The issue isnt the back presure. This has been a problem from the very beginning just trying to get it to drip through the tube into the nozzle. Ive been talking about this problem for years now and all the changes Ive made to my nozzle arange has evolved around this problem trying to solve it. Its why I was trying this new boiler set up hoping that back presure would keep adjitating the water flow.
The first systems I tried using the crank case vent to pulse the water tank that sort of worked but then you have a hose from your engine to the water tank and its just too crude of a set up.
Then I moved the drip tube to the end of the nozzle to use intake air to flow over the tube that works until the unit starts producing gas and then the presure drops. Like you said there is very minimal vacuum once up and running.
So Cody posted on this concept and I too had this same thought. So I built it and yup same issue flow keeps stalling. This first attempt I think the steam vent holes are too big and not creating back presure. I want it to do this. I want the steam expantion to back feed the water line to agitate it.
But I also think the needle valve is getting some crud build up in it. One of my clients put the housing of an ink pen into the NPT side of the needle valve that threads into the water tank. So that crud that settles in the bottom of the tank cant flow into it. He said its been workiing. So that may be a simple solution.
This Atomizer and the Oxy Concentrator are just experiements. We will never know unless we try you cant predict the unknown and we dont know it all.
The boiler nozzle when its working is pretty impresive. The steam expansion is desirable this is what is cutting out the atmosphere we want this to happen the more the marrier. More displace = more cooling and higher gas energy density. But inconsistancies are killing it because as soon as there is a flood the reaction crashes. If we can sustain this with better presision I think we can get a ton more water in these systems. So far recent test are proving this.