Injector Flow

I wanted to ask that whether it's right way how I calculated the injector flow?

So, first I used an Arduino to provide the pulsewidth of 2-13 ms and off time of 15ms and collected the fuel. I provided 1500 cycles for each pulsewidth and then calculated the cc/min by summing up the time, like for 10ms and 1500 cycles it would be 15s then divided this time by the fuel collected and got cc/s and then multiplying the result by 60 thus got cc/min. I would also like to mention that no fuel was injected on pulsewidth less than 2ms.

I am attaching the obtained result.

Can someone please verify the method.

If you turn the injector on for exactly 1 second (at your reference fuel pressure, and reference voltage), how many CC's flow. Multiply that number by 60 to get the theoretical flow rate of the injectors.

The reason you have various flow rates, is because the delay required for the injector to open (and close). A typical delay may be 1mS, so the longer your pulse width, the less this influences the total flow -- that's why your values are approaching the theoretical flow rate (probably about 330-340cc) -- but a simpler test is what I described above. The delay will change with voltage and differential fuel pressure.

Thanks I will try it get back as soon as possible!!!

I haven't performed any injector flow testing personally however I do have a little insight into what is required and one of the problems is that you're likely to face is that you end up with quite significant oscillations in fuel pressure which can affect your results quite dramatically. Essentially you get a water hammer effect that causes oscillating fuel pressure based on the injector pulse width/duty cycle. Those who specialise in testing injectors in a professional environment go to significant lengths to control the fuel pressure and keep it stable.

Hi David I performed the test you suggested and got the following results. The injector flow was 360 cc/min like you told (close to 340 cc/min). And I calculated the gm/pulse for different pulsewidth taking .78 kg/L as density, and obtained the following result attached below.

Thanks