engineerboy6561

As someone who does product development and engineering design for a living, I'd say that most modern subway car systems are incredibly complex in their own right, and the car itself has to make sure that all of those systems play nicely with each other under a wide variety of conditions. Under ideal conditions each system comes with a datasheet and a specification that describes how it will behave under normal operating conditions, and the car manufacturer will have specified everything so that the most extreme conditions the subway car is likely to see in service fall under that heading; under absolutely ideal conditions they get the car in, boot everything up, run it for a few miles, and turn it loose on passengers. In practice, it's almost never that fast; a few weeks back I had a simple test that would have taken 20 minutes if everything had been put together properly and more realistically half a day to a day to do, and it took us five and a half days because of a bunch of harness assembly issues and connection mistakes that were a mix of technicians misreading things and a couple errors on cable drawings. The MBTA is having a hell of a time right now with their new equipment, for a variety of reasons (this is CRRC Changchun's first time working in America, and in order to actually make good on their low bid for the cars themselves they're trying to hire assembly techs at like $17 an hour, so of course you'll get nonsense and mistakes), and even if the car assembler is perfect it's possible to get weird bugs because something in the subsystem specifications the assembler gave the subcontractor was either slightly off or unspecified because you didn't think it would be an issue but it absolutely is and now you're in trouble. Like the Class 180s in the UK that Alstom built were absolute shitshow lemons, even though the family of trains (the Coradia) they were derived from runs perfectly fine on the continent; the issue was that Euro trains are significantly wider and taller than British trains, and so they were trying to cram a full suite of underfloor systems under something like 10-20% less floor area, and this resulted in some really weird design decisions. That sort of thing is also why the Class 220/221/222 units in the UK intermittently smell of sewage; space constraints led Bombardier to put the toilet tank right near the exhaust (which heats up the contents and makes them offgas) and put the pressure relief valve right near the HVAC inlets (so sewer gas goes into the passenger cabin). Like if you want trains to not require a ton of testing the way to do that is to overdesign and overbuild most subsystems (which shifts the design and testing risk onto the subcontractors who make those subsystems), and then once you get a subway car working with a given configuration change as little as possible in future orders; you lose out on all the benefits that new technology will offer you but you wind up being able to get away with far less testing on introduction, and your operators and maintenance people only really need to understand the quirks of one particular set of systems (to my understanding that's how the SMEEs were, and the NTTs are that way to a lesser but still nontrivial extent). I'm biased as an engineer who likes designing state of the art stuff; I'd love to see aluminum honeycomb construction, Jacobs bogies, plug doors, and silicon carbide traction inverters come to NYC, but I don't want to minimize the amount of work needed to make that happen.

I took a look at this and I have a lot of thoughts; unfortunately I only have half a braincell because work was a lot this week, so it may be the weekend before I can put together a critique.

I'll check it out when I get home tonight and offer feedback then

Honestly, if we wanted to use that as a primary route into the city we'd probably want to provide express service along Ft. Hamilton Parkway (bring back the , send it to Coney Island, send the to South Av via the SIE and Ft Hamilton Parkway local, and then send the to St. George via the old South Shore ROW and Ft Hamilton Parkway express). The would make all current Culver local stops to Coney Island, and the and would make the following stops starting at Bergen St: Bergen St (express, joins the ) Carroll St (local) Smith-9th Sts (local) 4 Av-9 St (local) 7 Av (express) 15 St-Prospect Park (local) Fort Hamilton Parkway (local) Church Av (express) 15 Av/39 St (local) Ft Hamilton Pkwy/39 St (local) Ft Hamilton Pkwy/New Utrecht Av (express, transfer to ) Ft Hamilton Pkwy/ 50 St (local) Ft Hamilton Pkwy/55 St (local) Ft Hamilton Pkwy/62 St (express, transfer to ) Ft Hamilton Pkwy/Bay Ridge Av (local) Ft Hamilton Pkwy/77 St (local) Ft Hamilton Pkwy/86 St (local) Bay Ridge/95 St (express, transfer to ) New York Av-Ft Wadsworth (express, branch point) trains would split off and turn north, stopping at Fingerboard Rd, Hylan Blvd/PS 13, Clifton, Stapleton, Tompkinsville, and St. George, while the would continue on in the median of I-278, stopping at Hylan Blvd (connecting to a new stop on the SIR), Richmond Rd, Manor Rd, Bradley Av, Victory Blvd, South Av, and Gulf Av. That configuration would probably still allow for fairly fast travel times between SI and Manhattan (and if we wanted we could add a mix of park-and-rides and new housing at the stations along the SIE to get people out of their cars).

I just read through that proposal and it's actually really interesting and I'd be down to do it; I'd combine a tunnel under Ft Hamilton with an extension of the if things were interlined that way; the other thing I would totally do to replace the missing Brighton express service is send a 2 Av train to Coney Island via Broadway (Brooklyn) and Brighton local while the take Brighton express. This would be fun as hell to actually map out, and at some point when my job calms down somewhat I would really like to sit down and pull together a few completed proposals (including construction phasing and rough cost estimates) to throw up on here for feedback; there are enough degrees of freedom here to make things interesting.

Honestly I wouldn't do a tunnel directly under the Bay to St. George because it's going to be extremely long and expensive and still leave a forced transfer at St. George, and I feel like the 2 Av local would be better off being connected to Court St as a Fulton St local than running one stop to Staten Island. I'd probably be more likely to suggest extending the under the Narrows and then along the northern half of the old East Shore Branch to St. George, and maybe if I was feeling crayon-ish an extension of the under 15 Av, 39 St, and Fort Hamilton Parkway, then under the Narrows and in the median of the Staten Island Expressway as far as South Av.

Honestly that makes a lot of sense, and it's a pity that 67' didn't work out; longer car lengths make more room for things like transverse seats, and the door widths make it even more so. Honestly, what I'd love to see is an aluminum-based trainset with swing plug doors so we could have lower weights and more window space again.

That's fair; the assumption I'm making about having TfL-style moquettes is basically that the biggest threats facing public transit seats are writing and bodily fluids, and so a seat that's impervious to bodily fluids and designed with a pattern against which graffiti is frustratingly hard to see would be fine.

My initial thought was to have three 3-car sets arranged as triplets the way the R110B was arranged, but since apparently the Eastern Division can take eight 67' cars it might make sense to do sets of four and then sets of five (so that sets of four could be used anywhere and then sets of 5 would do non-Eastern Div work. You can have a 60' car with a more comfortable layout, though the way I would probably do it is similarly to the Class 345s, putting a couple bays of 2+2 near the ends of the set while leaving most of the middle as purely longitudinal seating. As far as the moquettes are concerned, I meant with an upholstered seat similar to what the Tube uses, so it would be soft to sit on. Part of why I pitched the hydrophobic coating is that they actively repel most water-based liquids (which includes most kinds of food as well as other bodily fluids that can end up on train seats if someone's having a bad day). Ideally you could design the seat covers to come off in a minute or two, and then once a month or so the covers get taken off and washed while the train's laying up in the yard.

That ridership discrepancy is weird, and yeah headway issues would make sense as a thing that keeps people away. On weekdays the is half an hour on the timetable from Bedford Park to Bryant Park, and probably 31-33 minutes IRL during rush, whereas the is 41 minutes on paper but probably closer to 45. Yeah, at night and on weekends the headways on the and could get weird and missing a train could be a serious issue, whereas there's always a coming. Like the fastest way for me to commute during rush back when I lived in the Bronx and worked in Newark would probably have been to get the Bx1/2 to the 7:25 at Bedford Park, which would let me make the 8:13 commuter rail train to the 1 bus to the job, but if I missed that train it would be a 10-12 min wait.

I don't know; I'm honestly assuming that Concourse takes a lot the passengers bound for the west side in that area, while Jerome takes folks going down on the east side plus some going farther down; if I had to guess it might also a frequency thing; the runs something like 13-15tph during rush, while the and combined run about 16, and in the mornings ten of those tph are express from Tremont to 145 St. Then again, during rush the and the are timetabled to both take 41 minutes from Bedford Park to their respective 42nd St stations (though in practice the usually winds up slower), so it becomes a question of where pax are going and whether the frequency is an issue for them.

Yeah; 75' stuff can't really run on the BMT Eastern Division, and I'm also not quite sure how something would work with 75' length and Jacobs bogies (on trains without Jacobs bogies a 75' car is probably 65'-70' over truck centers, whereas a 75' car with Jacobs bogies would be actually 75' or so over truck centers; such a design would probably work better with 67' car lengths like the R110Bs had because the actual distance over truck centers would be similar). Honestly, making the seats uncomfortable doesn't make the homeless people go away or get off the train; safe injection sites, better shelter infrastructure and cheaper housing does a lot of that. I'd think that the London moquettes are fairly vandalism-resistant (at least they claim they are; I'm not sure how they would stand up to someone with a knife, but my impression is that not many people use knives vs drawing implements to mess up a seat, and a stock of spare covers could probably used to mitigate that. Coat the moquette in this stuff and you'll find you can basically spray it down and it'll clean really easily: https://www.flotechps.com/advanced-coatings/super-hydrophobic-coatings/)

Honestly same; I'd really hoped to see more 75' cars with some longitudinal seating in the vein of the R68s on some of the longer express routes; like I'd love to see something Class 345-inspired on the , , and other lines with long express segments and long end-to-end runtimes (I loved the R44s and R46s on the in high school when I had a research internship at NYU Polytech because I'd just grab a forward facing seat at Jay St-Metrotech and knock out all the way to 207 St before grabbing the bus up to Riverdale and walking the rest of the way home); like the Class 345s are lower-capacity because they're only 9'2" wide, but you could widen them up to 9'10" to fit the BMT and shorten them to 600'; their platform height is aligned with NYC's, and they look pretty damn comfortable to ride in while still having decent capacity. Images supplied for reference: Like these were some older renderings I'd done of what I had hoped the R211s would look like (the interior was modeled on a fusion between the NTT and R46 interior, while the exterior was modeled on the Acela intermediate cars):

Saaaame; for me watching the R62s and R68s go is going to hit different. I was only 6-7 when the Redbirds retired, and I was only able to ride them once or twice (grew up between the and the , mostly took the because my parents had memories of being on the during the 80s, and I only really started riding the in high school when it was all R142/R142As)

Time bloody flies; it's hard to believe that 2010 was thirteen years ago

That's fair; it would beat the to the for the Pelham-to-core market, which would be nice. I don't think it would do much of anything for the Jerome-to-core market, because the entirety of Jerome already has a one-seat option into the core ( the and from 161 St up, and the at 149 St). Re: Queens, that's why I keep repeating plans for a Northern Blvd 4-track trunk feeding into lower 2 Av, which in turn replaces the Jamaica Line with a four-track trunk offering 30 express tph from SE Queens into downtown/Midtown; it lets you knock out a whole bunch of birds with one stone. Now, actually doing that within 10-20 years without eating half the city's GDP is going to require a fair amount of organizational change at the MTA, and a whole bunch of neat tricks from abroad (which I've bought up in the proposal thread; the situation is unfortunate, but it's worth pushing for change).

Yeah, looking at this study about half their proposed alternatives were drawn with a crayon; I do wonder if that was somewhat deliberate, though. If you know what option you want to go with but you've been told to study all the options, then putting it up against a bunch of stupid ideas like converting Lexington Av to B division service in Manhattan or running trains from downtown Jersey City to 125 St or contracting with jitney operators to run supplementary dollar vans along 2 Av basically guarantees that your preferred alternative gets chosen (because it's the only one that actually makes some degree of sense). Now granted, doing that also boxes out other ideas that might make more sense than what you're proposing, but if you're primarily interested in building a cool thing with your name on it that's a feature, not a bug.

That sounds about right; of course part of the problem is that making service better requires a fair amount of expansion if you think about it; otherwise you get a situation like eastern Queens where you have like two subway trunks and they're overloaded as f**k because a ton of people from all over Queens are descending on the first couple of stops during rush. The current subway system doesn't avoid taking people from certain parts of the city into the core; it just makes their trips longer and crowds everything badly enough that people's commutes become ridiculous. If you want service to run better, you have to spread trips out across enough different subway lines that no one trunk gets overwhelmed, but you can't just plop a bypass on the most congested bits of the network and call it quits (which is what the current SAS plan is trying to do). Like if you add a transfer but change very little else (average speed, number of stops, etc. stay roughly the same) then your commute gets longer rather than shorter, and since most of the crowding on Lex is basically crowding from folks on Jerome/Concourse/WPR trying to get to East Midtown you're not going to be able to get many of those folks to change at 125 St for an overall longer commute (which is what it seems like SAS is trying to get them to do).

That's really cool I have a few older Sketchup designs I did a long time ago; right now I've started doing things in Solidworks because it's absurdly powerful and easy to make things in once you get used to using it

This sounds like the sort of terrible plan that's going to either be a giant waste of money or that we're going to wind up spending a shitton more later to fix. Reverse branching at 63 St only works if there are four tracks, and if there's a Queens trunk that's feeding lower 2 Av; that's why all my plans involve four services on 2 Av below 63 St, and then three services go to Queens and one continues up to upper 2 Av where it meets the for the run into the Bronx. If you do that, then 2 Av and Broadway work like 8 Av/CPW and 6 Av, you've fixed the merge at 57 St/7 Av, people traveling across the boundary at 63 St don't all have a compulsory transfer to continue, and you get good operational value for your money. What they're doing now is mostly wasting money on poorly thought-out plans that won't deliver anywhere near the ridership or aggregate time savings per dollar that a proper subway line ought to.

I'd agree that SAS did it backwards; they made these massive identical things and then turned around and made all the parts people didn't actually care about different (which f**ks up maintenance people and increases procurement costs). Like I'm personally somewhat of a fan of minimalist and modernist architecture, so that's going to be reflected in the sorts of things I think look good, but I don't have a problem with the idea of leaving space for architectural uniqueness within a standard design. If you're interested I can look at knocking together a few stations in different orientations with different track numbers to show you what I'm thinking.

Honestly I'd agree; I think for the most part aside from a few exceptions like sending the to Emmons Av via Nostrand and/or the to Kings Plaza via Flatbush Av it would be best to leave the existing IRT system as it is given the combination of suboptimal trackwork (like Rogers and 149 St/Concourse) and lower capacity, and focus on building more B division infrastructure. The idea is that if we build mostly B division infrastructure, especially new full trunks, we're likely to be able to provide more capacity, faster rides on average, and more overall system flexibility; furthermore, there are enough interesting little bits of track and bellmouths available to work with that it's possible to get a great deal of value from fairly short segments of track (like connecting Hanover Sq on the lower Manhattan end of the 2 Av subway to Court St, or rerouting the along a mix of a revised Jamaica Line, the Crosstown Line, a below-ground version of the Franklin Av shuttle, and Brighton local tracks to allow for Brighton to still have access to both 6 Av and Broadway while pulling one line off DeKalb Junction). Like the nice thing about having a bunch of four-track trunks in a bunch of places is that it's possible to have reverse branching that constrains service frequency a lot less than it would otherwise.

A to Gateway Center Mall is simple because it's just a six-block extension of two tracks down Elton St; it's the sort of thing that even the current can't screw up too badly (so I can understand why they're prioritizing it); a subway to Red Hook is meaningfully more complicated and should really come after 2 Av gets resolved in a meaningful way. As far as extending the out to JFK, I don't think you can extend them both without either a nasty merge at Utica or a complete rearrangement of the line from Franklin Avenue east. Livonia Av is only 61 feet wide between buildings; that gets you two tracks and an island platform with about five feet to spare. There's no way you could bring four elevated tracks through there and have any kind of room for platforms, and the elevation drop (a little over 50 feet) just past Utica Av means you have to either go elevated with all four tracks or rebuild the tracks between Kingston Av and Utica Av to slope down at about a 4% grade so that the tracks end up about 30 feet below East 98 Street before continuing along the existing route and then under New Lots Av, the Cypress Hills Houses, and Linden Blvd. You'd only be able to resurface there if you wanted to do the rest elevated (over Linden Blvd and the Belt Parkway). Alternately, you could run four tracks elevated over Rockaway Parkway and Linden Blvd, though you'd either have a really tight turn at that intersection or need to rise an additional 30-40 feet above the ground in order to clear Brookdale Hospital (which you'd need to do; there's no way you'd get away with demolishing it given how many people it serves). That would be about a 6-mile extension, though most of it would be elevated. Like at that point, the service pattern would wind up being something like this: Wakefield-241 St to Flatbush Av, Eastern Parkway/Nostrand Av local Harlem-148 St to Howard Beach, Eastern Parkway/Linden Blvd local Woodlawn to Howard Beach, Eastern Parkway/Linden Blvd express Dyre Av to Gateway Center, Eastern Parkway express/Livonia Av local That also decongests Rogers Junction at the price of Flatbush Av losing its express service; I'm assuming that in order to make this work a fairly large storage facility with maintenance shops is built over the southern half of Long-Term Parking Lot 9 at JFK, to preserve yard access for the in this configuration (either that or you'd need to knock down most of the block bounded by Stanley Av, Linwood St, Wortman Av, and Elton St (that would allow trains to exit the yard onto the new tracks to Gateway Center, and then a new track pair over Linwood St would swing up and over the Amazon warehouse to connect to the local tracks on Linden Blvd. This would be interesting from a redundancy perspective (if the goes down you can still get into the city), but I'm not sure you'd see much time savings over the on the . The assumption I'm making here is that the stopping pattern on the extension from Utica is Rockaway Bl/Rutland Rd (local) Rockaway Blvd/Kings Highway (local) Rockaway Blvd/Church Av (express) Linden Bl/Rockaway Av (local) Linden Bl/Van Sinderen Av (express, transfer to ) Linden Bl/Pennsylvania Av (local) Linden Bl/Van Siclen Av (local) Linden Bl/Ashford St (express, transfer to ) Linden Bl/Euclid Av (local) Linden Bl/Eldert Ln (local) Linden Bl/81 St (local) Conduit Av/Cross Bay Blvd (express, change for Q52/Q53 SBS) Howard Beach/JFK Airport (express, change for ) The makes 11 stops between Fulton St in Manhattan and Howard Beach, and 17 between Port Authority and Howard Beach. Under this configuration the also makes 11 stops between Fulton St and Howard Beach, and 14 stops between Grand Central and Howard Beach. Based on that metric I'd expect the to be about even with the into downtown, while a bit faster to the midtown core than the . The is currently scheduled to take about 51 minutes to make the run from Howard Beach to Port Authority, while the is scheduled to take about 30-31 minutes to make the run from Grand Central to Utica Av; I could see the making the additional run in 10-15 minutes, offering a 5-10 minute time savings over the .

I have complicated feelings about standardization; I'd argue that the ideal thing to do would be to internally develop a base set of standardized station components that are designed to be somewhat modular for easy adjustment to different construction methods and construction depths, but all use standard elevator designs, escalators, lighting, furniture, etc. and all have similar overall MEP layout. That way when you build a new line you can basically pick one or more combinations of construction method and depth (which may well vary along the length of the line depending on geological conditions, topography, choice of street to run under, etc.), decide how many tracks you want to have and what platform arrangement makes sense for each station, and then select the corresponding off-the-shelf design. I'm imagining (and people with more experience can correct me if I'm wrong) that we could do things like have a standard station access box with a given number of elevator shafts and escalators, adjustable in height from 15ft to 30 ft (elevator bank against the back wall, combined stair and escalator flights on the sides), and then for stations deeper than 30 ft we could just stack the boxes going down until we hit platform level, with the lowest level of the access box above track level serving as a mezzanine); similarly, we could basically standardize a 60' track and platform segment (track placement, platform width and placement, lighting, platform furniture, etc.) and just tile segments to produce full station boxes (540' for A Division stations, 660' for B Division stations). The idea would be that once we decide on a line alignment we can standardize construction enough that we wind up with a smaller number of vendors supplying identical or at least very similar parts and equipment in higher volumes, which should allow for price breaks during construction and also massively ease long-term maintenance and upkeep (if all the escalators and elevators are from the same product family by the same manufacturer you need a much less varied stock of spare parts and you can focus technician training on the one to two types of equipment they're actually going to work on). Furthermore, doing the design that way means that we could focus MTA engineering staff on designing a small number of station topologies (and then optimizing them for lower costs, higher ease of infrastructure maintenance and upkeep, higher pax throughput, and easier accessibility).

Just read through Section E and that's an interesting idea; it's also frustrating that the IRT has a bunch of spots that weren't really well designed to maximize throughput, but I guess that's what happens when you're building a subway system for the first time.