So I got to thinking, do I really "need" flex lines to go to the selector? It might be nice, but it is by no means a requirement. It may actually be better without them, because the fuel selector would be suspended in space in a fixed location, with the six hard lines going to it. Then the tunnel cover will come down over the top of it and screw to it. I think I'll try to go this route, or at least see if we can make it work before shelling out a ton of cash for some unnecessary "bling" that will be concealed beneath the tunnel cover anyway.
I started by making two hard lines going from the send and return fittings previously installed in the tunnel last night. Just those two lines were enough to hold the selector valve rigid in space in the correct location. Nice! Now just to make 4 more lines. Oops, I'm running low on AN fittings again. Gotta place another order with Spruce and we're be back on hold for a few more days. No big deal. In the meantime, here's a revised diagram of the fuel system layout:
N165MJ Fuel System Design - Second Draft
This diagram is pretty similar to the previous one (link here) with a few minor changes. Most notably, the regulator (or pressure relief valve) has been moved into the tunnel, and taps off of the inlet to the mechanical fuel pump instead of the servo. I've also ditched the filter between the mechanical pump and servo in favor of a gascolator in a more conventional location. I also threw in a check valve on the aux tank feed and a few more fancy labels.
Labeled photo of current fuel system (masking tape covering 'tee' inlet and outlet fittings)
Since the firewall-forward portion of the system is still TBD, the only part of the diagram that has been semi-finalized is the portion within the tunnel, visible in the photo above. The plan, however, is to run the "Send" line up to the inlet of the mechanical fuel pump, and tap off a "Return" line back to the firewall at that location. With this design, a failed relief valve would only affect the boost pump and not the mechanical pump. In this way the engine cannot lose pressure without a dual failure of both the mechanical pump; and the boost pump and/or the relief valve. This type of reliability is now identical to a traditional aircraft fuel system, with a significantly longer purge/return path. More importantly, the system is still fully forward-compatible with upgrading to an EFI system in the future if/when we go for that.
The only design limitation I can think of is that the boost pump should not be run with the fuel selector in the OFF position (and the mixture in cut-off position). This would result in a condition known as "deadheading," where both Send and Return lines are cut off, resulting in the fuel not being able to go anywhere and causing a spike in pressure. This is a condition that should really never occur anyway, especially while in the air, and the only thing that would happen in this scenario is a tripped breaker for the boost pump. Likewise, fuel cannot be transferred into the system from an auxiliary tank in this configuration for the same reason. That seems intuitive enough, though... fuel gets transferred into the selected tank, so if the selector is OFF then no fuel can be transferred! Also, there is no reason to run the boost pump with the fuel selector OFF, either. This will be placarded though, just to be clear and thorough.
That's pretty much it for tonight! Not much building really... a lot more thinking and planning, with a side of ordering more stuff from Aircraft Spruce. Tomorrow, I feel, may consist of a bit of a priming session if we're lucky! Stay tuned...