Canard Community Forum  

Go Back   Canard Community Forum > Firewall Backward and Forward > Fuel Systems
FAQ Members List Calendar Search Today's Posts Mark Forums Read

Reply
 
Thread Tools Display Modes
  #1  
Old 04-27-2004, 08:42 AM
John Slade's Avatar
John Slade John Slade is offline
Flying TurboRotaryCozyIV
 
Join Date: Apr 2004
Location: KWST
Posts: 3,836
Default Fuel injection return systems

Chad robinson & discussed fuel return issues in private messages. With Chad's permission I'm posting the discussion here in case anyone has thoughts, comments or suggestions.
__________________________________________________ ________________

John, I've pestered you with questions on your fuel system plumbing in the past, so I hope you'll tolerate another. I've been toying with ideas on various failure modes and it seems to me that the most critical with your layout is having the return solenoid fail and you end up returning to a tank you don't expect. If you're low on fuel you might not notice this with a balance shift in time to catch it.

What I was wondering is if there was a way to SIMPLY create a checks/balances mechanism for the return flow. Fuel pressure sensors can be expensive, so I hate to litter them about, but it might be a good solution if the return line has enough pressure near the tank to trigger a good indication. Or maybe a simple sight gauge with a trapped ball that would be pushed against one end of the gauge if fuel was flowing that way.

Have you given this concept any thought? Your system is extremely elegant in that it minimizes the parts involved, reduces the plumbing in the cockpit area, and allows a cross-feed, all things I'd like to do.

The only alternative I can come up with that keeps a mix of both worlds (I hate the facet pump idea) is to use the pumps to select the source, and use the plans fuel valve for the RETURN instead of the FEED. The manual valve would give you final control/identification of where the return goes. But I'd like to avoid this if I can address the safety issue.

Regards,
Chad
#1147
(Working on the instrument panel, and dreaming of the future)
Reply With Quote
  #2  
Old 04-27-2004, 08:43 AM
John Slade's Avatar
John Slade John Slade is offline
Flying TurboRotaryCozyIV
 
Join Date: Apr 2004
Location: KWST
Posts: 3,836
Default

Hello Chad,
> I've pestered you with questions on your fuel system plumbing
Its not pestering. Its thought provoking. Thank you.

> seems to me that the most critical with your
> layout is having the return solenoid fail and you end up
> returning to a tank you don't expect.
I agree. Two issues here. Tank pressure and fuel starvation.
1. Tank pressure. My catastropic tank failure was due to a blocked vent in worst possible conditions - i.e. zero consumption, max cross return. This is very unlikely in any flight condition. Keep the vent free (I added a wire mesh cap) and I think we're safe from this one.

2. Fuel starvation
First prevention mode is pilot procedure. e.g. I plan to draw and return from the right tank most of the time during local flying. For long cross country I'll swap every hour, so this is the danger area. Obvoiusly I plan to keep an eye on fuel levels after the switch. This is drummed in habit anyway. Worst case I can think of is that I dont notice the crossfeed, the left tank goes dry and the right tank overflows from the vent. Engine quits. Emergency procedure item # 1 = switch on both pumps. Engine picks up again, and I have a full tank to fly home on. Can you think of a worse case?

> What I was wondering is if there was a way to SIMPLY create a
> checks/balances mechanism for the return flow.
I'd like to do something on this, but haven't come up with a good, inexpensive idea. My solution for now is to keep a damn close eye on it.

> Fuel pressure sensors can be expensive
Not too bad. I got one from Vans for < $100. You only need one on the left return, but would it register pressure on a free return?

> if the
> return line
> has enough pressure near the tank to trigger a good indication.
Right. I doubt that it would. Perhaps some sort of flow sensor.

> Or maybe a simple sight gauge with a trapped ball
Hmmm. The returns are hidden right at the back of the cockpit. You'd have to pipe the left return forward, then back again. Not a good solution. Need something that'll alert you automatically. I REALLY like my LED annunciator / voice alert system. Cheap and effective. Since you're designing you're panel, consider an LED annunciator top center built into the glare shield as I did. Take no space, costs $10 and is fun to do. (note: I have a recent picture of it in the camera. I'll post it shortly.) Also - I recently added a white print on clear tape label showing the failure items with car like icons for quick identification.

back to you're question...
> Have you given this concept any thought?
Oh yea, but I must admit that I'd moved on. I'm glad you brought it back to mind.

> Your system is extremely elegant in that it minimizes the parts
> involved, reduces the plumbing in the cockpit area, and allows a
> cross-feed, all things I'd like to do.
Thanks. Don't forget the redundancy issue. Redundant tanks, pumps & filters is nice. Few single failure modes.

> The only alternative I can come up with that keeps a mix of both
> worlds (I hate the facet pump idea) is to use the pumps to select
> the source, and use the plans fuel valve for the RETURN instead of
> the FEED. The manual valve would give you final
> control/identification of where the return goes. But I'd
> like to avoid this if I can address the safety issue.
This would work, but you're just as likely to forget to switch as you are to get a solenoid failure

How about a simple low fuel warning on the left tank? I don't have one, but they're not too expensive or hard to install. Put it at, say 1/4 tank. If the solenoid fails, AND you're running the left tank, you'll get a warning (voice in my case) before the engine quits.

Am I missing anything here?

> (Working on the instrument panel, and dreaming of the future)
Keep on dreaming. It's the journey, not the destination (says someone who's almost there).

Regards,
Keep thinking. Keep asking questions.
John
Reply With Quote
  #3  
Old 04-27-2004, 08:45 AM
John Slade's Avatar
John Slade John Slade is offline
Flying TurboRotaryCozyIV
 
Join Date: Apr 2004
Location: KWST
Posts: 3,836
Default

John Slade wrote:
>>I've pestered you with questions on your fuel system plumbing
>
> Its not pestering. Its thought provoking. Thank you.

=) Then I'll keep, ah, provoking you.

> I agree. Two issues here. Tank pressure and fuel starvation.
> 1. Tank pressure.
> My catastropic tank failure was due to a blocked vent in worst
> possible conditions - i.e. zero consumption, max cross return.
> This is very unlikely in any flight condition. Keep the vent free
> (I added a wire mesh cap) and I think we're safe from this one.

I agree. This is the sort of thing that could happen if a vent clogged anyway, even without this type of setup, so the traditional means of addressing it are still OK. Double-venting helps.

> 2. Fuel starvation
> First prevention mode is pilot procedure. e.g. I plan to draw
> and return from the right tank most of the time during local
> flying. For long cross country I'll swap every hour, so this is
> the danger area. Obvoiusly I plan to keep an eye on fuel levels
> after the switch. This is drummed in habit anyway. Worst case I
> can think of is that I dont notice the crossfeed, the
> left tank goes dry and the right tank overflows from the vent.
> Engine quits. Emergency procedure item # 1 = switch on both pumps.
> Engine picks up again, and I have a full tank to fly home on.
> Can you think of a worse case?

I'm not sure, but I'll think about it. The nice part of having to deal with a return is that you can so quickly purge/prime an air-charged system. Just get ANY fuel flowing and the pressure will drive all the air out in a second or two. You might be restarting instead of letting the engine restart itself, but that's not the end of the world.

I suppose one thing you could do to address this issue is draw as much as possible from the tank that requires the return solenoid to be energized. That way your failure mode has you falling back to the tank with more fuel in it.

To break this down, the instigators of a problem that I can think of are:
1. The solenoid fails into an unintended position.

2. The solenoid fails during a switch. Is this a break-before-make device? It would be interesting to know if there is some middle state where fuel can flow to BOTH output ports, or if fuel could flow to NEITHER. If the device jammed mid-switch (maybe a seal breaks and jams it?), making both sounds bad because you have no control over the return and it'll flow to the path of least
resistance. But breaking both is worlds worse because now you have NO return line. Your fuel pressure gauge, if you're watching it, should be an immediate indicator that this has happened, you'll probably start running rich really fast, and all sorts of bad things can happen. I guess the biggest danger is fire if your fuel system plumbing can't take the sudden pressure surge before you kill the pump, and sprays fuel into the engine compartment.

3. The solenoid becomes clogged. Does it have the same cross-sectional flow area that a manual valve has? If so, anything that would clog it should have been picked up much earlier.

So I suppose the worst case is a jam of some sort. To address this, it might be a good idea to have a double-pressure (80psi?) relief valve somewhere plumbed directly to one of the tanks. You can have this with a manual valve, too, of course, but you get warnings that it might be about to happen, and you can also jiggle it to see if you can get it into one or the other states.

Obviously, this only applies if the device is break-before-make.

>>What I was wondering is if there was a way to SIMPLY create a
>>checks/balances mechanism for the return flow.
>
> I'd like to do something on this, but haven't come up with a good,
> inexpensive idea. My solution for now is to keep a damn close eye on it.

I've been thinking about two things. First, I might try to get some fuel pressure sensors from a junkyard on-the-cheap. Instead of wiring those to actual pressure gauges, I'd just use them with a simple interface circuit (comparator, probably $0.50 of parts) to drive a logic level on/off output that could be wired to a warning light on the annunciator panel. That is, if a side is on and fuel ain't flowing, turn on the light. And vice-versa.
The second is the above emergency-return. Let me know how your solenoid works, because it may be irrelevant.

> Not too bad. I got one from Vans for < $100. You only need one on
> the left return, but would it register pressure on a free return?

Yeah, it would pretty much always register pressure, at least atmospheric. Most of the cheaper devices are "absolute". But a comparator circuit from a 555 timer or op amp is really easy to make, and could be used to create a level-set trigger that only turns on above a certain point.

>>if the return line
>>has enough pressure near the tank to trigger a good indication.
>
> Right. I doubt that it would. Perhaps some sort of flow sensor.

It's tricky. It might have enough just past the solenoid, for instance, especially if the diameter of the return line is small enough. The pressure drop wouldn't really start kicking in until you got close to the tank. For safety, I'd be putting this right after the solenoid, maybe even on a T from the solenoid's output port.

> Hmmm. The returns are hidden right at the back of the cockpit. >You'd have to pipe the left return forward, then back again. Not a >good solution. Need something that'll alert you automatically. I >REALLY like my LED annunciator / voice alert system. Cheap and >effective. Since you're designing you're panel, consider an LED >annunciator top center built into the glare shield as
> I did. Take no space, costs $10 and is fun to do. (note: I have a > recent picture of it in the camera. I'll post it shortly.) Also - > I recently added a white print on clear tape label showing the
> failure items with car like icons for quick identification.

I'd love to see the picture. Currently, I'm planning a somewhat special top/side-view annunciator to replace several devices. The design is to create a panel with a pictorial representation of the Cozy, and various pieces of information in their relevant locations. For example, I'd have a small RPM display right where the prop is. LEDs would show fuel pump activity, and these would be located in the wing tanks on the picture. You get the idea. It would be rather large - 4"x4" or so - and would consume panel space, but I think it
would be worth it. And since it would replace several instruments I think space won't be a problem.

The reason behind this is that my plane will be wired with CANbus, a
communications bus designed for automotive use and all sorts of critical communications networks. My annunciator panel will be a bridge between the primary and backup busses, so it will have access to all of this data (fuel pressures, RPMs, etc.)

> How about a simple low fuel warning on the left tank? I don't have
> one, but they're not too expensive or hard to install. Put it at,
> say 1/4 tank. If the solenoid fails, AND you're running the left
> tank, you'll get a warning (voice in my case) before the engine
> quits.
>
> Am I missing anything here?

No, that third piece is a good idea. I must not have been thinking straight - I'm going to have fuel-level sensing anyway. Might as well add this. You know, another interesting way to address this is a simple wet/dry sensor in the sump. If your sump starts running dry, you can start warning like crazy before you run into an actual outage event.

Still beats a header tank.
Reply With Quote
  #4  
Old 04-27-2004, 08:46 AM
John Slade's Avatar
John Slade John Slade is offline
Flying TurboRotaryCozyIV
 
Join Date: Apr 2004
Location: KWST
Posts: 3,836
Default

Hi Chad.
Interesting stuff.
> I suppose one thing you could do to address this issue is draw as > much as possible from the tank that requires the return solenoid
> to be energized.
On a long cross country this would be good procedure. On short local flights, just use the fuel from the non energized side.

> 2. The solenoid fails during a switch. Is this a break-before-make > device?
I think I checked this when I got it, but I'll remove it and check again. This is important information to have.

> It would be interesting to know if there is some middle state
> where fuel can flow to BOTH output ports, or if fuel could flow to
> NEITHER.
right. I'll check and let you know.

> 3. The solenoid becomes clogged. Does it have the same
> cross-sectional flow area that a manual valve has?
Yes. It's pretty big. No problem there, I think.

> I've been thinking about two things. First, I might try to get
> some fuel pressure sensors from a junkyard on-the-cheap. Instead
> of wiring those to actual pressure gauges, I'd just use them with
> a simple interface circuit (comparator, probably $0.50 of parts)
> to drive a logic level on/off output that could be wired to a
> warning light on the annunciator panel. That is, if a
> side is on and fuel ain't flowing, turn on the light. And vice-
> versa.
Sounds like a good idea.

> I'd love to see the picture.
I'll send a link when it posts

> Currently, I'm planning a somewhat special
> top/side-view annunciator to replace several devices.
Sounds excellent! Sounds like a potential product!

> a simple wet/dry sensor in the
> sump. If your sump starts running dry, you can start warning like
> crazy before you run into an actual outage event.
How is that done?
Reply With Quote
  #5  
Old 04-27-2004, 08:47 AM
John Slade's Avatar
John Slade John Slade is offline
Flying TurboRotaryCozyIV
 
Join Date: Apr 2004
Location: KWST
Posts: 3,836
Default

>>Currently, I'm planning a somewhat special
>>top/side-view annunciator to replace several devices.
>
> Sounds excellent! Sounds like a potential product!

I've been tinkering with visual information to display. Some things like RPM are easily represented there. Things like fuel pressures/flows could be done that way too, but might be harder to "scan". But I also intend to include warnings such as gear up/down, etc.

>>a simple wet/dry sensor in the
>>sump. If your sump starts running dry, you can start warning like
>>crazy before you run into an actual outage event.
>
> How is that done?

About what you'd expect. You have a separate fuel level sensor in the sump, just not a complex one. A simple air gap capacitor, could be two sheets of aluminum with some insulating spacers, mounted about 2/3 up the wall of the sump. If there's ever air there, you're about to run out of fuel. A comparator, which again is all of a $0.50 circuit, could be used to trigger a logic 0/1 output to drive a warning light, buzzer, voice warning, or any combo of the three.

Regards,
Chad
Reply With Quote
  #6  
Old 04-29-2004, 05:02 PM
crrobinson14's Avatar
crrobinson14 crrobinson14 is offline
Member
 
Join Date: Apr 2004
Location: Niantic, CT
Posts: 39
Default Fuel returns, more thoughts

I've had some more thoughts on fuel return systems. I'm still not terribly fond of a solenoid-driven return, mainly because there are failure modes for the solenoid that could cause serious trouble (failure mid-switch, for instance; also, is it make-before-break or make-after-break?). I've thought of a few alternatives and figured I'd throw them out to the crowd. Note that all three options continue to use John Slade's method of supply - a no-pass-while-off fuel pump on each tank with a T to connect the two supplies at the fuel rail.

The first option would be to exchange the solenoid for push-pull controlled fuel valves, a-la Marc and Nadine Parmalee. This is the simplest option, and is probably my first choice (for an alternative, anyway). There are some drawbacks, though. If you want sensing (like to light up a crossfeed indicator) you need to add a microswitch somewhere, and you start building back up to a reliability concern again.

The second is a combination of the solenoid and an "emergency" return. The idea here is that in the event your fuel system pressure suddenly spikes, you would use an emergency valve that bypasses the solenoid. It would have to feed back to just one of the tanks, but you'd avoid all the troubles that leaving the system overpressured would cause, while still using the solenoid for primary switching. It's a fairly easy install - the same valve and setup Marc Parmalee used, and just a T before the solenoid to the valve, and fromm there to a T on one of the solenoid outputs. It's one more push-pull cable but it would do the job.

The third option would be an automated system, and I'm tinkering with this right now to see if it would actually work. The fuel pressure regulator needs a certain output pressure (actually, a lack thereof) to work properly. If the return was suddenly pressurized the regulator would probably maintain a pressure that was too high. But where is this line? Is a few psi OK? And what pressure is normally in the return, at say, X inches from the regulator, with say 3/8" lines? if you could work out a safe way to do this, you could put a low-pressure regulator further down the line that Ts around the solenoid, like above. If the solenoid fails the low-pressure regulator would open to relieve the return line back to a default tank. Sensing this has happened would be tricky, but it's an option.

I'd appreciate all thoughts.
Reply With Quote
  #7  
Old 04-29-2004, 06:05 PM
Dust's Avatar
Dust Dust is offline
Super Moderator
 
Join Date: Apr 2004
Location: Troy, Michigan
Posts: 7,963
Default new forum, same story

Yes, my old one, the $45.00 US used double throw vales i bought from wentworth aircraft. I say valves cause i bought two, uno por yo y uno por clifford.
__________________
Enjoy the build,njut av byggandet, godere il costruire, nyd bygningen, geniesse den Bau, apolafse tin kataskevi, disfrute la construcción, curta a construção, Pidä hauskaa rakentamisen parissa, bouw lekker,uživaj grade?inaslajdaites postroikoi, geniet die bou
dust

maker of wood, fiberglass, foam dust, metal bits and one day a Cozy will pop out and swiftly whisk me from meeting old friends and family to adventures throughout the world
Reply With Quote
  #8  
Old 05-03-2004, 08:30 PM
Rob Wright Rob Wright is offline
Cozy #1056
 
Join Date: Apr 2004
Location: Belfair Washington
Posts: 110
Default valve

I saw a valve for sale in the wicks catalog that was designed to control both the return and the supply. It is made for fuel injection requiring a return. It is 450 dollars. Has this been considered and ruled out do to the cost and plumbing required or for some other reason?
Reply With Quote
  #9  
Old 05-03-2004, 08:35 PM
John Slade's Avatar
John Slade John Slade is offline
Flying TurboRotaryCozyIV
 
Join Date: Apr 2004
Location: KWST
Posts: 3,836
Default

That's the Andair valve out of UK. Very nice piece of kit. I had one in my hands at Sun & Fun a few years back. I resisted it because I didnt want to spend the money, and I didnt want six 40 psi fuel lines in the cabin.

If you like the valve approach, talk to Bulent Alieve. He got a bulk buy on a similar valve which was MUCH less expensive.
Reply With Quote
  #10  
Old 05-04-2004, 11:12 AM
LargePrime's Avatar
LargePrime LargePrime is offline
redesigning every part before build
 
Join Date: Apr 2004
Posts: 502
Default

We should get al wick over here and create some "standard" layouts and FEMA them and get a real data point for evaluating our options here.
Reply With Quote
  #11  
Old 05-04-2004, 11:31 AM
rui rui is offline
Member
 
Join Date: Apr 2004
Location: Calgary
Posts: 192
Default

I'm really attracted to the Tracy Crook system of feeding and return to only a single tank. The only problem with this system (that I know of, are ther others?) is the contaminated fuel problem. Does anybody have any hard numbers when it is the case where one tank is contaminated and the other isn't?

I think the cases where a plane goes down because of fuel starvation due to wrong tank being selected (or fuel valve in the off position) are much greater than the bad fuel problem.

But I'm torn, the control freak in me wants to be able to switch tanks. The analytical side says that its safer to do the single feed/return system. Less complex, less likely to be affected by pilot error (which, like it or not, causes the majority of small plane accidents), easier to use, especially with some sort of automatic fuel tranfer system.

Basically an idiot-proof system, why do I even consider others? Ah, ego, thats it. I won't screw up and select the wrong tank on final approach, nope not me...
Reply With Quote
  #12  
Old 05-04-2004, 12:35 PM
John Slade's Avatar
John Slade John Slade is offline
Flying TurboRotaryCozyIV
 
Join Date: Apr 2004
Location: KWST
Posts: 3,836
Default

Quote:
We should get al wick over here ...
Yes. He'd be an excellent guy to run some of this stuff by. Al's flying a Subaru Cozy with his own very fancy glass panel. Consider yourself appointed as the recruiter.

The reason I wanted two independent tanks was redundancy. Fuel contamination is one possibility, but there are others - crud in the tank blocking the filters, air leak or fuel leak in the line, blocked vent....

I like the idea of having a backup "system".
Reply With Quote
  #13  
Old 05-07-2004, 04:29 PM
Rob Wright Rob Wright is offline
Cozy #1056
 
Join Date: Apr 2004
Location: Belfair Washington
Posts: 110
Default

Can you expand on the issues around having six 40 psi fuel lines in the cabin?
Reply With Quote
  #14  
Old 05-07-2004, 05:57 PM
John Slade's Avatar
John Slade John Slade is offline
Flying TurboRotaryCozyIV
 
Join Date: Apr 2004
Location: KWST
Posts: 3,836
Default

Sure.
Aluminum has been known to crack or fracture, and fittings CAN leak. This is especially true if you bang them on the side of a house at 130mph

The plans system is gravity feed, so if do you have a leak it's not too catastrpohic. By contrast, a fractured high pressure line will dump a lot of fuel in your lap very quickly. Admittedly, AN fittings and aviation grade aluminum pipe are pretty secure, and the chance of a problem in normal usage is slight. But if there IS a problem, its a much bigger one at 40 psi.

Take Rich Hughe's Cozy III accident for example. He walked away without hardly a scratch. Imagine if he'd had high pressure fuel lines runing from the seat back to the firewall. All the stresses of the impact would be almost bound to cause a fracture. There's a really hot engine right there, and a good chance of a shorted wire somewhere. If the fuel pumps are still running the fire hazard with such a senario is very significant. One way around all this is to return to a header tank, but this, in itself, becomes a potential problem in case of accident if it's somewhere inside the cabin.
Reply With Quote
  #15  
Old 05-07-2004, 06:33 PM
Dust's Avatar
Dust Dust is offline
Super Moderator
 
Join Date: Apr 2004
Location: Troy, Michigan
Posts: 7,963
Default

ok, so i am the only one that is going to have an airplane engine and a $45 simple double throw valve on this forum, thats fine, just so others on this forum know, you are allowed to use an airplane engine on your cozy AND post here
__________________
Enjoy the build,njut av byggandet, godere il costruire, nyd bygningen, geniesse den Bau, apolafse tin kataskevi, disfrute la construcción, curta a construção, Pidä hauskaa rakentamisen parissa, bouw lekker,uživaj grade?inaslajdaites postroikoi, geniet die bou
dust

maker of wood, fiberglass, foam dust, metal bits and one day a Cozy will pop out and swiftly whisk me from meeting old friends and family to adventures throughout the world
Reply With Quote
Reply


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -4. The time now is 09:56 PM.


Powered by vBulletin® Version 3.6.7
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.