Monday, July 29, 2013

Fundamentally Redesigning Metro Train Travel

Increasing the speed of  metro trains is really expensive. The technological challenge of quick yet comfortable acceleration are to be overcome & accordingly safety infrastructure needs to updated. Engineers have been working on this challenge for many years now. However all efforts seem to be focused on having faster trains.

Lets start afresh, with a blank sheet of paper. If we step away and analyse the time taken for a typical journey. The typical graph would look like this:

Typical Travel Time Distribution at Peak Hours in Delhi Metro

Data analysis is such a bitch! You can twist it the way you want, to take out the juice you want. Here, I intentionally created this category of acceleration and de-acceleration. You will see later, why :)

The other thing to be observed is that in most of the stations, the number of people who have to board and de-board are really small. Taking case of Delhi Metro, for example from HUDA City Center to Civil Lines, there are only 3 stations in which large number of people get in and get out. Obviously, they are Central Secretariat, Kashmere Gate and Rajiv Chowk. The intersection stations. In all other stations (Qutub Minar, Hauz Khas, Green Park, INA etc.) only a small % (may be maxima of 10%) of commuters board/de-board the train.

So, based on these observation and some chai-sutta session brainstorming with my colleagues at Mechartes, here is the proposed design:

  1. Rails: Each train can be designed such that it is running on 3 set of parallel rails. Apart from central rails (which is meter gauge), there are two more thinner rail tracks (0.25 meter gauge) on the sides. 
  2. Coaches: All the coaches in the train are wider than current coaches. They are divided in to three compartments. The central compartment has 80% area, on the sides are compartments each with 10% area. These two side compartments are detachable, just like jai-viru motor-bike in "yeh dosti, hum nahi chodenge" :)
  3. Train: The train moves at constant velocity. Once it starts from Huda City Center, then it stops only at Central Secretariat. In all the remaining stations in between, it keeps moving at constant velocity. 
  4. Stations: The intermediate stations are designed such that there are side compartments always parked there. Those side compartments have their own engine and they accelerate when the train is about to reach the station. When these compartments achieve velocity same as train velocity, then it get attached to central compartment. 
  5. Commuters boarding the train : They come in and start filling up the side compartments already parked at each station. The gates of these compartment automatically close, when the train is about to arrive or if the compartment is completely filled (whichever is earlier). When the train is about to arrive at station, then these side compartments start accelerating. The acceleration is intelligently tuned so that when they achieve the speed of the train, the train is running parallel to them in same speed. Then without any problem the side compartment gets attached to the main train. The separation door between the main compartment and side compartment is opened and commuters can freely move in to the central compartment. 
  6. Commuters de-boarding the train : They move to the side compartment of their coach. (Incidentally, there will be only one of the two side-compartment attached at all times. So, there is no confusion on which side I've to go to de-board.) When the station is about to arrive, the separation door between side compartment and central compartment gets closed. Then the side compartment gets detached from the central compartment and it slowly starts de-accelerating. This is again intelligently tuned, such that it comes to complete halt at the station. Then commuters can de-board, and new commuters can board this side compartment and wait for next train to arrive. 
Voila! Chai khatam, problem solved :)  
Think about this one. Although this sounds very sci-fi, but it will work. The current state of technology is more than sufficient to realize this new design of metro system. I dare say, this would be the fastest and most energy efficient metro system of the world. 


10 comments:

Shishir Gupta said...

Thanks Dude. As always you are the only one who I can convince of my ideas :)

Shishir Gupta said...

@Jaini, with your permission, my replies are as follows :)

1A. Relative velocity between the side compartment and main compartment is Zero. So they are not really hitting each other.

1B. The change in mass (and hence momentum) will cause problem. It will be better if side compartments are attached, one at a time. With sufficient time in between for train to absorb the additional mass.

2. The system will consume less energy, since most energy is required for acceleration & deceleration. Hence this cost can be easily recovered. The speed and hence the capacity will increase. Revenues could be much higher. I've not done detailed cost analysis, but I've a feeling this would be hands down winner in cost recovery analysis.

3. With my experience, 10% has sufficient factor of safety. What I've observed is that this number is usually in the range of 5% for intermediate stations. Do notice, that at high mass-flux stations like Rajiv Chowk, the train will actually stop and mass will be exchanged.

Shishir Gupta said...

Yo!China Rocks. They will build it too, pretty soon. Its so much like: "Elementary my dear Watson". I cant see a flaw in this design.

Dizzy said...

Brilliant! This actually seems very doable, patent kara lo :D!

Shishir Gupta said...

With regard to patents, I'm a non-believer :)

Abhishek Jain said...

1A. The velocity will be the same with certain margin of error. The question is if the train can run smoothly within the error bounds. Let's assume there could be a small 0.5km/hr difference in speed. What then?

Abhishek Jain said...

1A. I did a ballpark calculation...weight of a 10 car metro train could be approx. 400 tonnes. Lets say, the side train weighs as much as 100 tonnes...For, even 0.1km/hr difference, the momentum transfer is going to be (400-100)X0.1 = 30 tonnes-km/hr. I wonder if rubber pads can absorb this kind of energy repeatedly on a single day..It will be useful to find out whats the best out there in the industry.

3. Well, if the influx/outflux is higher, the driver decides not to move and let the big train arrive and stop first, joins and they move together...basically follow the traditional way.

Dizzy said...

lol...agreed, but something can be developed based on this (in all probability something must have been done already in some other part of the world). Many simple solutions come to mind regarding the impact problem, maybe a magnetic field cushion? or direct impact with rubber like energy dissipation material? Doesn't seem impossible, but details might be tricky

Shishir Gupta said...

Yup, it sounds very feasible. I don't think it has been implemented so far anywhere in the world. May be soon. BTW tomorrow Elon Musk is going to reveal design of this new concept train called Hyperloop. I'm so looking forward to it. Read about it, that too is very interesting.

Anupama said...

Very very interesting concept!
It may be possible to achieve what you have envisaged with only 2 sets of parallel tracks and only one side compartment being attached to the main train at any given point. As the side compartments will be the ones detaching and decelerating OR accelerating and then attaching on the side, track only one side is required.
This can be achieved because while the compartment which is to attached will start accelerating and hence leave the station before the compartment that is to detach has even detached from the incoming train, to decelerate and then stop at that station at the same place the compartment that left was standing.