General MBTA Topics (Multi Modal, Budget, MassDOT)
- Thread starter Waldorf
- Start date
Don't text and drive, people.
Brattle Loop
Senior Member
- Joined
- Apr 28, 2020
- Messages
- 1,163
- Reaction score
- 2,077
They're probably groaning the same thing at the NTSB. As long as the Green Line doesn't have anything that can enforce separation and signals, this will continue to be a possibility.
I thought that a contract for a protection system was already signed pre-covid as part of GLT? When does that come online?
Edit: yep. Here. According to that, 2024, but will this collision speed up implementation, or is this one of those things that just take time and no good way of doing it faster? https://www.mbta.com/projects/green-line-train-protection-system
Saw this on Reddit... damage to the train that hit looks pretty bad. Looks like it hit at full speed (for the B line)
HelloBostonHi
Senior Member
- Joined
- Apr 17, 2018
- Messages
- 1,422
- Reaction score
- 3,685
Certainly looks like it hit faster than it should have been going for that section of track. It's a 10mph limit and it's in an active construction zone where they're building the new platforms...
Last edited:
Brattle Loop
Senior Member
- Joined
- Apr 28, 2020
- Messages
- 1,163
- Reaction score
- 2,077
I'm no expert, and I don't recall which thread it was in, but someone (F-Line to Dudley maybe?) at one point went through how complicated a problem it was trying to graft a train protection system onto the Green Line's ancient signal system without utterly ruining its capacity, so I kind of doubt it's something that can really be sped up very much. At least the fact that they have a contract and are moving forward is an improvement over some of the past collisions' responses, because at this point at least they can say they're working on it.
BostonTrainGuy
Active Member
- Joined
- Sep 18, 2020
- Messages
- 159
- Reaction score
- 142
Looks like the Bredas are stronger than the Kinkis. They are newer of course so the buff standards might have been increased. This is certainly not as bad but it is reminiscent of the horrible fatal in 2008. Seems like a lot of deformation for the speed which I would guess was much less than the 40 mph Riverside Line.
Looking at the reports and pics online, it looks like the 8 hit the back of the 7. If it was the other way around, that operator would not be walking to an ambulance if that was the correct operator shown in one of the reports. He was lucky although possibly now out of a job. The 2008 crash was a 7 into a 7.
Looking at the reports and pics online, it looks like the 8 hit the back of the 7. If it was the other way around, that operator would not be walking to an ambulance if that was the correct operator shown in one of the reports. He was lucky although possibly now out of a job. The 2008 crash was a 7 into a 7.
Last edited:
F-Line to Dudley
Senior Member
- Joined
- Nov 2, 2010
- Messages
- 9,196
- Reaction score
- 9,002
Plus the collision avoidance overlay would work different on the surface where the signal system is not in effect. It's essentially one system, two different implementations...so Subway+GLX+D are likely to get deployed on a different timetable than B+C+E.
F-Line to Dudley
Senior Member
- Joined
- Nov 2, 2010
- Messages
- 9,196
- Reaction score
- 9,002
No...the crash management profiles of the two makes aren't much different. The difference in damage profiles is almost entirely the result of one train having its brakes fully engaged while the other didn't. The Type 7 was the rear car on a stopped consist. Because all brakes on 2 cars were engaged, the forces were borne entirely by the crumple zone because none of it could disperse forward. The Type 8 was moving and didn't have its brakes engage, so there was a pronounced 'jerk-back' across the consist dissipating the forces instead of the crumple zone having to bear that work. It's done intentionally that way, as the rear operator cabin is always unoccupied (therefore you want the crumple zone to be doing the work), whereas the front cabin on the moving car is occupied and you don't want the operator's life endangered by it crumpling.
Had the roles of who was stopped vs. moving been reversed, so would the damage profiles.
EDIT: For further illustration...the Type 8 that was leading the stopped consist sustained no major damage because the brakes throughout the consist + the crumpled rear end of 3705 took all the forces. However, on the moving set the leading Type 8 (# yet unknown) sustained articulation/drawbar damage and the trailing Type 7 (3694) took some head-end damage from the coupler gear smashing together and laterally displacing. That's all the 'jerk-back' forces being distributed throughout the moving set from the brakes not being set. They felt whiplash for lack of a better word, but no potentially lethal crumpling. The stopped set was all-crumple, zero-displacement.
Last edited:
- Joined
- Dec 10, 2011
- Messages
- 5,599
- Reaction score
- 2,707
Will part of the solution be the train equivalent of my car's Foward Collision Warning (FCW) and Automatic Emergency Braking (AEB)? FCW would go a long way (it warns (light & sound) when it calculates closing speed vs stopping distance and alarms if impact will be in fewer seconds than the time that it would take for hard braking to stop the vehicle).
F-Line to Dudley
Senior Member
- Joined
- Nov 2, 2010
- Messages
- 9,196
- Reaction score
- 9,002
The FCMB presentation on it from May 2020 has some simplistic diagrams about how the overlay will work. It's a modified form of Positive Train Control using short-range radio pings to determine distance. Integrated with the signal system on the Subway + signaled branches. Somewhat reduced form on the mixed-traffic B/C/E (and signal-less Mattapan) where the trains detect the outer 'envelope' of each other's ping range for collision avoidance and there's some tie-in to the prioritized traffic signals. Trains will also have hazard-avoidance cams/sensors to detect unexpected objects/cars/pedestrians in the ROW, as that'll help make platform passage and the un-sealed ROW's on the mixed-running branches much safer. No detail provided yet on how much operator leeway will remain (a key point for sustaining Central Subway traffic density and clearing platforms faster), and no install sequence for subway or which branches specced yet.
BostonTrainGuy
Active Member
- Joined
- Sep 18, 2020
- Messages
- 159
- Reaction score
- 142
Very interesting. It sounds like you know of which you speak. Thanks for the insight.
F-Line to Dudley
Senior Member
- Joined
- Nov 2, 2010
- Messages
- 9,196
- Reaction score
- 9,002
Per NETransit. . .
Both Type 7's involved in the accident have been wrecked beyond repair. 3705 (the stopped car that took the direct hit) has a severely crumpled B-end, and 3697 (the trailing car in the consist that did the hit) has A-end frame damage caused by a telescoping coupler. Type 8 3894 (the impacting car) has unspecified collision damage, and its rear coupler was responsible for 3697's A-end damage. The other unspecified Type 8 that was the lead car on the hit consist was apparently undamaged and is still marked active (though it's very unlikely to have taken any revenue turns since the accident, and will probably be impounded for a bit for the NTSB).
The 3600's/1986 and 3700's/1997 lineages of Type 7's are mechanically different enough that they can't cross-mix the salvageable A-end from 3705 and salvageable B-end from 3697 into a remated new car. They'll be stored as parts donors unless/until some other Type 7 wrecks serve up other remating opportunities. A handful of too-far-gone-for-repair 3600's and 3700's were scrapped for parts sources at the end of the rebuild program, but apparently none of the salvaged body parts from that stock are any match for what needs repairing on these two.
There were already 8 Type 8's out-of-service long-term awaiting parts before this, needing a T procurement for midlife refreshes for them to be able to run again. 3894 makes that 9 cars, though it should be repairable since there are extra Breda body shells laying around. Means there's getting to be quite the mismatch in low-floor vs. high-floor cars available for service as there are only 83 active Bredas left (with none of the sidelined expected back soon) vs. 97 active Kinkis (+3 in-shop for short-term repairs that, post-rebuild, have not been taking all that long for turnaround).
Very costly fuckup so far.
Both Type 7's involved in the accident have been wrecked beyond repair. 3705 (the stopped car that took the direct hit) has a severely crumpled B-end, and 3697 (the trailing car in the consist that did the hit) has A-end frame damage caused by a telescoping coupler. Type 8 3894 (the impacting car) has unspecified collision damage, and its rear coupler was responsible for 3697's A-end damage. The other unspecified Type 8 that was the lead car on the hit consist was apparently undamaged and is still marked active (though it's very unlikely to have taken any revenue turns since the accident, and will probably be impounded for a bit for the NTSB).
The 3600's/1986 and 3700's/1997 lineages of Type 7's are mechanically different enough that they can't cross-mix the salvageable A-end from 3705 and salvageable B-end from 3697 into a remated new car. They'll be stored as parts donors unless/until some other Type 7 wrecks serve up other remating opportunities. A handful of too-far-gone-for-repair 3600's and 3700's were scrapped for parts sources at the end of the rebuild program, but apparently none of the salvaged body parts from that stock are any match for what needs repairing on these two.
There were already 8 Type 8's out-of-service long-term awaiting parts before this, needing a T procurement for midlife refreshes for them to be able to run again. 3894 makes that 9 cars, though it should be repairable since there are extra Breda body shells laying around. Means there's getting to be quite the mismatch in low-floor vs. high-floor cars available for service as there are only 83 active Bredas left (with none of the sidelined expected back soon) vs. 97 active Kinkis (+3 in-shop for short-term repairs that, post-rebuild, have not been taking all that long for turnaround).
Very costly fuckup so far.
They've historically operated with about a 20-25% operational spare ratio, excluding long-term OOS cars. 97 Type 7s, 83 Type 8s, and 24 Type 9s effectively gives them 95 pairs. Pre-COVID it required 73 pairs to operate peak, and the extension will require about 5 pairs, so they stay in that ballpark. I'm not that worried about car shortages in the short-term. For better or worse their fleet planning accounts for these crashes happening.
Of course the best time to have done a mini-overhaul on Type 8s would have been 2019-2020 while they were flush with cars with the Type 9 deliveries and lower car demands from station closures.
Brattle Loop
Senior Member
- Joined
- Apr 28, 2020
- Messages
- 1,163
- Reaction score
- 2,077
I think the problem is that having lost another Type 8 hurts more than losing 2 Type 7s because there are more T7s, but now even fewer T8s to pair with them. The Type 9s can't trainline, so the Type 7s are dependent on the T8s in order to have low-floor cars in the train to meet ADA requirements, so losing them hurts more than losing T7s, especially if the T8s reliability gets worse, which isn't outside the realm of possibility given that the T7s were recently overhauled and the T8s, as noted, weren't.
Theoretically there were options available, (that the T said they weren't going to be exercising) but I'm unsure if the exercise window for those has passed given the completion of deliveries.
Sam Staiti
New member
- Joined
- Jun 6, 2019
- Messages
- 27
- Reaction score
- 329
3862 was the leading car of the first train, and it is listed with draft-gear damage from the force of the crash going into its coupler.