Questions for the knowledgeable ?
Airguy offered E10 pump gas requires fuel temp management and higher rail pressure than my engine driven pump and/or AFP electric pump can produce.
Assuming, seasonal / regional vapor pressure variance:
1- How high should rail pressure be to suppress bubbles ?
2- Who sells an RV compatible pump that can supply this pressure ? Apparently cars have one in the fuel tank.
3- Is bypass recirc a must at this pressure ?
4- Is this higher rail pressure acceptable to an AFP servo and splitter ?
Difficult to say whats REQUIRED, considering the myriad approaches available and the fact that this is experimental, so technically nothing is REQUIRED in the first place. However, here's a quick summary and what I did over time and why. The only real trick here is to develop a "hydraulically correct" fuel system - which needs some knowledge of physics and chemistry.
I was running standard Bendix injection with electric pumps from AFP when I originally built my plane (first flight January 2016), with a Borla backpressure regulator on the fuel loop. The fuel came from the tanks with great attention to low-pressure-drop routing and fittings to the pumps (dual parallel, in the doghouse forward of the fuel selector) through an Andair full-duplex valve. From the pumps, the fuel ran to a Tee with one leg going to a Borla pressure regulator holding 30 psi backpressure and dumping the rest to the tanks through the return line on the Andair fuel valve, with the other Tee side going forward through the firewall to the Bendix servo, and standard from there in a deadhead configuration. This worked mostly OK - but was still prone to vapor lock with heat-soaked fuel on the ground especially during a hard climb with warm fuel in the tanks, and hot-starts were the usual fun dance with throttle and mixture.
One of the immediate issues that became apparent was vapor bubbles forming from heated fuel in the FWF area boiling where the pressure was lower. Coming out of the servo the pressure drops from the 30 psi inlet to about 6 psi or even less on the output side between the servo and the fuel divider, and even lower (only a few ounces of pressure at idle) in the divider lines to each injector. In June of 2016, still in Phase 1 at about 36 hours, I put a second layer of firesleeve on all FWF lines to help prevent heat migration to the fuel, changed out the injector orifice restrictors from the standard .024" to a smaller .022" orifice to raise the pressure in the individual cylinder injectors lines (to help stop boiling), put a 4-pound spring in the flow divider to raise the servo-to-divider fuel line pressure, and raised the Borla backpressure regulator from the initial 30 psi to 45 psi. This was required as a side-effect of installing the smaller orifice on the injectors - the small orifice helped raise fuel pressure at idle to stop boiling, but that also meant that it had a high pressure drop at full power fuel flow, which required more feed pressure to overcome, hence the increase from 30 to 45 on the input to the servo. This helped my vapor lock issues quite a bit during hard climbs with warm fuel, but did not completely eliminate them - and hotstart issues were still present as the fuel would heatsoak and boil, flooding the manifold with fuel on shutdown.
In April 2017 with about 144 hours on the airplane, I installed the AFP purge valve between the servo and the divider block, right up against the divider block, with the return line coming back through the firewall and dumping into the downstream leg of the Borla regulator in the fuel returning to the tank. This was an attempt to fix the hot-start issues, and worked pretty well - shutdowns were now done by pulling the purge valve open and leaving it that way. On startup, leave the valve open and turn on the fuel pump immediately on entering the cockpit, let it run while getting seatbelts on and avionics powered up, and by the time I was ready to crank it had been pushing cool fuel from the tanks FWF through the servo up to the divider and through the purge valve back to the tanks for 45 seconds or a minute. This flow rate is not large with the Bendix system when the engine is not running, but there is some present, and it helped move the hot fuel out and put cool(er) fuel in it's place for a startup - and that worked pretty well. Hotstarts were tamed and life was mostly good.
August 2018 at 293 hours, I lost one of the dual fuel pumps, pulled the original AFP pumps out and replaced them with the new dual-pump module being offered from SDS (very nice). This required a new fab of the doghouse but was worth the effort. This system continued to work well but I was growing weary of the hotstart routine and use of the AFP purge valve, I knew there had to be a better way and I was looking for it.
During my January 2020 annual at 563 hours, I raped out the current fuel system and installed full SDS from Ross Farnham. I kept the 45psi input pressure on the Borla regulator, but reconfigured the FWF fuel system and eliminated the FT-60 fuel flow transducer since the SDS sends out flow pulses to the EMS to serve the same purpose. Now my fuel goes forward from the pump module through the firewall, still double-firesleeved everywhere FWF, direct to a manifold on the top of the engine with the 4 cylinder flowlines to each electronic injector, and then exits the manifold back to the cold side of the firewall to the input of the Borla regulator which holds 45 psi of backpressure on it and dumps the rest to the tanks via the Andair full-duplex valve. I have my ignition map pushing as low as 22 degrees BTDC on 30" manifold pressure all the way to 25 degrees below 24" manifold pressure, and move that to 29 degrees for LOP operation below 24" manifold pressure. I also installed a common automotive oxygen sensor and controller in the exhaust to full-time monitor my AFR, and it is reported in the cockpit through the SDS programmer so I can see my "right-now" AFR at all times in flight.
This current configuration has completely cured all vaporlock instances and all hotstart nonsense. Once you get the SDS programmed and mapped correctly (it comes with a basemap that is very close, doesn't need much tweaking) it simply starts like your car - doesn't care about hot or cold, or sea level versus Leadville, Colorado - it just works. I've operated from +116F on the ground in summer in Texas to -12F in the winter in North Dakota, and at cruise altitudes between 14,000' and FL190 pretty regularly, and I've even had it up to FL210 twice, all running 91E10 until summer 2022 when the local Walmarts changed over and started offering 93E10 which I switched over to. I now have 8 years and 1620 hours on the airplane.
It should be noted in fairness that the EFII32 product offered by Robert Paisley offers pretty much the same level of performance as the SDS system developed by Ross Farnham - mostly because it was a direct rebranded copy of the SDS system until Ross and Robert split up their partnership. I prefer the SDS system not due to any particular hardware or software preferences (though SDS has a better injector setup, with a lot of technical reasons to justify my position), but entirely due to the customer support and technical assistance available from Ross. He was right there for me on every phone call and every issue through the course of my journey. His level of support can only be compared to that of Allan Nimmo or Tom Swearingen or Rhonda Barrett - top shelf.
Anyway, not telling anyone else how to do it - just telling what I did...
www.lycoming.com
