Fuel system options for fuel injection

[Jim Sower]

John,

You've got me pretty well convinced about the "Single Failure Point" or SFP of my idea. So as long as we're on that subject:

1. What drives your fuel pump? Hellacious switch or relay (SFP)?

2. What drives your return selector - switch or relay (SFP)?

3. I presume that the return selector is a screw driven valve controlled by a DPDT switch and will fail in the position it's in when the lights go out. True? I further presume that the motor/screw turns itself off when it arrives at the selected position (FP?). True? Have any of those "scrwe operated electric valves" ever been known to fail midway?

4. The Cozy design selector, your duplex pump/return system and my draw-from-right system involve roughly equal pilot workload for fuel management, so that issue is sort of a wash.

5. I VERY much like your idea of installing a "blank" 1/4 NPT Al plate at the top and bottom of each tank (along with an 1/8" NPT for the vent:o) as the simplest way to mix and match various design ideas and allow things to evolve with minimal effort.

6. However, EVERY joint and fitting is a FP, even if not a glaring one.

7. I feel that the sump-in-hell-hole gets some unwarranted knee-jerk negative reaction. Would a 3-5 gal sump in the hell hole, perhaps under the gear bow, pose any more danger to anyone than the couple of feet of ribs, floxed blindly to the strake between the cockpit/cabin and the fuel tank that you've ALREADY got? A sump tank in the hell hole (not to be confused with the cabin) should not constitute any more of a danger, IMO, than all the plumbing the design system puts in the cabin area proper.

 

'Nuff for now .... Jim S.

[niceez]

Jim & John,

 

Has anyone tried a submerged fuel pump in a "sump" as such using avgas? I have not researched it yet.

 

I still like the two tanks into one sump also (worked well for Vance all these years) but of course the SPF is back. I was thinking that a valve and a line could be set up to one tank to provide an emergency back up for clogs but fuel comtamination might not work so well in this scenario.

[John Slade]

>What drives your fuel pump? Hellacious switch or relay (SFP)?

Switch, but there will be two of them.

 

>What drives your return selector - switch or relay (SFP)?

Same switches as above.

 

>Have any of those "screw operated electric valves" ever been known to fail midway?

So long as it doesnt fail closed, I'm happy. I'll be doing some testing on this.

 

>I VERY much like your idea of installing a "blank" 1/4 NPT Al plate at the top and bottom of each tank. However, EVERY joint and fitting is a FP, even if not a glaring one.

True, but stuffing a pipe through a composite panel and securing it with flox is also an fp. Also, the per plans bend in the pipe as it comes out of the tank seems like a weak point.

 

>I feel that the sump-in-hell-hole gets some unwarranted knee-jerk negative reaction.

Agreed, but it's something else that can leak. Pipes are easier to seal. My main reason for rejecting it is SPF. Nicez's belt & braces idea has some merit, but adds complexity.

 

>Has anyone tried a submerged fuel pump in a "sump" as such using avgas

Not sure about avgas, but the automotive industry has pretty much settled on this method of delivery. I think a pump in the tank would be tough to implement in a Cozy. I have a Mazda in-fuel pump unit. I couln't see a way to install it either in a main tank or in a header.

[Jim Sower]

Dale,

<... two independant tanks with a mechanical valve. KISS seems to be screaming at me ...Return feeds ... have a mechanical valve...>

What screams at me is SIX fuel lines INSIDE the passenger compartment

 

John,

The feed-from/return-to-right configuration and the dedicated-strake-to-strake routing is, IMO, toss-up by any standard. I don't mind a sump, but I sort of distrust remote valves. You don't like the sump and distrust my single route to the pumps. It's a judgement call, and I guess it's all been said.

 

Pre positioning blank outlets top and bottom of fuselage structure into strakes is still a champion idea. Gives you all manner of flexibility, maintainability and growth/evolution potential at virtually no cost.

 

How big a footprint will your duplex pump/filter plumbing occupy on the firewall? will they be on the "hot" side or in the hell hole under the gear bow (my choice if there's room). How about isolated, independently fused, dedicated circuits to the pumps/return valve? Photocell low-level detectors in the lines/blisters to warn of an empty tank just before the motor stops?

 

Best, Jim S.

[niceez]

Real quick,

 

My ides for a sump is to remove the rear seat back and make one intragal with it. On the LEZ we have this contoured half tire thing on the back side. Seems like a little research once removed would allow a guy to make it a fuel tank on the back.

 

My valve is mounted on the right side below tank and sump level and the valve stem and handle are just aft of my right elbow so infligt I never have to let go of the stick. I eliminated over 12 feet of aluminum tubing. The best method I have seen fro the LEZ is Bob Davenports. Took the sumps all the way aft to the cowl and routed the fuel to the valve on the firewall.

 

Just an idea!

 

Must get to movin'

[dust]

I know you have made your mind up, but, I can't help thinking about it, too many pilots die from fuel starvation. if you switch to larger diameter tubing and a double throw valve, you now have a lower presure system and no pilot workload!

 

You don't like having high preasure tubing in the cockpit, well it is there, I only suggest doubling it. If it bothers you, just safty it with an extra safty cover over the fuel lines.

 

I think the double throw vales you will find that are made for this are set up for the larger diameter tubing, the beech tubing is larger.

 

You may even find a used beech or cesna valve at an unreasonable price.

 

Mike

[John Slade]

Hey Mike, I'll be receptive to input about this fuel system (and anything else in the plane) even AFTER I fly it. Of course you're right - fuel starvation comes right after VFR into IFR which is why I've agonized on this one so much. There's pros and cons with every option....

 

Changing the size of the pipes doesnt make sense to me. The outlet of the fuel pump is 1/4 inch. There's not much point in connecting this to a half inch line. You've gotta have high pressure at the rail anyway. Large lines make more sense for gravity feed systems.

 

The double throw valve from Andair is nice, but fitting it would require four long aluminum lines in the cockpit, or Andair's mechanical remote. Neither option is attractive, partly because of the potential for fracture and partly because this makes the pumps suck a much longer distance, not to mention the expense. The pilot workload is the same in either case - i.e. you have to remember to switch tanks. I'm comfortable with this, and prefer it to a "both" system after many hours in a Piper. At least this way fuel is like Georgia - "always on you're mind".

 

Anyone who see's a problem with any of the above rational, PLEASE let me know.

[Jim Sower]

Mike,

<... too many pilots die from fuel starvation ...>

Too many of them crash a mile or two from the airport with 'several' gallons of fuel in the UNselected tank. How does your system prevent this? I want, at the end of EVERY flight, to have ALL of the fuel on board in the tank that's feeding the engine. During recovery phase, I'm closest to the ground and the workload is highest. I don't want to have to worry then about fuel management and switching tanks.

 

<... in the cockpit, well it is there, I only suggest doubling it ...>

My system has NO high pressure (or low pressure for that matter) fuel lines in the cockpit. There are a few in the hell hole, none in the cabin area. I am not sure what bigger lines accomplishes other than making cramped spaces more cramped.

 

Your system has roughly twice the plumbing in the cabin as mine has in the hell hole. Your valve is a failure point and constitutes pilot work load (and won't comfortably purge a tank). My transfer pump is a failure point that constitutes (I would argue somewhat less) pilot workload (and it always, very comfortably, purges a tank).

 

<... you have made your mind up ...>

Not yet. I'm a long way from plumbing my fuel system, and I won't 'make up my mind' until I'm right there.

 

It's a judgement call. Lots of factors that we each weigh ourselves - I don't know how often you've had reason to run a tank 'pump empty', but I've done it several times and don't EVER want to have to do that again.

 

Regards, Jim S.

[John Slade]

Jim,

I'm sure many pilots (and a lot of CFIs I've known) would argue that discipline keeps good pilots from getting to the point where they NEED that last couple of gallons. Personally, I've never needed them in 37 years of flying. That being said, it would be fairly easy to add your "pro" to my system with a facet pump, switch, check valve and a short piece of pipe in the hell hole. Still no need for a header tank. Damn! I just sold my facet pump. I'll have to relay on dicipline.

[Jim Sower]

John,

Agreed. The bulk of the fuel starvation cases I would guess involved guys flying "instruments" (there's no 'visual reference to the horizon' when you've got your head up your ass ....). One of my times was exactly that - running my mouth and enjoying the flight and the scenery when ....

 

The other involved a rather long IFR flight (with adequate but not excessive reserve) and most legs in excess of 100 nm. I was in the clag or between layers the whole hop. The winds were over twice the headwind forcast and when I called FSS 100 mi into the flight. I only had VORs, so accurate positioning and 'howgozit' was possible only at the widely separated waypoints. I knew by half way that I had unfavorable winds, by 70% or so that they were very unfavorabe and eating badly into my reserve. Figuring out positions to do fuel checks using crossing radials is unreliable on a _good_ day. This was a choppy *bad* day. I landed on fumes after running one tank pump empty while at altitude. It was a long way between air patches with approaches so I had little (but not *no*) choice as to where I would go. If I'd relied on my best guess of empty tank when I did my last switch, I would have run out scud running through a mountain pass 8 or 10 miles from my destination. One could, with 20-20 hindsight, come up with a number of alternative courses of action that would have got me on the ground with 5 gal of so. Gawd knows _I_ did. But I think I did a credible job under the circumstances.

 

On account of that, I am really anal about having ALL of my fuel where I need it toward the end of _any_ flight.

 

We are all moulded by our experiences .... Jim S.

 

PS A tactical jet with 200 or 300 lbs of fuel is a real attention getter too.

[dust]

My friend the accident investigator/7000 hour pilot overules all else in concerns to safty.

 

One mfg had vapor lock at 9500 feet because bubbles came out of fuel at that altitude and blocked the return lines, thats why they went to larger diameter fuel lines.

 

isn't the fuel pump near the engine? if that is the case, then you just need volume to feed the high preasure at the engine and the larger lines do that and the larger returns insure that any bubles won't block the flow.

 

I know i am kicking a dead horse, but, few people have seen what my friend has seen! The one mfg lost a number of planes before they found out that the returns needed to be bigger and piped to the top of the tank.

 

I know i seem to simplify it but my system will be just that, one valve at my fingertips, large lines to and from.

[Rob Wright]

just curios,

 

How did Greg Richter set up the Razzle Dazzle?

[John Slade]

If memory serves, he has 1/2 inch lines direct to a small header tank in the hell hole, then two pumps and filters on the firewall. Return is to the header.

[Jim Sower]

<... My friend the accident investigator/7000 hour pilot overules all else in concerns to safty...>

No argument here. I am an ex Marine Aviator with a commercial SE, ME, Inst rating 4000 hrs in tactical jets, over 1000 in combat. I majored in Aero Engineering in school. I did NOT fall off the turnip truck last week. I know about accidents. I have personally crashed and burned (blew the canopy, leapfrogged over the windscreen, slid down the nose and sprinted out of the fireball). I have spent afternoons out in the boondocks with a plastic bag and a butter knife scraping friends of mine off the trees and rocks. I am no stranger to the causes of accidents. I am walking around today because I learned from other folks' mistakes and misfortunes. Also a LOT of pure dumb luck.

 

<... One mfg had vapor lock at 9500 feet because bubbles came out of fuel at that altitude and

blocked the return lines, thats why they went to larger diameter fuel lines...>

I have to believe that there was a little more to it than that, but that's OK. On a pressurized system, with the fuel supply at or above the pump inlet, and no serious preheating of the fuel, there will be no bubbles and bigger fuel lines serve no purpose. Do you really delude yourself that bigger fuel lines that solved some problem some where will automatically benefit you in your configuration and situation?

 

<...isn't the fuel pump near the engine? >

NO. It's on the other (cool) side of the firewall.

 

<... if that is the case, then you just need...>

You don't "just" need anything (with the possible exception of a "better-than-nodding" acquaintence with Mr. Newton). Fuel system, or any other system design is more than picking solutions that sound appealing and install them in the hope that perhaps they're relevant to your issues.

 

<... the larger lines do that and the larger returns insure that any bubles won't ...>

Larger than WHAT?? John's lines are large enough. They will work just fine. They have on any number of other applications. There is no reason in the world ("reason" is a key word here) to rush around "fixing" sh*t that ain't "broke".

Harvey's Scond law: " ... if it works, dont f*** with it..."

Harvey's First law: "... if you don't know what it does, don't f*** with it..."

 

<...I know i am kicking a dead horse...>

What are you talking about? What does a dead horse look like? Sound like? Do you find it useful to kick one? I've been open to discuss any idea that comes along that seems to impact the issues at hand in some relevant way. There are positives and negatives for every proposal. Some are relevant and some are not. We have to evaluate and weigh them, each within the context or our own application and come up with the best rational solution that we can. There is no compelling or convincing reason whativer to assume that the solution to someone else's problem (who we don't even know) will even be relevant to, much less slove, your problem (or John's or mine).

 

<... but, few people have seen what my friend has seen ...>

I have seen a enough of it to suit me. How much of it have YOU seen? How involved have you been in coming up with a solution for the problem that caused an accident? Your friend certainly has been more often than I. I certainly have been more often than you.

 

<...The one mfg lost a number of planes before they found out that the returns needed to be bigger and piped to the top of the tank...>

This the same mfgr as above, or a different one?

 

<...I know i seem to simplify it...>

You got THAT right!

 

<...one valve at my fingertips...>

What airplane are you building? Only the stick and throttle are at my fingertips. On my Long-EZ, and the fuel selector is "within reach" - where it can easily be (and occasionally is) overlooked.

 

<...large lines to and from...>

You seem fixated on big fuel lines like they are going to fix your life or something. If that was even _remotely_ true, EVERYONE would have big old fat fuel lines and life would be terriffic and everyone would live forever.

I sincerely hope that big(er) fuel lines will actually be relevant to your design and somehow make it better.

In addition, I truely hope you NEVER forget to switch tanks, that you somehow never run into a situation that leaves you short of fuel, that you ALWAYS have sunny days and tail winds and that you never ever run into a problem that can't be fixed by installing bigger fuel lines.

 

Best regards, Jim Sower

 

"...If your onliest solution is a hammer, the problems all start lookin' like a nail ..."

My dear old Dad ...

[John Slade]

>I am walking around today because I learned ....

Well... hobbling, at least.

[Jim Sower]

Hobbling today, playing basketball tomorrow. I'll take it. Been crippled for the past 4 or 5 years and that's my limit. Wish I'd done it a couple of years ago (but the technology hadn't reached Albany by then). See you in Port Gibson soon and we'll do some sprinting ..... Jim

[John Slade]

Back to the fuel problem...

I'm now in the process of installing the plumbing. Another builder (whose name I won't mention in case he wants to remain annonymous) suggested an interesting alternative to the motorized valves for switching returns. The pressure in the return line is very low. Why not just T the returns, he says. Hmmm. I talked to Tracy Crook about the fuel pumps. He tells me that the pump will run at about 35PSI and, worst case at idle, I'll be returning about 30 gals/hr. Tracy also mentioned that the filters should go on the engine side of the pump, not the tank side. This saves the pump sucking through the filters and, presumably, protects the injectors from eating bits of pump. He has a gascolator on the suction side of the pump. I wasnt planning on installing gascolators. Maybe I should.

 

Tracy felt that the return flow direction might be influenced by minor differences in restrictions to the flow. I think the fuel would tend to flow to the tank currently being drawn. Once a tank is full I doubt the return will force fuel out of the vent rather than simply flow to the other tank. Worst failure mode I can see is an overfull tank venting fuel and an increased pilot workload.

 

The idea is so simple that I think I'll try it and see what happens. I'll put the T in a place where I can easily retrofit motorized valves later if I dont like the behavior.

 

I also ordered a set of capacitance probes and duel guage from http://www.recreationalmobility.com/ These seem like a reasonable deal. The probes can be bent and cut down to 12.5 inches as needed. Rather than have them "sticking out of the top of the wing" like one plane I could mention, I'm planning on installing these sideways in the rear baggage area. Has anyone installed similar probes?

 

Thoughts and suggestions welcome.

[Mevans]

John;

 

I was up at Tracy's last week picking up my redrive and EC2 and we talked fuel systems. I am at the same point in my LEZ as you regarding fuel systems. I have a question. By your last post, are you saying you are going to put a T in your return line? Is it your thought that the return flow will naturally go into the tank being drawn at that moment? I agree worse case is you vent extra fuel out of a tank. It would seem the return would flow into a less than full tank before it would force a full tank to vent overflow. A simple T with the return choosing its own flow would be simple, I like that. You will be flying long before me. I might rig mine that way and wait for your experiance. It would be fairly easy to change the return later if not working.

 

Keep building, one day we will fly!

 

Matt Evans

Clearwater, FL>

[John Slade]

Yes, Matt, you have it. Its a kind of "suck it and see" experiment.

 

There is one "worst case" that bothers me ... low on fuel during decent you're running at idle and transferring fuel at 30 gals / hr. Lets say you have 10 gals / side ... if, for whatever reason, the return goes the wrong way you'll run a pump dry in 20 mins. What we really need is a simple on off motorized valve with metal fittings. One each side and the transfer problem is solved.

[Jim Sower]

John,

<... I think the fuel would tend to flow to the tank currently being

drawn. Once a tank is full I doubt the return will force fuel out of the vent rather than simply flow to the other tank...>

...think...doubt... are key words here. The numbers are very VERY small re pressures, etc. so a very minor restriction in one line would cause surprisingly dominant return to the other tank. In your favor is the notion that the vent line being smaller and pressurized by ram air would tend to prevent fuel loss over the side. OTOH, once fuel began dumping overboard, you would have a siphon effect of about twice the ram air pressure wanting to continue syphoning fuel.

 

Your idea of "T"d return dumping half of a LOT of fuel into the unselected tank on descent late in the flight is an IMPORTANT consideration. I would suggest that we're now back in the "...all the fuel on board in the selected tank..." business with a vengance. One solution that suggests itself is to install a facet transfer from left tank to right. Just make a practice of ending your flight with the engine drawing from the right tank and the facet pump ON.

a) Minor additional plumbing

b) "T"d return OK

c) Always end the flight with all your fuel in selected tank

d) Facet pump overcomes assymetric return through "t"

Only failure mode is if Right tank pump fails. If that happens, you know IMMEDIATELY (engine goes quiet) and as soon as you switch to the other pump (about 1.2 sec) you're heading most direct route to the nearest suitable air patch.

If right pump fails very late in flight when left tank is empty, yo're screwed, but then again, if you want guarantees, buy a toaster.

 

I'd say "T" the lines, install facet pump.

 

Best, Jim S.

[John Slade]

>I'd say "T" the lines, install facet pump.

This would work OK until the day someone else flys my plane and decends on the left tank. Besides, I just sold my facet pump.

I still prefer the idea of a motorized on/off valve in eash return line. If only I could find a suitable valve....

 

If you INSIST on having access to every drop of fuel, then a facet pump could be added to this system.

Regards,

John

[Mevans]

Slightly Clearer now;

 

I think will install a facet pump running left to right, The one in Spruce runs at about 30 GPH. "T" the return line. Switch tanks as normal throughout flight. Last stages of flight would always include right tank "ON" facet pump "ON". Whatever fuel is onboard is being directed into the feeding tank, regardless or where the return is sending it. I like that better than a motorized method of switching the return line, but that would also work. Thanks for the input fellas.!!

 

Matt

[Jim Sower]

John,

<... If you INSIST on having access to every drop of fuel...>

It's not ME that insists. It's my sister-in-law PRUDENCE. You've probably got one too. She's kind of obnoxious most of the time, but typically ends up with a compelling argument. Particularly AFTER THE FACT :>))

 

I have links to JC Whitney electrical L-R screw driven pumps (fail where they fail) that might work for you as well as a solenoid operated unit that can be plumbed to fail L or R. Since your solenoid will fail without warning at irregular (but arguably long) intervals, fail-Left into the tank you can purge would seem prudent here. Your return selector fails into my original design.

 

Go with the solenoid (cheap) return selector valve, and clean up with the Facet pump (buy another you cheap screw, you can pay for it with the money you save on solenoid v. screw valve).

 

The link for two of the JCW valves is:

 

http://www.tvbf.org/gallery/?dir=subsystems%2FElectric_Fuel%20Valves

 

jcwhitney.com has all manner of stuff.

 

Tomorrow we'll look into it deeper :>)) Jim S.

[niceez]

Guys,

 

It seems that the complication of this system grows. I have yanked out some of my old engineering systems books and been doing more reading. Then I contacted some old classmates that work still work in the business. We all agree that;

 

Two tanks into one sump should eliminate gads of plumbing. I am forced to submit that from either both tanks or from one sump, the fuel should be pushed not sucked. The pumps are readily available and it would appear that a tee'd return would be simple, yes/no?

 

If fuel pushed from a sump I would install the return line to the sump if the mains are gravity feed to the sump.

 

A major concern seems to be fuel in the cockpit. Seems like a Cozy has a large enough firewall to accomidate a firewall mounted sump near the bottom of the firewall.

 

I think simplicity has a lot to be say. Further, our recommendation seem to change with different engines types.

 

Cheers,

[Jim Sower]

<... It seems that the complication of this system grows....

Agreed.

I am in favor of a sump. John is not. My last post was oriented toward fuel management for his no-sump system. He does have a few points. One is that with a sump contamination of one strake contaminates the system. I personally think it's not a compelling argument since you're supposed to check for water in BOTH strakes, and if you don't check you're not likely to have dramatically more water in one than in the other.

I would lean toward the strakes feeding a "T" through manual maintenance valves. The 'stem' of the 'T' would contain the fuel flow transducer. Fuel would be drawn from the sump, through an emergency shut-off valve, branch into one or the other pump/filter unit and thence to the fuel rail and returned (perhaps through a rudimentary cooler) to the sump. No moving parts other than FF transducer, emergency shutoff and pumps. Damn few failure modes. No fuel lines in the cabin.

 

But it DOES have a sump .... Jim S.