How many cycles before you trust knockoff injectors?
 

I've been given an opportunity to test a range of what appears to be knockoff injectors.
A local guy wants me to test several sets of injectors. He doesn't just want them flow matched, he wants them cycled to test for durability.
I've seen Chinese injectors split in half, have flow rates that vary by as much as 20%, and just stop working all together.

The plan was to build a test rig that would recirculate fuel, through the injectors and back into the tank.
Pulse width and duty cycle would be varied with a small little rise/fall formula to constantly go between 10 and 100% duty cycle and 1ms and 10ms pulse widths and back down again.

So how long would you test them before you trust them? I can do 1 million cycles in about 3 hours considering an average of 10 ms for each cycle. (10,000,000 ms total = 10,000 seconds or 2.77 hours).
They will be re-flowed on my DIY flow bench afterwards to check for wear and failure.

Is 1 million cycles going to be enough? You would get 1 million injector cycles driving 1000 miles with a 60 mph and 2000 RPM average.

I think if they're gonna blow, it'll happen early, within 1k cycles. It's not so much about wear, more about build/assembly quality. Just my.02....

Bathtub curve Joe.

The bathtub curve is fine, except I'll have to find the other end of it.
That could possibly mean millions and millions of cycles. I can do roughly 62,000 miles worth of cycles (62 million cycles) in 1 continuous week. I'd assume that would help me find the wear-out failures.
Fluid temps will be a huge factor at that point. Constantly recirculating the same 5 gallons of mineral spirits is going to cause extreme temperatures after a few hours. A basic cooler will only go so far.

That's exactly where I was going.....you need to find the other side. Why not just make it a HALT test and run it worst case (highest DC/lowest ms)? Run your proposed test to simulate normal use, and then run another test that is just worst case until something goes BOOM. You can then have a rough idea of useful life given that the motor won't be at WOT all the time.

These won't be in a motor. I've built a small test stand. That will allow me to run fuel through the injectors for several hours at a constantly varying pulse width and duty cycle. The fuel will recirculate closed loop. What sprays out of the injectors will return to the tank to be used again.

Fuel temps are a major issue. I might try to come up with a budget fuel cooler.

I want to test a few sets of these for reliability and flow match 8 of them to put in one of my cars as payment.

Yes I'm aware the test setup isn't in a motor. You can control DC and opening time with an Arduino or some PLD....I'm saying HALT test it for worst case and run it until it breaks. Then based on the number of cycles you can speculate how long it would last in a motor.

In reality it will be a very difficult thing to test.

Going boom or stopping working will be easy.

Poor injection at some random time, enough to destroy an engine but not a total injector failure.....will not be easy to check on such a test unless you're doing continuous flow/spray monitoring throughout.

I don't see how this test would tell you much of anything in the real world of heat cycles and oil fouling that goes on in all engines. The injectors as you know get baked and everything in between PLUS the constant hydraulic hammering they take.