912 ULS: Partial power loss on takeoff - vapor lock?

Thanks Roger,

"Why is it wrong when so many around the world do I without any issues."

Comon Roger you're an engineer - fuel return back to the gascolator is not best practice.

That it does "work" is , in my mid, not a good argument for a permanent installation - its a make do at best.😈

Thank you for all your responses!

We did some more testing yesterday, trying to approach it as systematically as possible.
The temperature was again around 25 deg C / 77 deg F.

- Cowling on, boost pump off, we started at around 4.5 psi. As the engine warmed up / time passed, the pressure however dropped to about 3.3 psi and began to fluctuate significantly. When I applied full power over the course of maybe a second or two, the pressure dropped to below 2 psi, fluctuating wildly.

- Cowling off, boost pump off, everything stayed nice and cool, the exact same thing however happened again. Vapor lock therefore indeed doesn't seem to be the issue.

- Turning the boost pump on increased the pressure levels by around 0.3 psi, the overall behavior however stayed the same

- We then replaced the mechanical fuel pump, no change

- Next, we capped the return line. First test, cowling off, the fuel pressure was stable at around 6 psi and stayed there, even after the engine warmed up. Under full power it dropped by about 1 psi, no fluctuations.

- Cowling back on, the pressure strangely stabilized at 5.3 psi and I was getting a brief overpressure alert at around 8.5 psi, when I turned on the boost pump. Some more testing showed again a drop of around 1 psi at full power, no more fluctuations.

- I then took the plane for a test flight, no problems, overall stable fuel pressure other than another brief overpressure alert.

- I am not quite sure about the fwf fuel lines, as they came prefabricated, as a kit. I however used -6AN everywhere else and was getting around 140 l / 37 gallon / hour free flowing at the hose coming out of the gascolator, what should be more than plenty.

- We also tested the fuel flow directly at the carbs, with only the boost pump on. Both were pretty much identical at 74 l / 19 gallons / hour, what should again be way more than enough.

Looking at all of this, I am convinced that somehow the setup with the return line going into the gascolator is the problem.
I am also wondering if the restriction jet, sitting right on top of the gascolator, might cause issues as I understand that most installations have it on the other end, at the fuel manifold.

Today we will look into whether we can find a way to run the return line back into on of the tanks, without ripping the wing apart.
If we can't make this work, a small header tank under the panel, might be a solution: https://www.cxracing.com/FST03-57AN

Thanks again, I will keep you posted!

Oliver

All the UL planes I hve seen in canada have a 2-5 litre gascolator behind the seats...

Most have a fuel warning light so ias soon as thw 5 lires is not full, a light goes on letting you know you have 15 minutes to sort things out.

Yhe last one i worked on had the return line going to that.  I think the only thing on the firewall was a fuel filter.

Jay wrote:

All the UL planes I hve seen in canada have a 2-5 litre gascolator behind the seats...

Most have a fuel warning light so ias soon as thw 5 lires is not full, a light goes on letting you know you have 15 minutes to sort things out.

Yhe last one i worked on had the return line going to that.  I think the only thing on the firewall was a fuel filter.

That's what I would call a header tank.

Our CG is however already very far aft, the fuel selector between the two wing tanks is under the panel and I want to keep the fuel system simple.

That's why would ideally see if we can route the return line all the way back to one of the main tanks or, if this turn out to be too involved, to install a small tank under the instrument panel.

Oliver

When fuel or any fluid passes through a restrictor orifice the pressure is reduced at that point. In this case, you are taking fuel under pressure and passing it through an orifice directly into a lower pressure environment on the suction side of the system.  When this happens, bubbles will form in the fuel at rates that are determined by temperature and the fuels reid vapor pressure (RVP). I suspect bubbles are being introduced directly into the gascolator by the return line. You could temporarily install a short section of clear fuel tubing after the gascolator to check for this. Like I mentioned earlier, I suspect this goes away if you try avgas as a test. The RVP of avgas is about 1/2 that of Winter Mogas.