Returnless FI?

[Archer39J]

I’ve been vacillating on this for quite a while and I’m still undecided, but I wanted to post my thoughts and hopefully benefit from the experience of the group.

I can’t seem to justify the need for either a header tank or a complete return system in the Bearhawk for the SDD FI system. This assumes a design where the “hot” fuel bypassed from the engine is teed in just before the fuel pumps, and gascolator in my case. Being a high-wing it seems to me that in any flight attitude and fuel quantity the fuel pump inlets will always see a positive head pressure from the tanks being above them, when on BOTH anyway.

I’ve read concerns about unporting tank pickups and sucking air into the loop, but to me this looks to be impossible since we’ve designed our systems to provide greater than the maximum fuel burn at WOT (125% for FI) on a single tank. So under no circumstance will the pumps be pulling more fuel than the main line can provide despite the high flow rate, closed loop remember, and the only way to get air would be to run out of fuel in the selected tank(s). Well that and sustained uncoordinated flight on a single tank with minimum fuel, so… don’t do that.

That leaves the concern of vapor locking the fuel pump, but again I don’t see how this is possible when the pump inlets are always at a positive pressure from the tanks. While the pump is flowing a high volume of fuel it’s a closed loop, the only differential will come from the fuel the engine is actively burning. This can also be mitigated by locating the fuel block off the engine and taking other steps to avoid excessing heating of fuel return lines.

Ross from SDS has told me that high wings can get away with a small header tank, and I would understand a supplier’s reticence in recommending something different, but this just doesn’t seem necessary. Also this would also keep the system overall more in line with Bob’s stock design, by not adding returns to the tanks or duplex valves.

What am I not considering?

[Battson]

Have I got this right - you are considering ignoring the designer's recommendations based on your own understanding of the system???

Be honest with yourself, the designer knows best. I see a lot of assumptions in the paragraphs above. Just don't go there. The designer knows the system best and understands the risks.

If you want no header tank, get a Lycoming with Bendix FI.

[Archer39J]

Have I got this right - you are considering ignoring the designer's recommendations based on your own understanding of the system???

Be honest with yourself, the designer knows best. I see a lot of assumptions in the paragraphs above. Just don't go there. The designer knows the system best and understands the risks.

If you want no header tank, get a Lycoming with Bendix FI.

​​​​He already capitulated that a return isn't required in a high wing. It just looks like there's room there and before mocking up the system I wanted to run this idea past others.

Can you articulate what is wrong with these assumptions beyond "don't go there"?

​​​​

​​

[jim.mclaughlin924]

Dave,
I must be missing something. Does the FI supplier say a header tank is required or just that a small header tank is acceptable as a return point. I am not quite sure how the need for a header tank has much to do with FI. There are certainly plenty of aircraft with FI returns and no header tanks. I did weld in a bung in the left main for the return from my FI and electric fuel pump pressure regulator.

[Archer39J]

Dave,
I must be missing something. Does the FI supplier say a header tank is required or just that a small header tank is acceptable as a return point. I am not quite sure how the need for a header tank has much to do with FI. There are certainly plenty of aircraft with FI returns and no header tanks. I did weld in a bung in the left main for the return from my FI and electric fuel pump pressure regulator.

A good point. No where is a header tank "required" or even mentioned, and the documentation says that fuel "must" be returned to the main tanks. Though outside the docs it is apparently acceptable to return to a small header tank instead and it looks like many low-wing installs do this, so return to the mains is not apparently a "must" after all. That's what's prompting me to ask the question about whether one is needed in this application, because at least right now it doesn't make sense to me that it would be.

Further, my discussions with the supplier lead me to believe that for a high wing it's probably not necessary either. I'll be discussing this more with him now that I'm a paying customer, but I wanted to ask folks here what they thought first in case I'm missing anything obvious.

I guess it's technically not "returnless" either, just not returning it to the main tanks.

[AKKen07]

I would be concerned about resolving a couple of items with a return line that feeds back to about the fuel selector. I would want to have solid evidence that the volume of fuel returning to the system wouldn’t back up or cause some other problem with no tank to empty into (that’s just a hunch - no experience to back that thought up). The second item is the ability of the retuned fuel to cool sufficiently in that smaller loop vs going to a tank. If it formed a closed loop it could keep warming itself maybe causing warmer and warmer fuel to go to the engine where vapor locking is a bigger issue than the selector. I don’t know if these are realistic issues since I am no expert on the subject. On that note, unless you are an expert on the subject please be very sure before you neglect the system designers advice. You just might not know what you don’t know.

[Archer39J]

Precisely! And those are exactly some of the things to think about. As well as the drawbacks of a full mains return system.

But let's not start off with "It can't be done!" without a solid discussion.

[Battson]

Big caveat, I have no idea how this system works, so I don't understand the materiality of these assumption - but seeing as you mentioned them, I assume they are important.

For me, the assumption that unporting a tank doesn't happen is unrealistic. I have unported a tank by accident on several occasions. The idea that it is always possible to fly with full-ish tanks and do coordinated turns is not realistic. There are factors beyond your control, mostly weather.

Vapour locking - have you considered hot starting and bleed back of the fuel between engine runs?

[Battson]

I would be concerned about resolving a couple of items with a return line that feeds back to about the fuel selector. I would want to have solid evidence that the volume of fuel returning to the system wouldn’t back up or cause some other problem with no tank to empty into (that’s just a hunch - no experience to back that thought up). The second item is the ability of the retuned fuel to cool sufficiently in that smaller loop vs going to a tank. If it formed a closed loop it could keep warming itself maybe causing warmer and warmer fuel to go to the engine where vapor locking is a bigger issue than the selector. I don’t know if these are realistic issues since I am no expert on the subject. On that note, unless you are please be very sure before you neglect the system designers advice. You just might not know what you don’t know.

Fuel heating.

One of the concerns which EFII raised with me about their pump system being left running for longer than the recommended 50% duty was re-circulation through the bypass valve, and heating of the fuel.

Again - not knowing the system - there needs to be enough thermal mass to absorb the heat transferred to the fuel by recirculating around the FWF area and heat transferred to the fuel by the pumps themselves, without raising the fuel temperature an unacceptable amount...

[AKKen07]

It sounds much smarter when you say it...

[Battson]

I don't know about that! It was your thoughts which jogged my memory in the first place.

[AKKen07]

Actually this issue is largely what pushed me away from the EFII system. I didn’t like any of the approved solutions to returning fuel.

[robcaldwell]

Same here. Bendix or Airflow Performance is less involved.

[Archer39J]

Big caveat, I have no idea how this system works, so I don't understand the materiality of these assumption - but seeing as you mentioned them, I assume they are important.

For me, the assumption that unporting a tank doesn't happen is unrealistic. I have unported a tank by accident on several occasions. The idea that it is always possible to fly with full-ish tanks and do coordinated turns is not realistic. There are factors beyond your control, mostly weather.

Vapour locking - have you considered hot starting and bleed back of the fuel between engine runs?

I figured that unporting a tank happens. But when set on BOTH in sustained uncoordinated flight wouldn't that just be transferring fuel from the wing high to wing low tank? The fuel valve and pumps would still be under positive pressure. I would think unporting of both tanks would be transient at most, and again with the fuel rate just being what the engine consumes this can't be any more of a problem than any other fuel system.

​​This systems uses a fuel rail which can be remotely mounted to the baffle or firewall, that can be shielded or even actively cooled if need be but I haven't seen this done yet.

I could do a fluid and heat transfer analysis of the entire system, but what's the heat transfer capacity of a small header tank at equilibrium in that environment compared to a little extra line? Off the top of my head I figure the surface areas would be at least close. In addition to the fact that the pumps and return T are under the head pressure of the tanks, which further reduces the chance of vaporizing the fuel. Care taken to shorten the runs that will be soaking up the heat then getting recirculated as well.

Obviously there would be measurement and possible iteration as a result, but it's not exactly rocket science either.

[Battson]

We are just talking in assumptions now.

Do we know the fuel just transfers between wings? I have no idea.
Do we know if the fuel rail needs shielding or cooling? Not at this stage.
Do we know fuel vaporisation is the risk of fuel heating? I wouldn't think so.
Do we know the fuel selector will always be on BOTH tanks when a tank unports? That is not a safe assumption.

It's called 'experimental' aviation, but I think I want to be as far from true experimentation as possible. I think I've flogged this horse to death, so I will leave it here.
YMMV, maybe your dream is test flying with lots of investigation into problems, trial and error, redesigning, maybe the odd forced landing. I am just thinking from my perspective.

[Archer39J]

We are just talking in assumptions now.

Do we know the fuel just transfers between wings? I have no idea.
Do we know if the fuel rail needs shielding or cooling? Not at this stage.
Do we know fuel vaporisation is the risk of fuel heating? I wouldn't think so.
Do we know the fuel selector will always be on BOTH tanks when a tank unports? That is not a safe assumption.

It's called 'experimental' aviation, but I think I want to be as far from true experimentation as possible. I think I've flogged this horse to death, so I will leave it here.
YMMV, maybe your dream is test flying with lots of investigation into problems, trial and error, redesigning, maybe the odd forced landing. I am just thinking from my perspective.

I mean, this is part of design. You and I are both mechanical engineers if I recall, so this is absolutely within our capabilities and we can already answer some of these questions. Whether you're comfortable designing systems like this is certainly up to you. But to me it's worth it to avoid having return lines full of fuel or a header tank in the cabin to take a hard look at what is possible and what's actually required.

Thank you for taking part in the discussion though.

[Archer39J]

So what else is the header tank for than cooling fuel or letting air escape, as it relates to FI that is?

I believe the main potential issue would be sucking air into the closed loop and having to wait a bit for it to be purged when fuel was restored. But again with Bob system I can't see how this is possible, unless Bearhawks are out there with stuttering engines in any uncoordinated flight and I've not heard anything about it.

[svyolo]

Dave;
When I ordered my engine from Bob, I also bought a rebuilt carb, despite wanting to use EFI eventually. I wasn't sure I was going to come up with a fuel system that didn't scare me for EFI.

EFII uses a "fuel rail" like most old school EFI cars. The fuel picks up significant heat as it passes over the heads. They recommended a minimum size header tank of 6 gallons if I remember right, if you wanted to use a header tank. Too much heating of the fuel. They prefer returning to the main tanks because of this.

SDS used to use something similar, but changed to a small fuel manifold, which then feeds fuel to each injector. They do recommend returning fuel from the fuel block to the mains, or a header tank. In their case they said the fuel picks up minimal heat from he fuel block, and a small header tank can be used.

FAA regs require 150% fuel feed for a carb engine, or 125% for a injected. The pumps SDS and EFII use pump 45 gallons per hour, continuously. I was not comfortable expecting Bob's fuel system to do this.

Most newer cars run return-less fuel systems. Some are pretty complicated electrically. It can be done fairly easily and simply, and I thought about it. But reading some other literature recommended still having a very small return line to bleed air.

In the end, I bought a 2 liter "surge" tank that sits under the right front seat. I have a full size return line running from the fuel block back to this tank. This tank is vented to the top of one main tank, with a T that connects to the other tank so both tanks are vented together.

That return to the surge tank makes me very comfortable that the system will work, and will handle any air that somehow enters the system, although the surge tank itself should prevent this.

So I have a stock Bob fuel feed system to the small header tank, feeding the amount of fuel Bob designed it for. The rest of the system is as designed by SDS. The only addition is the vent to the mains, and venting the two mains together. I think venting the two mains together is kind of recommended by AC43 anyway.

I am completely comfortable with this, and will sell the rebuilt carb.

This is also a fairly standard way to convert an old carberated car to EFI. The "surge" tank I bought from a speed shop, for less than 100 bucks. Eventually I will replace it with a larger (2-3 gallons) header tank, with two internal fuel pumps. Multiple companies also sell these, but they are made for cars,and I can't find one in the shape I want that fits. Otherwise I would just buy one. My aluminum welding is almost good enough to do it now, but I am trying to get-er done, and avoid side projects.

Anyway, that was my solution. One of the easy return-less solutions was to mount the fuel pressure regulator in or on the surge tank. Fuel is returned right after the pump. Several companies make and sell them, but again, I couldn't find one that would fit.

[svyolo]

I found out about "surge" tanks from SDS. Not on the aviation site, but on the site for "other" applications, which is far bigger than the aviation part of SDS. I had never heard of them.

It is plumbed like a header tank is recommended to be plumbed. Viking also sells some very expensive header tanks with integral pumps that are return-less, but their pressure is referenced to atmospheric pressure instead of MAP. If the pumps were big enough, you still might be able to get them to work.

[svyolo]

The first pic is similar, but not exactly what I used. The ports are a little different. The second is one with 2 internal pumps.

[svyolo]

One last thing. The EFI folks don't recommend a gascolator, but the way I did plumbed mine it should work fine. But I am not planning on one as the very bottom of my surge tank, when sitting on the ground, has a fuel bleed valve to check/drain water and contaminants from the low point of the surge tank. So that function of the gascolator is taken over by the surge tank.