hi

while going for the secondary injection,

u need to install secondary injection system seperately.

through tat, the fuel/water tank can be placed on electronic weighing guage--- u can measure the quantity injected for each time.

otherwise go for theoritical nozzle flow calculations.

but tat ll be theoritical.

you need to adapt GDI for water injection and quantity can be regulated by Programmable ECU 

The injector "flow curve" - i.e. the relation between the commanded electric signal for injector open-close duration and the injected mass of liquid is not straight forward. This is due to multiple factors such as the "magnetic"  and "hydraulic" opening and closing delays, nozzle discharge coefficient, etc. You need to measure it (there are dedicated equipment for it)  or request it from the injector supplier (it would normally be measured for n-heptane/gasoline or a non-volatile substitute, e.g.  "stoddard". For water injection, you may re-scale the flow curve based on the density ratio (of water vs. the tested liquid). This is may not be exact, but sufficient accuracy for your application. 

In this case, experiment seems to me to be easier than the mathematical model approach, and you do not need any high-tech apparatus for this.

Install the injector in an engine, or in a pressurized tank (such as from a garage air compressor) filled with air to the desired temperature. Run the injector until you get all your bubbles out. Then run the injector, and measure the water consumption by any means you have available - change in the water tank mass as suggested by Sudhakar, or using a burette and a stopwatch. Divide the consumption by the number of injector pulses and you get quantity of water injected per pulse. Repeat this for a) different injector pulse width, b) different injector pressures, c) different pressures of the charge in the engine or air in the tank into which you are injecting.

What pressurising system are you using? 3-5 bar seems like port fuel injection pressure supplied by a conventional efi fuel pump, not the extreme pressures needed to inject against combustion chamber pressures that direct injection pumps supply.

A conventional port fuel injector (operating at 40-70psig) is measured by supplying the injector with fluid under a constant pressure, then opening the injector with a known pulsewidth at a known frequency. You can then measure the fluid that flows through by any other means (mass removed from tank, collected volume, etc), divide that by the duty cycle (open time percentage) to calculate the injector flow rate. It's not a good idea to just power port fuel injectors constantly as they can burn out under some circumstances. If you are doing this, it's worth checking the injectors at different pulsewidths individually, as they do tend to open and close at different rates even amongst injectors with the same part number. Going through a whole bunch of injectors to find a matched set may be worthwhile if the 5% variation that some can have will affect your experiment.

In theory the measuring of a direct injection unit is the same, but I have not dealt with them before. They do use a much faster reacting injector (stacked piezo crystals) that not only opens almost instantaneously but can also open by varying amounts (not just time), making the control process that much more complicated. It's also a lot harder to supply fluid at the 1500psig that these injectors use. They also use high voltages to trigger them.

If you are trying to directly measure the flow in real time (while you are running your engine) you have three choices:

You can measure the ideal flow rates as I said above, then calculate what was supposed to happen based on what you asked the injector to do

You can measure the flow rate into the injector rail, and the flow rate out of the rail. The difference is the volume injected. Finding a flow meter that can handle direct injection style pressures will be difficult.

Or you can set a flow meter between the tank and the pressure pump, with the return line from your pressure regulator installed between the flow meter and the pump.