By design, certain blocks change their availability to arriving entities depending on the circumstances. For example,
A queue or server accepts arriving entities as long as it is not already full to capacity.
An input switch accepts an arriving entity through a single selected entity input port but forbids arrivals through other entity input ports.
Some applications require more control over whether and when entities advance from one block to the next. A gate provides flexible control via its changing status as either open or closed: by definition, an open gate permits entity arrivals as long as the entities would be able to advance immediately to the next block, while a closed gate forbids entity arrivals. You configure the gate so that it opens and closes under circumstances that are meaningful in your model.
For example, you might use a gate
To create periods of unavailability of a server. For example, you might be simulating a manufacturing scenario over a monthlong period, where a server represents a machine that runs only 10 hours per day. An enabled gate can precede the server, to make the server's availability contingent upon the time.
To make departures from one queue contingent upon departures from a second queue. A release gate can follow the first queue. The gate's control input determines when the gate opens, based on decreases in the number of entities in the second queue.
First port that is not blocked mode
of the Entity Output Switch block. Suppose each entity
output port of the switch block is followed by a gate block. An entity
attempts to advance via the first gate; if it is closed, then the
entity attempts to advance via the second gate, and so on.
The Entity Gate block offers these fundamentally different kinds of gate behavior:
The enabled gate, which uses a control signal to determine time intervals over which the gate is open or closed
The release gate, which uses a control message to determine a discrete set of times at which the gate is instantaneously open. The gate is closed at all other times during the simulation.
Many models follow a gate with a storage block, such as a queue or server.