An unobstructed passage through the coupled pair, rather than a valve mechanism sitting in the bore.

A conventional quick-disconnect has a spring-loaded poppet in each half that seals when disconnected and is pushed open when the halves engage. Even fully open, that poppet, its spring and the seat remain in the flow path. The effective flow area is less than the nominal bore suggests, and fluid has to divert around the obstruction.

Full-flow designs either omit the valve altogether, giving a clear straight passage, or use a mechanism that retracts clear of the bore when coupled.

The practical effect is measured as pressure drop at a given flow rate, and the difference between a conventional and a full-flow coupling of the same nominal size can be substantial. Compare couplings on published flow coefficient or pressure drop curves rather than on nominal size, which tells you very little.

In hydraulic systems it becomes heat and wasted power; in cooling circuits it reduces flow; in transfer lines it slows the job.

A hydraulic circuit running continuously through a restrictive coupling converts that pressure drop into heat in the fluid, every second the pump runs. The energy comes from the prime mover, so it appears as fuel or electricity consumed and as a hotter system needing more cooling.

On a liquid cooling circuit the flow rate is what carries the heat away. A coupling that throttles the circuit lowers the flow and the equipment runs hotter, which is usually diagnosed as a pump or heat exchanger problem before anyone suspects a coupling.

On compressed air, restriction at couplings is one of the most common causes of tools that never quite deliver rated performance.

The cost is continuous and invisible, which is exactly why it is worth designing out at selection.

Some do and some do not, and this is the key selection question.

Non-valved full-flow couplings give the clearest bore and the lowest pressure drop, because there is nothing in the passage at all. But disconnecting one spills the line contents from both halves and lets air in. They suit lines that are depressurised and drained as part of the procedure - transfer lines, wash-down, systems where a small spill is acceptable and easily contained.

Valved full-flow designs retain automatic shut-off on one or both halves while keeping restriction low, using flat-face or sleeve mechanisms that clear the bore when coupled. They cost more and give slightly more restriction than a non-valved coupling, but far less than a conventional poppet type.

Decide from the fluid and the consequence of spillage. Hydraulic oil, chemicals and anything hazardous point to valved; water and air in a controlled area may not.

Generally not without a design intended for it, and attempting it is both difficult and hazardous.

Pressure inside a coupling half acts on the sealing area and produces a force pushing the halves apart. On anything but small sizes and low pressures that force quickly exceeds what can be overcome by hand, so the coupling simply will not engage.

Worse, a partially engaged coupling under pressure can blow apart, releasing fluid and becoming a projectile.

Where connection under pressure is genuinely required - on hydraulic quick couplers for attachments, for instance - purpose-designed connect-under-pressure couplings exist, using threaded sleeves or screw-together designs that mechanically overcome the separating force.

For everything else, depressurise before coupling. Trapped pressure in a disconnected hose is a common cause of couplings that will not reconnect, and the correct response is to relieve it, never to force the connection.

By flow rate and acceptable pressure drop, not by matching the port or hose size.

The temptation is to fit a coupling with the same nominal size as the line, on the assumption that it will not restrict anything. But nominal size describes the connection thread, not the flow area through the coupling, and two couplings with identical nominal size can have very different flow capacities depending on the mechanism inside.

Work from the design flow rate, look up the coupling's pressure drop at that flow from the manufacturer's curve, and check it against what the system can tolerate. On a circuit with several couplings, add them up - four modest restrictions in series is a significant loss.

It is common for the correct answer to be a coupling one size larger than the line, with adapters either side. That is cheaper than the energy a restriction wastes over the equipment's life.