, and signals), highway design, and urban design.  I'll tell you everything you want to know (and more) about New England's roads, specifically Connecticut's and Rhode Island's, since that's where I've worked. Engineering rules and theories are quite region-specific in some cases. I'll try my best to specify whether the things I'm writing about apply to one state, to the country, or to the whole world. Please correct me if you see something that could be incorrect.  A few suggested lines of inquiry, as I don't want this OP to get too long: - Why New England's freeway network is so screwed up (Hint: it's your fault) - How we decide where a road is needed, and how it should be built - Dealing with the mentally disturbed and the woefully ignorant - Highway standards in Europe vs. the USA - A primer on interchange design (so you don't have to read the Green Book yourself) - How did this happen? (You show me a highway gently caress-up in your area) - Highway oddities: Spupclos, stop signs on on-ramps, horrible designs that nearly came to be… - Why mass transit sucks in the USA  One disclaimer, and keep this in mind: My opinions on speed limits or red-light running or anything else of that nature should not be construed as giving you permission to perform illegal acts. Some of our traffic laws may be arbitrary, misguided, and obsolete, but you still have to obey them. 
|
#
¿
Jul 27, 2009 23:08
|
|
|
|
| # ¿ Apr 30, 2024 03:40 |
|
|
Crackpipe posted:Can you guys go back to paving highways with concrete again? Please? I-84 crumbles every five years because The Powers That Be insist on going with asphalt, yet the concrete sections have been holding up since before I was born. Ask as often as you like, but your complaints will fall on deaf ears. We've been using asphalt for so long here that it's very unlikely things will change in the near future. A big issue with concrete pavement is that it can last 50 years, whereas asphalt is meant to last 7-10 years. A lot changes in 50 years. Maybe your road needs repaving after 10 because you widen a shoulder, and you have to tear up concrete and waste all that investment. Maybe a sewer line breaks and you have to bust up your nice flat concrete. There's also the issue of ride quality. Theoretically, concrete is rougher to ride on than asphalt. The concrete industry explains that this is because most of our concrete is 30-50 years old, so of course it's rougher than the 5-year-old asphalt that was built to modern specifications. Concrete paving can also lead to problems at joints if the dowels aren't placed properly. One example is on US 1 in Rhode Island, where riding on the concrete road was something akin to trying to drive over railroad ties at 60 mph. Given the limited money we have to work with, I don't doubt a lot of our roads will be built shoddily. I'd rather be able to fix it in 10 years than in 50. And finally, this is more of a political thing, but SOME jurisdictions(no names!) purposely build lousy pavement so that their contractors can soak up more maintenance money and stay in business. Edit: If you think the concrete sections of I-84 are in good shape, assuming you mean the CT portion, take a close look at the Aetna Viaduct, which is currently under emergency repair work. It looks like cheesecloth in some places.
|
|
#
¿
Jul 27, 2009 23:22
|
|
|
Frinkahedron posted:Did you study Civil Engineering? First off, yes, I've got degrees in Civil and Environmental Engineering, French, and German. That was an interesting curriculum, to say the least. Highway oddities will take a while, as there's a lot of ground to cover. Look for some hilarious (and frightening) drawings and photographs in the days and weeks to come. DC is a complicated place, to say the least. Looking at the surrounding cities and pulling some numbers out of my rear end, here's a diagram of how I'd design it. I'm making some assumptions: Baltimore hasn't been reduced to a smoldering crater, I have to preserve the downtown DC area (freeways there are built with tunnels, very expensive), and the airports are full intermodal hubs with train stations for the Metro and commuter rail.  You can see, the basic idea is redundancy, and providing a nice bypass for through traffic on I-95. Most major origin-destination pairs have a direct freeway route.
|
|
#
¿
Jul 28, 2009 00:02
|
|
|
Chairon posted:I was driving along a highway(99, if you have to know) near Sacramento, CA when the road gets all these lines carved in it, maybe half an inch deep or so, across all four lanes of traffic. They cris-crossed diagonally, very close together. There wasn't any road work going on, and this went on for miles and miles, in both directions. What would be the reason for this? Is it just the design of the road? It's really noisy and I think it'd chew up tires pretty quickly. I also made a horrible Paint to show what I mean! The black is the road, and the blue is supposed to be said lines. Replying to you first since it's an easy question. That's grooved pavement. Asphalt isn't all laid out at once; we put down the base course, then the leveling course, then the wearing course. Often, days go by before the next course gets laid. If you've ever made pottery before, you'll remember that wet clay doesn't stick to dry clay, and you have to score it first to get more friction. This works the same way. We score grooves in the first course of asphalt so the second one will bond with it.
|
|
#
¿
Jul 28, 2009 00:04
|
|
|
Crackpipe posted:Have you planned any busways? We're in the (very long) process of designing the New Britain - Hartford Busway. It's arguably the biggest project in the state, and there are no end to the problems we're having, just in the design phase. It's a well known fact that the traffic composition, which is a fancy way of saying "how many heavy vehicles," is the single biggest influence on fatigue. One semitrailer, for example, can do as much damage as a million passenger cars. Asphalt that's meant to last for 10 years under car traffic (and let's be honest, when money's short, it's more like 20) isn't going to last nearly as long when heavy buses are digging ruts in it. quote:Bah, have it your way traffic engineers! It's not us, we don't care too much one way or another. This is a matter for the pavement management and financial folks. quote:Yeah, I was thinking of the sections immediately west of Manchester which aren't in the best shape, but they're older than several typical goons combined. Like I said, concrete lasts a long time. This section isn't as old as you think, though. That whole area was massively redone in the mid-80s to accommodate all the I-384 and I-291 on-ramps, as well as bring a 40-year-old freeway up to modern standards.
|
|
#
¿
Jul 28, 2009 00:13
|
|
|
Billy Maize posted:Awesome. I'm going into Urban Planning (although just starting out, so I'm not entirely positive on what I would necessarily be doing) so all of this is incredibly interesting to me. My country's obsession with the automobile has ruined it for a long time to come. Suburbs are, as I'm sure you're aware, a pretty horrible thing from an urban planning standpoint. Gas isn't going to get any cheaper in the decades to come. Whereas most countries have highly developed urban centers surrounded by farmland, a system that is very well suited to mass transit, the USA is going to have some MASSIVE growing pains. If gas gets scarce enough, and we don't somehow make cars that don't need it, we could very well see abandoned suburbs and a resurgence in urban construction. That's all common sense, though. What I think is that the average American will figure out, about 10 years too late, that he should live near where he works. Our road infrastructure is going to crumble soon without some major investment, and a lot of that money will get thrown toward rail and bus lines. No matter how much you love your car, if it's three times cheaper to take a bus or train to work, you're going to start thinking.
|
|
#
¿
Jul 28, 2009 00:17
|
|
|
BigHead posted:Do you also design red lights and timings of red lights? If so, why the Christ can't every city be designed so that, if you travel the speed limit on major roads, you hit only green lights (like in Anchorage) as opposed to hitting every single goddamn red light (like in St. Louis)? There's a fun question! I'll make an entire post about signal coordination, rather than hide the answer here. quote:Are all the pictures roads you've designed, or are they a representative sample? Of what I showed in the OP, I've only personally done the Hartford freeway model. I'm only 24; my construction experience is limited to working on one freeway in France, some transportation management, and a few dozen repair/rehabilitation projects in the US. quote:Do you have a favorite "style" (for lack of a better word) when it comes to designing highways? Do you prefer all straight lines, such as the last picture in the OP, or all curvy circular roads like the other picture? I prefer freeways to have gentle curves and slopes. It's much more alluring, more fun to drive, and keeps the driver alert. There's nothing interesting about driving for an hour in a straight line! Realistically, though, the engineer has very little say on where exactly a road goes. That's more dependent on who owns the property it needs. quote:Do you have a favorite specific city or area you've designed? Like I said, I haven't designed much. One city I really like is Chandigarh. It was designed by LeCorbusier, one of the pioneers of urban design. I play a lot of SimCity, and something about its modular nature really speaks to me.  quote:Do you deal with construction companies? What happens when a construction company has a terrible terrible reputation - for instance they take three times longer than another other company - but somehow keeps winning the low bids? Do you all grumble about it? In other words, would you analogize construction companies and roads to, say, hamburgers and Morton's vs McDonald's? Unfortunately, we always go with the low bidder. If someone REALLY pisses us off, for example, by faking inspection reports, we blacklist the company. Then the head retires, his son takes the top position, the company changes its name, and it's back to business as usual. quote:What are the best ways to start a career in this line of work? Get a civil engineering degree with a concentration in transportation and pray that someone's hiring in this economy. quote:Do you ever work with other urban planners to design highways that conveniently service certain areas, or is it all about traffic control? Speaking of which, what are several factors you take into consideration when designing a highway? We prefer to build highways proactively, rather than reactively. If a new airport, a big casino, or an industrial park is going in, or if the population's growing so fast that we'll need something in 20 years' time, we will plan the project ahead of time to handle the anticipated need. Unfortunately, around here in New England, there is very little unsettled land and a lot of environmentally protected areas, so we are often forced to work reactively. That means adding lanes where we can, widening roads, etc. As for design factors, it's a combination of geometrics (we have a lot of standards), available right of way, environmental concerns, and leaving room for future expansion. I'll talk about it more in detail as the thread matures.
|
|
#
¿
Jul 28, 2009 00:31
|
|
|
nm posted:Have you ever done a speed survey? Have they every followed it? Answering this alone because it's important. In our theories of design, there is one number that shows up again and again. In France, they call it the V15. Here, we call it the 85th percentile speed. This is the speed slower than which 85% of drivers travel. It's usually 5-10 mph above the average speed. (There are two ways to calculate mean speed, but that's a story for another time.) This 85th %ile speed is supposed to be equal to another number, the elusive design speed. The design speed is the maximum safe speed for the worst design vehicle on dry pavement. For example, if a freeway has a design speed of 75 mph, then a turnpike double (huge tractor-trailer) could safely negotiate its curves at 75 mph on a good day. Most cars, obviously, could safely go much faster, even in adverse conditions. And then we have the third important speed, the speed limit. This is ALSO theoretically equal to the 85th %ile speed and the design speed. Sounds simple enough, right? Every road I have a project on will have a speed survey done. They generally show that the 85th %ile speed is pretty close to the design speed. This is the Northeast, so I wouldn't be surprised if it's a little faster; we don't like to waste time. I try to set the speed limit around that point. Then Granny McGee and her 40 bingo buddies write an angry letter to the mayor, complaining that people are speeding and how unsafe the road is! Next thing I know, I'm putting up a "25 mph" sign on a road that I know can handle 40 with no problem. Now lowering the speed limit sounds like it would be pretty safe, right? Welp, buddy, that's why I'm the traffic engineer and not you. Turns out, it's not so much the speed that makes accidents, but the speed differential! When Granny McGee's blustering along at 40 mph, trying to keep up with the rest of the cars, nobody's getting rear-ended. But once she slows down to 25 mph, and the guy behind her is trying to go 50, that's a dangerous situation. What we see when we check speeds on almost any road, is that the 85th %ile speed is generally 10 mph faster than the speed limit. Does that mean that the MAJORITY is speeding? Probably! Does that mean the speed limit should be raised? Probably! Does that mean you can drive as fast as you think is safe? Well, I already answered that one in the OP. You've gotta follow the laws, no matter how dumb they are.
|
|
#
¿
Jul 28, 2009 00:43
|
|
|
Now to address the rest of your post.nm posted:Also, what the gently caress were they thinking with this? That's known as Weaving, a very important term in highway design. You're right, it's a substandard design for several reasons. The weaving is due to a left entrance from 62 onto I-35, which is pretty much taboo these days. I-35 is the main line here, so all entrances and exits should be to and from the right side. quote:In fact, Minneapolis had a whole bunch of these shared roadway interchanges (for example, move a few miles north on 35W and you get a similar disaster up at 94). They are all a disaster if you get more than 10 cars on the road. In general, when you have multiple routes on the same roadway, it's called a concurrency. Unfortunately, in this case, I-35 and I-94 are both important roads, so there's no obvious priority. Ideally, the ramps would cross over the outer highway so all entrances and exits were on the right, but money and space are important constraints. Overall, though, that interchange doesn't look too bad. Except for I-94, all the entrances and exits are on the right. The biggest problem there isn't weaving, it's that the ramps are too close together! Anyone out there who lives near a city can laugh at this with me: exits shouldn't be closer than a mile apart. I repeat, you shouldn't see more than one exit per mile, on an ideally designed road. Around here, though, it's not unusual to see 3 or 4 exits in a half mile.
|
|
#
¿
Jul 28, 2009 00:51
|
|
|
nm posted:Another question, so i stop editing I'll make a new one. Many, many many cities actually had excellent public transportation (referred to as transit in the traffic engineering lexicon) prior to the second half of the 20th century. Light rail, specifically, was quite prevalent around the world. Unfortunately, many jurisdictions (I'm looking at you, California) decided it would be prudent to turn their rail lines into extra lanes for their freeways. After all, gas was cheap, commutes were short, and automobiles were the wave of the future! Most light rail got replaced by bus in the 30s-70s, but here's the hilarious part: light rail is on the rise again in a big way. Cities all around the world (Europe's at least a couple decades ahead of the USA in this respect) are rebuilding their old tram lines, or making new ones. We figuratively just finished ripping up our old train tracks, and now we get to build them all over again! I love this country. Billy Maize posted:I just wanted to point out how much this pisses me off. For anyone reading this who hasn't heard about it, go read about the Great American streetcar scandal. Exactly. Boston used to have a much more extensive system, too. I wish engineers in the 50s had had some sense of foresight.
|
|
#
¿
Jul 28, 2009 00:57
|
|
|
AAPsel posted:I'm a civil engineer myself, though not a traffic guy. I remember reading reading a comparison of highway capacity per country way back in class though. Most euro countries were able to transport around 1900-2100 cars per hour per lane before congestion set in. Germany somehow managed to outperform everyone at 2500. I think US and Italy were the worst with 1600 or something similar. Let me answer your first question really quickly, because the second one is awesome and deserves its own post. The capacity of a freeway depends on the traffic composition I talked about earlier, the grade (slope) of the road, and the geometry of the roadway. On our freeways in Connecticut, we can get 2500 pcplpm (passenger cars per lane per mile) in good conditions, and 1500-1800 when the road is under construction. One of the many paradoxes of traffic engineering is that traffic flows best when the road is mildly congested. Time to get technical! We measure how smoothly traffic is flowing by LOS (Level Of Service). A is the best: you're the only one on the road, you can go as fast as you want. At LOS B, there are a couple other cars, but they don't impede you much. LOS C is a bit rougher, and you have to be careful when changing lanes. LOS D is quite stressful. Everyone's moving at a good clip, but you don't have much liberty. LOS E, you're going 10 below the speed limit and everyone's crammed in like sardines. Finally, at LOS F, it's stop-and-go, or just plain 'stop'.  You can see here that speed goes down as volume (and density) go up. However, since there are many more cars on the road at LOS E than at LOS A, even though they're going more slowly, the road is handling more traffic and therefore more efficient. Fascinating, eh? That's why a German road that's often congested can actually be better than an Italian road that nobody uses. Or maybe the Italians just can't drive. 
|
|
#
¿
Jul 28, 2009 01:13
|
|
|
psydude posted:Who determines the speed limits on highways? Much as I'd like to say the engineer sets the speed limit, Granny McGee and her bingo buddies have about 100X more influence than me.
|
|
#
¿
Jul 28, 2009 01:15
|
|
|
Steppo posted:Who in the hell thought that the route 395/2 interchange in Connecticut was a good loving idea? Yes, I almost got into a crash at that very spot. So now's a great time to talk about WHY CLOVERLEAF INTERCHANGES SUCK! Here is a handy diagram to demonstrate why cloverleafs introduce weaving.  Any time you have an on-ramp followed closely by an off-ramp, weaving will be an issue. Mister Red in the diagram needs to get to the left, so he doesn't get trapped into taking an unintended exit. Mister Blue, however, wants that exit. See how they're pretty much on a collision course? That's weaving, and it's a major problem in highway design. Mister Green here isn't helping matters; he wants to take the next exit, and he's going to shift to the right, not paying much attention to Mister Red moving into that same lane. Cloverleafs are bad because they introduce FOUR of these weaving sections into a single interchange. That's a massive safety issue. We'd much rather replace them with diamond interchanges or partial cloverleafs (parclos), even if it means lowering capacity. Oh, one last thing: cloverleafs have no expansion potential, since putting two lanes on a loop ramp (the 270 degree ramps that make the "leaves") is next to impossible, increasing the ramp design speed by a few mph means making the interchange MUCH bigger, and cars have to go way out of their way instead of going in the direction they want.
|
|
#
¿
Jul 28, 2009 01:25
|
|
|
nm posted:Thanks I'd love to have the cities and towns back me up like that. There's so much animosity between the residents and the people who set the speed limits! They feel like the speed limit is the reason that people speed. Yes, lowering the speed limit in conjunction with strict enforcement will make 80% of drivers go slower, but the 20% who don't are the ones who drive the fastest anyway.
|
|
#
¿
Jul 28, 2009 01:29
|
|
|
potato of destiny posted:So bookmarking this thread. Yes, we have some systems that can change the speed limit. I remember seeing some while driving through New Jersey (not a journey I'd recommend to anyone else). I think they respond to the road conditions more than the speed of the drivers. I know on the Autobahnnetz, variable speed limits are common because the roads are much "smarter." There are cameras and sensors all over the freeways, keeping track of road conditions, speed, occupancies... we just don't have the money here to install that sort of monitoring everywhere. There are some places where we're experimenting, but most of the work I've done in transportation management has focused on reacting to incidents, not to normal operations.
|
|
#
¿
Jul 28, 2009 01:52
|
|
|
Tom Steele posted:This is actually very interesting and something i've always wondered about. Can you do that post about traffic signals please? Since I promised it earlier, here's how coordination works!  COORDINATION 101 This is called a time-space diagram. The little figure on the left represents what we call a corridor. This is a major road, and we want to keep the traffic moving on it. At each of the four intersections shown there, there's a signal. Since this is the main road, it'll get about 50% of the signal's cycle as green time, and, logically, the other 50% is red. These red times are shown to the right as horizontal bars. It should be pretty obvious what their widths represent if you look at the axis labels. Now, we've got vehicles driving through that corridor, shown as blue and magenta lines. If X represents time and Y represents space, then the slope of the lines represents the speed of each vehicle. Each time the car has to stop (and we don't bother curving the lines to represent acceleration and deceleration), the line goes horizontal. Makes sense, right? So, we want as few stops as possible. In this diagram, the lights aren't coordinated very well, since there are a lot of stops. These stops waste time and gas! Therefore, we move around the red times and offets between lights and get this:  You see? Much fewer stops. All we had to do is move things around a bit. The number of cars that can get through the corridor without stopping is called the bandwidth. Unfortunately, this sort of situation is VERY hard to achieve in real life, but that's a topic for Coordination 102.
|
|
#
¿
Jul 28, 2009 02:00
|
|
|
Let me tell you a little more about my first transportation engineering job. I worked in what's called a TMC (Transportation Management Center.) How does one manage transportation, exactly? Well, check this out: It's basically a bunch of "operators," sitting around in front of some BIG TV screens and watching video feeds from highways around the state. We had nearly 100 cameras when I was there, along with several dozen electronic signs, a few radio towers, and live feeds to TV and the website. I had a lot of fun at that job. We'd sit around for 8 hours, watching cars crash and then responding to the crashes by putting up advisories on the website and radio and making messages on the signs to warn people of the incident ahead. I saw some amazing incidents in my time there, stuff you only see on the news, and some horrible things, too. I could fill a whole thread with the things I saw there... and maybe I will! I learned a lot there, and here's something everyone should know about incident management. First off, if you see an electronic sign on the road, it's either the state's or a contractor's. Never trust a contractor's sign. The state's sign, though, should be the easiest and most reliable way to find out about an accident ahead. If it says to seek an alternate route because of an incident, you should really consider it. The phone travel advisory (dial 511 in most states, try it now!) is also usually up-to-date. It carries scheduled construction information, as well as accident information. Most DOTs also have a website that shows where there is an incident. These are all pretty reliable. But the radio, that's a different story altogether! They're a TMC's lowest priority, because they generally have horrible reception, take a long time to scroll the 60 seconds of station callsign BS and routine advisories, and usually have a delay of 5-10 minutes or more. We have to record those voices ourselves, and that takes precious time that we need for alerting the police, setting up messages on the signs, and blacking out the cameras so that our good friend Granny McGee doesn't turn on the traffic channel and see her grandchildren smeared across the high-speed lane on I-95. Speaking of the Highway Advisory Radio, it's entirely possible you've heard my anonymous, velvety voice before and not even known it! Imagine that.
|
|
#
¿
Jul 28, 2009 02:12
|
|
|
teejayh posted:Have you heard of the Diverging Diamond Interchange, and if you have, what do you think of it? I love them. Crazy stuff like the DDI makes my job interesting! Let me draw out a quick diagram for those who don't know what I'm talking about, though:  This is a standard diamond interchange. Note that there are two signals, each of which has at least three phases because of the two left turns at each one. It's a pretty standard kind of interchange, though not the safest, strictly speaking.  Here is a DDI! It has the same number of signals, but each one only has 2 phases, making them more efficient. Additionally, there are no left turns. The weird part is that, on the overpass, traffic's driving on the "wrong" side of the road. Pretty cool, eh? I'd love to see more of these, but, unfortunately, it'll be a long time before they show up here. Americans are just too darn stupid to use complicated interchanges like that! Just about anybody can get a driver's license here, which means we always have to design for the stupidest motorist. You'll see that theme pop up again and again. Edit: There are a couple of these in France. Also, we have a half-formed DDI right in Rhode Island! It doesn't have the best reputation, unfortunately. 
Cichlidae fucked around with this message at 02:52 on Jul 28, 2009 |
|
#
¿
Jul 28, 2009 02:49
|
|
|
Panzerfaust posted:Thanks for this thread, Cichlidae. I'm moving to Toronto at the end of next month to start an urban planning degree, so all of this is very interesting, though my interests lie more in the transit sphere than the highway sphere. Yes, there aren't many cities that can boast an original system. Eastern Europe's done extremely well with their light rail; on a system 40 years old, a Hungarian (if I remember correctly) tram is carrying 400,000 passengers per day on a single track! Compare that to a standard freeway lane in great shape (25,000 cars per day), that single track is carrying the equivalent of 16 lanes. "LRT everywhere" might not seem like that sound of a strategy, but with efficiency like that, it's hard to argue with putting it wherever you can reasonably afford. A city half the size of Toronto, Brussels, has multiple metro and premetro lines, as well as several trams and a dense network of buses. My approach to mass transit is that more is better. It's very hard to saturate rail lines, and more light rail means more direct trips and shorter headways. Edit: Check this out. 24 tram lines, 6 metro lines. Cichlidae fucked around with this message at 03:04 on Jul 28, 2009 |
|
#
¿
Jul 28, 2009 02:59
|
|
|
FyreStar posted:Great thread, OP. Technology definitely makes things a lot easier. Traffic engineering used to be a field based more on guesswork and experience, but now you can just enter the traffic volumes into Synchro, press ctrl-T, and get your preliminary signal timings. Lazy engineers stop there, but the good ones do a lot more analysis, even going out in the field to tweak the timings manually. Honestly, computer models are great, but they don't account for regional variations. Synchro's drivers are all naïf, and VISSIM's drivers are idiots. There's so much tweaking required to make people act "normal," and even then, it's all just guesswork. Unfortunately, traffic engineering is on the bleeding edge of technology. Other countries, like England, France, Germany, and Australia, are doing some amazing things with real-time traffic monitoring. Here, though, we're just starting to get into it. Within 20 years, I hope we'll have a system to track every single vehicle in the system and allocate green time accordingly to minimize delay for everyone. It's possible; we could even do it now with enough money. As cars get "smarter," it'll get cheaper to monitor them. Two decades is an eternity for technology, though. Who knows? We could all be telecommuting by then, and then traffic wouldn't be much of an issue. I've got to get to bed, I'll check in tomorrow after work. Feel free to rack up some interesting questions. I'll also do a presentation on different types of interchanges, so you can learn about : Parclos! Spupclos! SPUIs! Volleyballs! Stacks! Directional Y's! Trumpets! Edit: Davenport posted:Following up on your post with the time-space diagrams, am I correct in assuming that none of those intersections have the traffic signal lights on a sensor? What is your opinion of the sensors? Where I live (Sacramento, fabled city of awful traffic engineering), there is a four mile stretch of road (Truxel, from I-80 to Elkhorn) that has about 2 dozen lights, and some days I will hit every drat one because one solitary vehicle on a cross street or opposing turn lane will trip the sensor and stop the flow of 30-50 cars over and over and over. That's a subject for Coordination 102. Basically, it's really really hard for current (1980s technology) signal controllers to adaptively allocate green time to different approaches by shifting cycle lengths. Check out England's take on the subject, both of you. Cichlidae fucked around with this message at 04:17 on Jul 28, 2009 |
|
#
¿
Jul 28, 2009 04:14
|
|
|
SlapActionJackson posted:Even for construction of new rail corridors that spectacularly fail the is-it-better-than-a-bus test? Yeah, from these three posts, it seems that Toronto's not really that good at planning. It's a little appalling to think that a municipality could be spending so much on improving its infrastructure, especially in expensive urban areas, but not do it properly. I'd be willing to bet, though, that it's not the engineers who are screwing things up. It's probably a powerful politician from whatever neighborhood who wants a tram line to himself. We all know the type, right? As to unprotected, or as we call them, permissive left turns, they're not as safe as protected left turns. We generally eliminate protected left turns when the number of left-turning cars in the peak hour gets around 200. Obviously, it's more efficient to have them turning whenever they can find a gap, but it's also a safety issue and it forces oncoming through traffic to slow down. It's especially an issue when there's a median, and those left turners need more time than they think to cross the oncoming lanes. Pedestrians crossing the side road on a dedicated ped phase are a concern as well.
|
|
#
¿
Jul 28, 2009 12:15
|
|
|
nm posted:Speaking of the autobahn. The 100 mph issue is more of a liability issue than a road issue. Most rural freeways (straight, flat, low volumes) can easily handle that. Heck, I know some people who drive 120+ on the roads here that were designed for 50 mph. Of course, it's not entirely safe. At that speed, it's going to take you 18 seconds to come to a complete stop, and in that distance, you cover more than a quarter mile. That's why you need a flat, straight road, so you can see far enough ahead to stop in time. Of course, there's always the looming spectre of a dear stepping out into the road, or a tire popping, or some other rear end in a top hat rear-ending you at 140 mph because he passed out while getting road head from some bimbo in his Porsche. Modern cars just can't take that sort of abuse. If you hit a deer doing 120, you're going to die, and if everyone around you is going 120, a lot of them are dead, too.
|
|
#
¿
Jul 28, 2009 12:22
|
|
|
DaisyDanger posted:How do you feel about New Jersey's (in)famous jug handles? I learned to drive there and actually liked them, but I can understand how they can be confusing. Jughandles are pretty awesome if you have a high left-turning volume. New Jersey, awful as it is, has brought a few interesting innovations to the traffic engineering world. The only reason jughandles aren't used more around the world is that it's a bit confusing if you've never seen one before.
|
|
#
¿
Jul 28, 2009 17:11
|
|
|
chiyosdad posted:Is it this one? Oh, please let it be this one! I'll have plenty of fun stories once I have time to tell them, don't worry! As to Burgers' Equation, it just looks like the usual fluid mechanics stuff, like a wimpy Navier-Stokes equation. I can definitely see how it would be applicable, though I've never used it before. Traffic, water flow, heat flow, and electric current are excellent analogues once you get past the obvious differences.
|
|
#
¿
Jul 28, 2009 17:13
|
|
|
Wallrod posted:Is the general absence of roundabouts/traffic circles in the US due to the aforementioned need to account for the lowest common denominator of drivers, or some other factor? (edit, rephrased): How is French highway design unique, compared with other countries, if at all? The biggest difference between the USA and other countries is that it's much easier to get a license here. That means that there are some REALLY stupid people on the road. For every 9999 people who know how to use a roundabout, there's always 1 who will try to drive the wrong way, or forget to stop, or be so obsessed with picking boogers out of his nose that he confuses the gas pedal with the brake pedal. France has more roundabouts than any other country in the world (and I think they have more than 50% of the world's roundabouts), and French drivers are awesome at using them. I saw dozens of distinct roundabout types there. I wish we could use them here! Look at this beautiful SOB I designed.  quote:My father works for a roads and highways maintenance office here in the UK, so i've managed to inadvertantly pick up bits of info about road design and construction and so on, so this subject interests me in a strange way I love designing crazy and underused things, but that's much more of a personal hobby than my job. I do the traffic design from beginning to end, so there's a lot of concept work, and a lot of what we call "engineering judgment." I've learned what kinds of things will work in certain situations, and some of my designs go through to construction with only minor changes. Then sometimes I design stuff like this, just for fun: 
|
|
#
¿
Jul 28, 2009 17:26
|
|
. Does your work mostly only cover the logic of traffic management, or does it involve in-depth designing of materials and construction of the roads as well? How far are your ideal designs from realistically possible ones, on average?
|
teejayh posted:The reason I asked, is that we're in the process of building one in Lehi Utah right now. You're right that it has the population all stupid WHAT DO YOU MEAN WE DRIVE ON THE LEFT, but after looking at the initial designs I figured it would be a great way for traffic flow, especially since it is connecting one of the fastest growing areas to the freeway. I understand how CFI's work (I'm on lunch break so I'll crib this image off Wikipedia), but I'm not a fan, personally. I'd rather see a dual jughandle, as it has the same effect of cutting out the artery left turn phase, but without making the left turners cross oncoming traffic.  The left turns still have to cross oncoming and side-street traffic. Basically, it's spreading out one intersection over three. I can see it working very well on wide arterials, but in New England, where land is scarce, I don't think we could come up with the right of way to build one.
|
|
#
¿
Jul 28, 2009 17:32
|
|
|
DeusEx posted:But it works here in Germany. Most deadly accidents don't happen on the speed limitless Autobahn, but on rural highways (Speedlimit 100 km/h, unless otherwise indicated), where people loose control of their vehicles while overtaking or taking a curve too fast. Actually Autobahn is usually considered to be very safe to drive, despite the high speeds there. You guys also have much stricter penalties for things like tailgating, and have to pay something like 2500 euro to get your license, right? I'd say Germans are safer drivers overall than Americans. Heck, 3/4 of our accidents on freeways here are caused by the nebulous mistake of "following too closely." quote:EDIT: The mantra of roundabout enthusiasts is "wide nodes, narrow roads." The idea behind that is that you need to buy up more land around the intersection proper, but you can omit the turn lanes that frequently go a couple hundred meters upstream of the intersection, making the road itself narrower. Now the roundabout vs. signalized question is too big for a single post, but let me summarize the key points on each side of the debate. - Four-way stops result in high accident rates, and require good sight distance on every approach, which is hard to do in densely settled or wooded areas. - Four-way stops are best used with only one lane of incoming traffic on each approach. More than that, and the accident rate skyrockets. - When there are too many cars for a four-way stop, or when accidents are too high, (or for some other reasons I'll likely explain later), we consider putting in a traffic signal. - Traffic signals often raise the accident rate, since they encourage rear-end collisions, but these accidents are much less severe than the angle collisions at a four-way stop. - Traffic signals can handle lots and lots of cars. - Roundabouts are safer than either signals or stop signs, UNLESS you're a bicyclist or pedestrian. - Roundabouts work best when traffic volumes are distributed evenly among the approaches. - Roundabouts can only be used on relatively flat terrain. - Roundabouts can't handle as many cars as a signal. When a roundabout gets congested, it acts as a de facto four-way stop. I'd say that roundabouts are generally preferable to four-way stops, but there are several conditions that would interdict their use. They're also excellent for traffic calming (slowing people down), acting as a transition between a freeway and a local road, and solving accident problems.
|
|
#
¿
Jul 28, 2009 17:40
|
|
|
gigabitnokie posted:TDM professional checking in! Here's some super awesome traffic engineering know-how for you! The length of a left turn lane (in feet) is, as a rule of thumb, equal to the number of left-turning cars in the peak hour. You have 200 left turners, you put in a 200-foot lane. See why we haven't switched to metric yet? Having longer left turn lanes are fine, but that takes up more land ($$$) and costs extra money to maintain. It's a financial issue, not a "200 feet is more than enough for anyone" issue.
|
|
#
¿
Jul 28, 2009 17:43
|
|
|
Stew Man Chew posted:First off this is an awesome thread and you are awesome, I've seen your commentary in other threads and always kinda hoped you would open up about your job. Yeah, the weaving there looks pretty bad. When was that area designed? I know Dallas has grown pretty quickly, and we generally only plan 20 years ahead when designing freeways. Weaving conflicts are strongly dependent on traffic volumes. Perhaps it wasn't a problem when the road was designed, but it's certainly one now that volumes are higher. If someone paid me to fix this, I'd rearrange the 114/121/97 interchange to offer left and right exits to the airport, and probably introduce what we call a collector-distributor (C/D) road to prevent weaving. If I have some time later, I'll draw it up, but I've probably already answered your question
|
|
#
¿
Jul 28, 2009 17:47
|
|
|
Roichlem posted:Thank you for this informative thread. I have no idea how Japan designs their roads, unfortunately. When signals seem "laggy" like that, it's generally because there are no detectors on them, so the signal is just running pre-timed. In short, no matter how many cars are on the side road, they'll always get the same green time. Another possibility is that there are detectors, but some of them are broken. We get broken detectors all the time, and if maintenance doesn't go out and fix them, it can make for some pretty bad backups.
|
|
#
¿
Jul 28, 2009 17:48
|
|
|
Billy Maize posted:I like roundabouts way more than 4 way stops but nobody knows how to use either one. One last one before I go back to work! That's called a dogbone interchange, and when implemented properly, it works super well. It looks like you guys kind of hosed up, though. That western roundabout should be tear-shaped, not round. It has too many roads coming into it. Check it out.
|
|
#
¿
Jul 28, 2009 17:51
|
|
|
Alright, I'm back from work! I took 10 minutes to re-look at the Dallas Fort Worth problem, and here's my back-of-the-napkin solution.  The different colors aren't significant. Basically, to stay on 121, you stay to the left, which is in line with modern design standards. This design also doesn't have any weaving problems, and doesn't use up much more land than what's currently out there. Now, for a lesson on interchanges!  This is a trumpet interchange! You'll see these all over. Very convenient for a three-way connection, where the minor leg has low volumes.  Here's a directional Y interchange. It's higher-capacity than a trumpet. You can tell this one isn't up to spec, because it has left exits and entrances.  Here's a cloverleaf; you've all seen these. Your grandchildren, however, might never see them.  One of New Jersey's lesser-known innovations is the at-grade cloverleaf. All the weaving problems of a cloverleaf with the added danger of roads crossing!  This is a partial cloverleaf, or parclo. When two loops are in adjacent quadrants, as here, weaving is an issue. There is a collector/distributor (c/d) road here to lessen the weaving.  This parclo has loops in alternate quadrants, and therefore no weaving!  This is called a volleyball interchange, or a three-level diamond. It looks pretty cool, but if you want to turn left, you have to go through 3 lights.  This is a turbine interchange. It feels like you're driving down a toilet.  Here's a stack, the most efficient and highest-capacity of all interchanges. Note that it only has one structure!  And here's a high-five, a stack combined with a volleyball! I wish we had as much money as Texas...  Finally, since there's so much roundabout interest here, this is a roundabout interchange. Very cool! Every interchange is unique, and I haven't even shown you freeway-road interchanges yet, so we're just scratching the surface! Yay!
|
|
#
¿
Jul 28, 2009 22:00
|
|
|
Billy Maize posted:I would also be interested in seeing anything in Sim City you've laid out that you're proud of. I love the game and go through phases of playing and not playing, but I usually plan too big of areas and then I get bored zoning them how I want. Do you use the Network Addon Mod (NAM)? I want SimCity 5 so badly  Why, Maxis, why did you have to contract out your franchise to some imitator? I want it more complex, not simpler! I want to place ramps, customize interchanges, set lane widths and signal phasing... but I'm probably the only one who would appreciate the micromanagement. I've never heard of the NAM before. What does it allow you to do?
|
|
#
¿
Jul 28, 2009 22:01
|
|
|
DeusEx posted:Not really, overall we have very lax laws considering traffic violations compared to the rest of the EU (except DUI/DWI). Tailgaiting will only be prosecuted if you were REALLY close or drive a truck. Obtaining driving licenses is indeed quite expensive though. "It works amazingly well" is kind of deceptive. I'm sure their accident rate is much higher than it is here. American roads have a much higher emphasis on safety than Indian or Chinese roads, and we sacrifice a lot of capacity and money to keep driver fatalities down. New England, by the way, does a great job of it! The safest four states to drive in are Massachusetts, New York, Rhode Island, and Connecticut. Hard to believe, eh? As to the "take down all the extra signs" concept, that's called naked roads. I'm glad to see it's catching on in Europe. Frankly, I agree that we have too many signs, and it's true that accident rates drop significantly when people are forced to pay attention like that. What's preventing that in the USA is liability. If I don't put up that "school bus stop ahead" or intersection warning or curve warning sign, and someone rear-ends a bus or drives off the road, I could quite easily get sued. It's in my best interest to put up as many signs as possible, whether or not they're really warranted.
|
|
#
¿
Jul 28, 2009 22:06
|
|
|
Phone posted:How do you feel about turning left on red arrows? I've gotta say, that feels pretty weird on a gut level. While the country as a whole has a signal control bible (free PDF, check it out to see what I have to deal with), it offers a lot of leniency in what individual states can do. Some states have a flashing red or yellow arrow to indicate a permissive left turn, some (CT and RI included) have a green ball (we call them balls, to you it's a normal circular light), whereas a green arrow indicates protected movements, and the new draft MUTCD proposes even more schemes. Personally, I'm a fan of the green ball, but if I saw some evidence that another scheme was safer, I'd switch over in a heartbeat.
|
|
#
¿
Jul 28, 2009 22:11
|
|
|
BadSamaritan posted:Having grown up in Rhode Island, I've always wondered about the Pawtucket S-Curves- why would you do that to a highway? That and the Thurber's Avenue curve- the exit/interchange itself is nice, but the heavily banked, densely packed curve is something else. Really that whole corridor is messed up... What I'm about to tell you is hearsay and absolutely must not be repeated, got it?  JUICY GOSSIP ALERTBack when I-95 was designed in Rhode Island, it was necessary to knock down a few buildings. Sometimes, a building was important enough that the proposed alignment would have to be changed. In Pawtucket, I-95 was supposed to be straight and smooth. Unfortunately for future generations, a powerful politician was a frequenter of an establishment of ill repute slated to be demolished by the freeway. Of course, he wasn't keen on losing his hang-out. I'm sure you can guess what happened, right?  While working at the TMC, it was a rare occurrence to actually see an accident happen in real time. Those S-curves, though... it was like clockwork. We had a camera right near the overpass there, and would catch an accident on film at least once a week. The camera itself, despite being behind guardrail and on an embankment, was demolished by errant vehicles every few months. quote:Why in general would you say New England's highways, especially Boston (oh god Boston) are like this? Is it because of the area's age/growth patterns? You mentioned in the OP it's our fault- what can we do to make it better? This merits a post of its own, because you guys are not going to believe the amount of hosed-up stuff that happened in the 70s.
|
|
#
¿
Jul 28, 2009 22:17
|
|
|
Undeclared Eggplant posted:I was just reading about Chicago's Circle Interchange, which is also a stack, but is also apparently the country's third-worst traffic bottleneck. So if a stack is already the highest capacity interchange, what do you do if it's still not enough? Spread everything out more? That's a turbine interchange, actually. A stack doesn't require cars to go significantly out of their way. It's a bit confusing sometimes because laypeople use "stack" to refer to any interchange with a lot of ramps. I can see why it would be congested; 300,000 cars is an awful lot. We have about the same amount going through the 95/6/10 interchange in Providence, or the 84/91/6/44 interchange in Hartford, and it leads to some massive backups if things aren't designed perfectly.
|
|
#
¿
Jul 28, 2009 22:21
|
|
|
Mister Gopher posted:How do you deal with the different design and funding inputs from a variety of sources (Fed/State/Local) for highways? Do you have a separate set of guidelines for each that you must do, or what? Yes, that's how it works. Different kinds of roads are eligible for different funding sources. There are too many to list here, but basically, you'll get more federal money for something like building a busway or fixing safety problems on an interstate highway than you will for filling in potholes or building a new road. Connecticut gets 80-90% federal funding overall, which is much higher than most states. Now obviously, if you're getting federal funds, you need to meet federal design standards. It's ok to skimp a little bit, but get too slack and the feds can withdraw funding. If you work for a town, getting state funding is pretty tough unless you've got connections or it has a major impact on a state road. Most town jobs I review are a mix of local and federal funding. Town standards are stricter than state standards, overall. Providence, for example, requires granite curbing on its roads (which runs about $30/linear foot!).
|
|
#
¿
Jul 28, 2009 22:27
|
|
|
KOFT posted:Have you ever come across three lane roundabouts? What do you think about them? I have, sort of. It was this one:  Hoooooooly poo poo. My French boss drove through that thing blaring opera music like it was nothing, but I think I'd have broken down and cried if someone forced me to drive there. It's simple, sure, but the sheer number of cars surrounding you makes it tough to concentrate. I'm a good driver, too! If you stuck that in America, there'd be so many accidents that they'd have a junkyard in the middle to store that day's totaled cars. I know three-lane roundabouts are smaller and calmer than that, and they're good for high-volume intersections, but I don't foresee them happening in the US anytime soon. Americans, me included, just aren't ready for something that requires them to use their brains while driving.
|
|
#
¿
Jul 28, 2009 22:32
|
|
|
|
| # ¿ Apr 30, 2024 03:40 |
|
|
Jarlaxle posted:What's your expert traffic-engineer opinion on the city of Worcester, MA? Because as a driver who occasionally has to pass through there, I loving hate those roads. I don't know exactly what it is about them, but it's the only place in New England I've ever had trouble driving in. Most people I know feel the same way, and I'd like to know if there's any actual reasoning to back us up. Oh god, I don't even want to touch Worcester. I stay away from it on principle. I'm so glad that the 146 expressway got finished, so even if I ever have to go there, I can stay off the local roads. (Also because it means Rhode Island will get a new interstate once I-190 takes over 146.) Since this is my thread, though... Worcester's a total nightmare, traffic-wise. All of its roads were laid out before automobiles were around, a bunch of the intersections meet at weird angles, things are so built-up that there's no way to fix the roads, I'd be willing to bet that nothing in the city meets modern design standards. Oh well, I've seen worse. At least it's not East Longmeadow, right? 
|
|
#
¿
Jul 28, 2009 22:38
|
|