Automated computer control chapter 2 - The basics of blocks.A vital part of automated computer control, is for the computer to know which train is where. Optimally, the locomotives would continously send a signal to the control center with such information as decoder number, speed, direction and which block its in. However, bi-diretional digital isn't that far yet, and even if it could send back most of these variables, determining which block a locomotive is in is difficult.
Instead, we need a system that detects on which part of the layout the train is. There are a variety of ways to do this, including using reed switches and magnets or infrared signalling. However, in my opinion, the most reliable way is block control.
So what is a block? A block is nothing more then the stretch of track between 2 signals. Those signals don't have to be physical signals, they can be imaginary. Prototypical railways generally use a block system to track their trains as well.

The general idea is that a layout is divided into multiple blocks. Depending on the system used, the blocks are either isolated on 1 or 2 rails. In most cases for 2-rail systems, only 1 rail is isolated. The reason for the isolation has to do with detecting when a block is occupied or not. More on that later.
To put everything into context, let's use an oval as an example. An oval can be divided into 4 block. The 2 straight sections are 1 block each, and the 2 180 degree curves are 1 blocks each.

Purely theoretically, 3 trains can run on an oval that's divided into 4 blocks without ever running into each other. The problem is that the moment a locomotive enters a new block, it will need to start decelarating because the next block is occupied. Its very similar to driving while in a traffic jam. You can only move when the car in front of you moves, and he can only move when the car in front of him moves. Imagine a traffic jam that runs in a circle, and you'll get a good idea of the problem.
Running 2 trains is much better, but if 1 train is faster than the other, you will still have the problem that the faster train will regularly need to slow down.
The most optimal solution would actually be running only 1 train. Generally, a train will want to have at least 2 free blocks in front of it. That way there's little chance it needs to slow down, unless you have a fast training running behind a slow train. For example, an express train would be faster than a freight train. (That situation can be avoided though, but adding either sidings so the express train can overtake the freight train, or make sure the express train has priority over the freight train at a train station or other stop.)
So, dividing in oval into blocks is really rather pointless, but an oval is something everyone will recognize, and the blocks are easy to set up ;)
Below is a flash movie that shows the 3 different setups. With 1 train, there's no problem. With 2 trains, there's no problem either because both trains run at the same speed. With 3 trains however, trains need to stop a lot, either because they sometimes run slower or faster, or because the length of some blocks is much shorter than others. Of course, the step with 3 trains is exaggerated, but it shows the problems you get when running too many trains at a time.
http://www.jr-chiisai.net/_div/computer_control/chapter02/03-blocks.swfNow that we have a basic idea of what a block is, the next chapter will talk about some more advanced topics related to blocks, including brake and stop sections, occupancy and feedback, and deciding where to install blocks.