What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Return fuel line restrictor fitting

Why? If you restrict the return line you would disrupt the ability of the line to deliver excess unused fuel. Thus rendering it useless. May as well just not install the return system. Remember the in fitting’s are number -6 which is considerably larger and has a higher flow rate, then the -4.

Hope this helps
RD
 
Which system are you guys discussing?

It's your basic fuel injection system. Previously on this aircraft I had an Avstar system and when I bought it from the support guy sold me a restricted fitting for the return line. I just got done rebuilding another engine and it came with a Bendix servo. (see attached). I circled the fitting. It comes out of the same chamber as the inlet.


To be clear I have return lines going through the selector valve that delivers the fuel back to the tank I am running from. The elbow in the attached pic will go to the return line.
 

Attachments

  • IMG_20221123_072612235_HDR.jpg
    IMG_20221123_072612235_HDR.jpg
    264.3 KB · Views: 126
Why? If you restrict the return line you would disrupt the ability of the line to deliver excess unused fuel. Thus rendering it useless. May as well just not install the return system. Remember the in fitting’s are number -6 which is considerably larger and has a higher flow rate, then the -4.

Hope this helps
RD

I had the restrictor fitting on the system before as advised from Avstar and it worked perfectly. No hot start problems. That said I see your point about the different size lines and maybe any size will work.
 
RSA Return Line Plumbing

Im not familiar with using the inlet fuel pressure tap on an RSA-5 injector as a return line fitting but it stands to reason if you are feeding the servo with a -6 inlet fuel fitting and you use a -4 to return the fuel to the tank your just going to pump the fuel in a continuous by-pass and never develop enough inlet pressure to run the engine. I suspect the restricter fitting was to choke off some of the by-pass fuel so it would develop the proper inlet pressure.

I would like to hear from someone who is familiar with plumbing an RSA system in this manor and how it helps with Hot-Starts.
 
Last edited:
IF you want a constant flow return line tee'd to the inlet fitting for a Bendix-style fuel control, yes, it needs a restrictor. The minimum supply pressure for an RSA-5 is about 20 psi at WOT full rich, i.e. at max fuel flow, so that determines the restrictor size.

Personally I think it's a questionable setup, for several reasons.
 
I remember the conversation with the Avstar guy. He was almost excited to hear that I had a return line. I'm going to call their support line and find out the part number.

I'm certainly curious as to why the system would be questionable. It performed very well for me. But if it's dangerous somehow I'd like to let the guy who bought the engine from me know.
 
Return Line

[QUOTE DanH

Personally I think it's a questionable setup, for several reasons.[/QUOTE]

Please opine……
 
I'm certainly curious as to why the system would be questionable.

This isn't black or white. More a matter of carefully considered compromise.

Let's start by asking "What is its intended purpose?"

Is there something special about your application which makes it unlike the multitude of Bendix systems operating without a return? An example might be the intent to operate on auto fuel with high vapor pressure.
 
Last edited:
This isn't black or white. More a matter of carefully considered compromise.

Let's start by asking "What is it's intended purpose?"

Is there something special about your application which makes it unlike the multitude of Bendix systems operating without a return? An example might be the intent to operate on auto fuel with high vapor pressure.

The original intent was to use an Aeromomentum engine which required a return line. That didn't pan out. Went with the Lycoming. The plumbing was there. Avstar recommended the fitting during a discussion with support. I didn't insist on using the return line but it seemed like the guy I talked to liked it. That is certainly speculation on my part from a conversation 10 months ago.

That said, it seems like a system that automatically bleeds down may be beneficial. Is this really different than the AFP purge valve?
 
The original intent was to use an Aeromomentum engine which required a return line. That didn't pan out.

Ahh.

A classic PWM port injection has a pressure regulator in the return flow path. The regulator holds the injector supply at a specific pressure above atmospheric regardless of engine demand. Think of it as a constantly variable restrictor. An alternative introduced later is a regulated pressure electric supply pump and no return line. Given a known pressure and a known injector flow rate at that pressure, the EFI computer merely calculates required injector open time. Injector flow is start/stop/start/stop.

A Bendix RSA type is a constant flow system. That includes Avstar and Precision (both near clones), and Airflow Performance (not a clone, but similar in how it regulates fuel flow). They meter using a pair of connected diaphragms further connected to a ball valve, which feeds the divider. The air diaphragm has venturi pressure on one side and dynamic pressure on the other. The fuel diaphragm has supply pressure on one side and pressure after a jet drop on the other. The combined pressure deltas move the ball valve. Being based on delta, the system is relatively insensitive to supply pressure, and will operate from roughly 15 psi to the limits of its internal seals, possibly as much as 90 psi.

Here's the thing. A Lycoming's engine driven pump is effectively a pressure regulated supply, typically 25 to 30 psi. It's done mechanically by sizing the pump spring. It's also a variable volume pump. An engine cam raises the pump diaphragm to max travel/max spring compression, then releases. With standard RSA plumbing (no return) the spring only pushes the diaphragm down a small fraction of an inch, in direct proportion to outlet demand. At idle, the movement is tiny. At WOT, it is more. Given a return line with restrictor, and WOT operation, the diaphragm travel would be greater yet. Note that being spring driven, as travel increases, spring force (thus fuel pressure) decreases....not a lot, but it's there.

More critical, note the limited system headroom. Best case, supply is maybe 30 psi, while minimum to run right is roughly 20 psi. If restrictor flow rate plus engine demand approaches pump capacity, supply pressure drops.

All the above is background. Let's return to the previous question.

Went with the Lycoming. The plumbing was there. Avstar recommended the fitting during a discussion with support. I didn't insist on using the return line but it seemed like the guy I talked to liked it. That is certainly speculation on my part from a conversation 10 months ago.

If recommended, why?

That said, it seems like a system that automatically bleeds down may be beneficial. Is this really different than the AFP purge valve?

Very different. An AFP purge valve is closed during normal operation. There is no return flow with the engine running.
 
Last edited:
Do I understand correctly that there is a constant retun bleed (calculated leak) that is installed in the servo to divider valve line?

Interesting. That is contrary to my understanding of the operation of a Bendix style system. Since you really want as much pressure as you can get to the fixed orifice nozzles (using common manifold hydraulic theory to keep regulation consistent), it seems that "stealing" pressure via a fixed bleed will only make the problematic low flow, LOP condition worse.

Will be following this thread in the hopes that I learn something new.
 
Do I understand correctly that there is a constant retun bleed (calculated leak) that is installed in the servo to divider valve line?

No. The subject is a return connected near the servo inlet.

Since you really want as much pressure as you can get to the fixed orifice nozzles ...

Strictly an aside, but there are only two practical ways an operator can raise nozzle pressure...advance the throttle, or change the nozzle inserts. IIRC, 0.028" to 0.022" is pretty common.
.
 

Attachments

  • Scan_0001.jpg
    Scan_0001.jpg
    371.9 KB · Views: 40
Last edited:
No. The subject is a return connected near the servo inlet.



Strictly an aside, but there are only two practical ways an operator can raise nozzle pressure...advance the throttle, or change the nozzle inserts. IIRC, 0.028" to 0.022" is pretty common.
.

Copy all Dan, thanks.

That said, and not to be pedantic, but high altitude, cruise flight for many of us in the West is accomplished with the throttle against the stop from takeoff to the TOD. But I understand your point that the servo schedules pressure to the nozzles via throttle position (among other inputs). Unfortunately, with the open ended Bendix style system that uses the nozzle sizing as a metering parameter, there is a conflict between a small enough orifice to meter effectively at low flow conditions (LOP and high altitude) AND the demands of substantially higher flow rates at TO power. This is tough to do in a system that tops out at 30 PSI.

So the scheme in this thread effectively pulls a bit of pressure off the total available for the competing demands of small nozzles for effective metering and high pressure required to overcome the small nozzles at high power. If it works now, great, but seems the scheme is adding extra plumbing and reducing performance margin for no real benefit.
 
It certainly does. I'm convinced. It's getting plugged.

Before we go, return to the photo from post #4

I just got done rebuilding another engine and it came with a Bendix servo. (see attached). I circled the fitting. It comes out of the same chamber as the inlet.

Is the circled fitting an elbow, or a tee?

If an elbow, and connected to a return line, all the return flow would be passing through the Bendix inlet screen. Trash tends to be cumulative based on volume, so the inlet screen would collect garbage at a higher rate than one which only screens the fuel burned. There are owners out there who fail to clean inlet screens, and...

If a tee, it would be connected to both the supply line from the pump and to the return. If so, it would supply fuel to the wrong side of the screen.
 
One additional thing to consider. The Avstar servos have a real tight idle cutoff. After shutting down the engine the fuel pressure starts climbing quickly. All efis manufacturers issued a service bulletin for the Kavliko pressure sensors that were popping under that overpressure.
My buddy has an Avstar on his -7 and fuel pressure hits 50psi+ after shutdown, then it goes to -- (as it hits the set sensor max). No idea how far up it goes but it popped his original sensor a couple of years ago. Resulted in fuel dripping out of the sensor, running down the wiring harness. Yikes.
 
One additional thing to consider. The Avstar servos have a real tight idle cutoff. After shutting down the engine the fuel pressure starts climbing quickly. All efis manufacturers issued a service bulletin for the Kavliko pressure sensors that were popping under that overpressure.
My buddy has an Avstar on his -7 and fuel pressure hits 50psi+ after shutdown, then it goes to -- (as it hits the set sensor max). No idea how far up it goes but it popped his original sensor a couple of years ago. Resulted in fuel dripping out of the sensor, running down the wiring harness. Yikes.

BTW, the fuel hoses FWF have a significant effect on this. The 5000psi hoses are very stiff and allow little "spring" for solid expansion of the fuel. I tested the Vans standard hoses on my 7 (Avstar) and the pressure rise was within the sensor range.

I had a real world product issue due to this very same thing (solid ethanol expansion in a sealed volume), it took our team a while to trace it to the hose as it exhibited the behavior after upgrading the field test to production hardware. I actually approved the hose change - - learned a lesson there.
 
Last edited:
BTW, the fuel hoses FWF have a significant effect on this. The 5000psi hoses are very stiff and allow little "spring" for solid expansion of the fuel. I tested the Vans standard hoses on my 7 (Avstar) and the pressure rise was within the sensor range.

I had a real world product issue due to this very same thing (solid ethanol expansion in a sealed volume), it took our team a while to trace it to the hose as it exhibited the behavior after upgrading the field test to production hardware. I actually approved the hose change - - learned a lesson there.

Woah! Great point Bill! Thanks! I'll have him swap back to a rubber hose.

Sorry about the detour guys, thought that might be why Avstar likes the idea of the restricted return line.
 
Before we go, return to the photo from post #4



Is the circled fitting an elbow, or a tee?

If an elbow, and connected to a return line, all the return flow would be passing through the Bendix inlet screen. Trash tends to be cumulative based on volume, so the inlet screen would collect garbage at a higher rate than one which only screens the fuel burned. There are owners out there who fail to clean inlet screens, and...

If a tee, it would be connected to both the supply line from the pump and to the return. If so, it would supply fuel to the wrong side of the screen.

It's an elbow and I capped it.
 
Im not familiar with using the inlet fuel pressure tap on an RSA-5 injector as a return line fitting but it stands to reason if you are feeding the servo with a -6 inlet fuel fitting and you use a -4 to return the fuel to the tank your just going to pump the fuel in a continuous by-pass and never develop enough inlet pressure to run the engine. I suspect the restricter fitting was to choke off some of the by-pass fuel so it would develop the proper inlet pressure.

I would like to hear from someone who is familiar with plumbing an RSA system in this manor and how it helps with Hot-Starts.

I use the -4 ORB tap on the RSA to feed a block that has both a transducer and a -4 return line to the tank. The return line has a valve to open and close the circuit. I use this to purge hot fuel from the lines up front after a heat soak and it works very well for this purpose. On really hot days, I will leave the return open until just before takeoff to insure no stumbles as I go WOT. This is a big issue for mogas, but not as much for 100LL with it's higher/lower vapor pressure.

While a wide open line would allow faster purging, it also creates the risk of starving the engine if you forget to close it before going WOT. I used a #60 drill bit for the restrictor in the return line fitting. This limits flow via the return to about 10 GPH @ 25 PSI. The mechanical pump delivers close to 45 GPH (boost is higher I believe) and the servo consumes less than 20, so there is no negative consequence if you accidentally leave it open.

I suspect that a wide open -4 return line to the tank would prevent the servo internals from seeing anything close to 25 PSI, though I have never done the calculations, as it seems like a bad idea. With my restrictor, I DO NOT see any pressure drop with the return open. Only downside is that it takes longer to fully purge the hot fuel.

I consider it a must have if running mogas in hot temps.

Larry
 
Last edited:
Back
Top