My wondrously performing 300KTAS(FL250, ISA+0, 75%) Lancair IV-P has two separate switches labeled high boost and low boost.  Before my recent adventures on hot days, enroute to high key, I only turned on the low boost on climb out, passing 10k. During the last month's shakedown of the fuel system, I discovered that high boost will consistently kill the engine, probably by flooding it. Low boost seems to offer some buffer against vapor, and I have begun selecting it at lower altitudes when the weather is hot. (Joe discussed this in the archives 1/31/07)
 I also was surprised to find that selecting high boost, after low boost is on, has no effect in my existing installation, because of wiring.  Contrary to what you might expect, you must turn OFF the low boost switch to actually get high boost pressure. This probably also occurs when with "low boost" ON similarly blocks the "prime" pushbutton, which I will test on my next trip to the airport. I am considering rewiring this so that both high and low on will command HIGH, rather than low, and pulsing the prime button will pulses high at all times! (anybody know a reason to justify the current setup?)
  I am still not certain that we found the cause of the fuel flow interruption which has now occurred at FL220, on deck after multiple hot day run-ups, and after calibrating the AOA at zero G. Recall what the reminder is, if you ever forgot to select low boost passing 10k. I confess that I have missed this checklist item before, and I think the rough engine sounds are caused by vapor in the fuel system, but I don't have any data to tell me where. My best guesses are the hot low pressure lines feeding the engine driven fuel pump(EDFP) or the even lower pressure (but cooler) lines leading from the tank through the selector to the boost pump by the copilot's right foot.

  My list when the engine goes silent these days is mostly the SEVEN  V's:
1. Fuel tank Vent blocked creating vacuum in the tank
(Why is there no "crossfeed vent" between upper tanks? or valve between vapor return lines? Or is there?)
2. Negative/zero G's exposing the tank gravity feed outlet to air (okay mostly N2, but still N2 Vapor).

3. Vapor in the lines before the selector Valve
4. Vapor leaking past the fuel selector Valve O-rings from the pressurized cabin
5. Vapor in the line feeding the EDFP around the accessory case especially in a hot climb
6. Vapor anywhere else the fuel line runs next to a heat source, though less likely in the higher pressure areas. ( Anybody have a triple point chart for 100LL? Marv has a nice graph on 2/2/07 -- but no pressure axis or contour/family of curves... yet )

7. ALL OF THE ABOVE: My recent experience indicates anything can start the bubbles anywhere back in the system. And as they propagate forward, you can actually aggravate cavitations of pumps by your countermeasures. The more suction applied to a vapor lock, I theorize, the bigger it gets until you have removed all the liquid from the system.

 I have managed to kill a rough engine with both high and low boost. I have not tried pulsing the prime pushbutton, but probably need to turn off the low boost switch because of my odd wiring.

 Procedures speculation:

So far I have not had the engine quit during a critical flight phase, which I define as below the low key altitude, on takeoff(about 30 sec after rotation passing the abort limits). If it does, I have already been scanning fuel pressure and flow during the abort-option phase and the initial climb. These days I am climbing to low key over the highway at Vy after liftoff.

My current thinking is to turn on low boost during takeoff as a preemptive measure on hot days. If the engine roughens or FF or FP drop (note- set alarm) or quits(#1-fly), deselect low and bump the prime(&fly). If this doesn’t work for 5-10 seconds(feels like 2 minutes here), switch tanks(&fly). Then select high boost, then low, then land on the highway.
This all happens in 20 seconds so I am "chair flying" it for proficiency.

   Yeah… mags and mixture, air, fuel and all that, but why did I not already abort the takeoff for low fuel flow when the mixture was set wrong.... and who rotated the mag switch down by my left knee after the runup mag check? There is not much time here so I have reordered priorities for survival. Also pay the insurance, so the decision is similar to the military ejection decision: OPM!

Disclaimer: this is a proposed experimental flight test procedure on an airplane which has demonstrated probable vapor lock problems on hot days. We have not found a "smoking gun" after a complete (one month) search of the fuel system including EDFP overhaul, recalibration, purge, and test of all components. The conservative approach is also to avoid very hot takeoffs until the problem is resolved (I am getting up at 4-5am these days). Long runways and a very low abort decision threshold are also good. You may not have the same odd boost pump wiring or you may have a kid or dog who stealthily twists mag switches and mixture undetected, so your performance and procedures may vary.

Thanks

Bill Miller

PS: Somehow in 16 months and 200 hrs, through all my due diligence, prepurchase inspection, initial training, condition inspections, insurance inspections, recurrent training, and a month's safety standdown accompanied by excellent consultation with superb experts, I only learned about Lancair SBs 020 and 033 yesterday. It was probably because everybody assumed they had already been covered, but there is no record of it.  If you are new to this stuff, read those SBs.