What's new
Van's Air Force

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

Lean of Peak ... I smell fuel?

bkervaski

Hellloooooooo!
Testing
So getting my head around Lean of Peak, I got the basics.

I've gone LoP twice, once on the way to Airventure very briefly, thought I smelled fuel so I went back RoP. My passenger didn't smell anything.

Today, went LoP (7500 feet) and smelled fuel again, this time the same passenger did smell it.

I never smell fuel any other time.

I'm assuming that it's just the overflow from the engine driven fuel pump vent line making its way back into the airplane from suddenly going rich to lean?

My fuel pressure usually stabilizes around 28psi (without the boost pump).

At any rate, scared me off trying again until I figure it out.

Thoughts?
 
Last edited:
I would pull inspection covers and look for fuel stains in the tunnel and along the route of all the fuel lines. If you have the standard fuel system running LOP shouldn't have any effect on presence of fuel in cabin IMHO.
 
I'm hardly an expert, but the engine driven pump shouldn't be pumping gas back into the plane...there's an internal bypass valve built into the pump so that excess pressure is bypassed but not pumped back into the plane. Are you sure that the pump is not pumping gas out of the overflow line and overboard? I'd think if you have leaks in the system, you'd know it by now, so I discount that. I'd look at your overflow line and see if you have any blue stains. Just a guess!
 
Are you sure that the pump is not pumping gas out of the overflow line and overboard?

Yes, this is what I think is happening .. and the fuel is somehow making its way to the air vents .. anyone ever experience something similar?

I would pull inspection covers and look for fuel stains in the tunnel and along the route of all the fuel lines. If you have the standard fuel system running LOP shouldn't have any effect on presence of fuel in cabin IMHO.

I never smell fuel any other time, only the two times I've gone LoP and those were 20 hours apart from each other. Once I went back RoP the smell went away.

Curious.
 
Pull the cowl and look for staining around the injectors.

Having fuel burp out of the injectors when LOP is not unheard of. There’s a whole thread over at the BeechTalk forums about this. It’s a long thread, but might be what is happening here.

Basically, the local pressures both inside the intake and around the cylinders is very chaotic. The pressure spikes caused by the intake valve closing can be higher than the upper deck pressure, and when fuel pressure at the injector is low enough (this lowering the venturi effect in the injector) the fuel can back up out the injector air vents. Sometimes happens to Mooney guys (Lycoming powered) with the ram air intake open, and apparently it’s quite common amongst the Reno racers as well. Maybe one of them will chime in.

If there’s no other sources of fuel leaks found, try doing running LOP again, but this time pull the throttle back by 1/2” of MP and/or change the RPM. That will often stop the burping.

If this is what’s happening and you can smell it, you might also have some firewall sealing issues as well, though it could just as easily be working its way out of the cowl vs through the firewall.

Someone actually filmed it happening in his Bonanza.

https://youtu.be/cvepzhj_OLE
 
Last edited:
Today, went LoP (7500 feet) and smelled fuel again, this time the same passenger did smell it.

There is a phenomenon called "auto rough", proven to occur when flying over bodies of water, right after spotting a shark.

I'm assuming that it's just the overflow from the engine driven pump making its way back into the airplane from suddenly going rich to lean?

There is no overflow from the engine driven pump, and no path back into the airplane.

My fuel pressure usually stabilizes around 28psi (without the boost pump).

Right. It's determined by diaphragm spring pressure, which is the same rich or lean.

...there's an internal bypass valve built into the pump so that excess pressure is bypassed but not pumped back into the plane.

No, there is not. See above.

Yes, this is what I think is happening .. and the fuel is somehow making its way to the air vents .. anyone ever experience something similar?

The RV-8's rear seat air vent intake under the right wing will pick up some fuel smell if the right tank is full, and burps a little out the tank vent during climbout.

Pull the cowl and look for staining around the injectors.
Having fuel burp out of the injectors when LOP is not unheard of. There’s a whole thread over at the BeechTalk forums about this. It’s a long thread, but might be what is happening here.

Just signed up with Beechtalk, as I'd like to hear the arguments. It's a recent interest. I made two sets of deltaP measurements, upper plenum next to the nozzle vs MP gauge. Then I went to a much more detailed method, using a fast differential pressure sensor recording intake tract pressure (at the primer port) vs nozzle bleed pressure in a turbo injector rail.

The first two sets are really overall averages, slow analog stuff. They did not indicate any circumstance in which the 390 suffered from a pressure reversal in the nozzle...bleed was always higher than manifold.

However, averages might easily miss some period in each 720 degree 4-stroke cycle where there could be a flow reversal in the nozzle bleed, thus the fast sensor recording on my laptop. The detailed record shows one very short period of pressure reversal, and the amplitude is quite small. I very much doubt there is any significant fuel escaping from the nozzle bleed.

The data is on my laptop, which is at the office today. I'll try to post a plot later.

Basically, the local pressures both inside the intake and around the cylinders is very chaotic.

I might buy into the idea of turbulent upper plenum pressure, thus highly variable bleed source pressure. I have not attempted to record 720 degree cycles using an open static source near a nozzle. Probably would not do any good, as it might vary a lot from airplane model to airplane model. I can demonstrate that the deltaP across the nozzle is very stable, cycle to cycle, when fed with a turbo rail.

The pressure spikes caused by the intake valve closing can be higher than the upper deck pressure, and when fuel pressure at the injector is low enough (this lowering the venturi effect in the injector) the fuel can back up out the injector air vents.

The assumption here is that air is drawn into the nozzle by a venturi effect. It is not, something pretty clear from the fast pressure measurements. It enters because there is a pressure difference between the bleed source and the intake port.

Someone actually filmed it happening in his Bonanza.

https://youtu.be/cvepzhj_OLE

Fascinating. That guy is losing a lot of fuel. Got a link to the story?
 
Last edited:
There is a phenomenon called "auto rough", proven to occur when flying over bodies of water, right after spotting a shark.

:D :D :D

Real fuel smell in this case, not faint or brief, pronounced. No leaks as far as I can tell. Baffled.
 
Sorry, Dan. I don?t have any more info on the video, other than that he was having issues with this phenomenon and decided to film it.

Is it possible that the less-optimized cooling intake from a certified Bonanza would allow the delta-p to be closer than your RV-8? There are several Bonanzas exhibiting this behavior, but most apparently do not. I wonder if the Bonanza cowl design creates conditions that, on average, fall just outside where this can happen, but there?s enough airplane-to-airplane variability that a small percentage of them do. 🤷🏼

In the few cases I?ve read about where it was tried, a slight reduction in MP prior to leaning prevented the fuel from burping out of the injectors. This makes me think that it?s probably a delta-p issue. I?ve been wrong many times before, though...

It would be nice to hear from some of the Reno racers who run turbo injectors with ram air feeds into them to know for sure if this is the reason.

Chris
 
I'm sure I'm wrong, but I may have misinterpreted the following from EAA's "Basic Fuel Pump System" by Tony Bingelis:

"When the correct engine driven fuel pump is installed, it will discharge (pump) more fuel than your engine needs . . . actually the pump should be capable of providing a minimum fuel flow of 125% of that required for maximum take-off power. This excess capacity will not be a problem in the operation of your engine as an internal relief valve-factory adjusted to deliver the fuel at the correct pressure for a particular carburetor or fuel injector installation - prevents the development of excessive fuel pressure at the fuel inlet."

That seems to imply that there's a bypass valve. Now I'm very confused. Dan, please explain.
 
There's a fuel vent line coming out of the engine driven fuel pump that dumps below the cowl .. my best guess so far is that is the source of the fuel smell.

Also, I have my two firewall penetration wraps off due to my pmag install, wonder if this is how vent fumes making their way into the cabin.
 
Last edited:
Bill,
I was parked next to you at OSH. I have experienced fuel smell a few times when going LOP in the past. It has happened when I pull the mixture lean too fast in combination with letting the mixture go too lean. At no time when this happened to me has it ever caused a problem. I have figured out that I do not need to ?yank? the mixture knob so hard and fast that this will happen. Consequently it has been a long time since it has happened. The key to operating LOP is to get to the lean side quickly, but once you get past the Big A** Pull and get to the lean side there is not much to worry about (if anything) that will cause some catastrophic apocalyptic destruction of the universe. Contrary to what some would have you believe.

My experience with this has shown me that to avoid this from happening I had to not pull the mixture lean so fast and so deeply lean that whatever component is allowing the fuel to escape lets this happen. I would say, don?t let that stop you from learning how to operate your plane LOP. Once the unknown becomes known you will then start understanding more about how your engine really performs. Go fly some more and experiment with HOW you go LOP. You will find what works best for your engine.
 
What type of servo are you running that has a vent dumping into the cowl?

Stock IO-390 engine driven fuel pump with a Vans provided tube that dumps below the cowl.
 
Last edited:
Is it possible that the less-optimized cooling intake from a certified Bonanza would allow the delta-p to be closer than your RV-8? There are several Bonanzas exhibiting this behavior, but most apparently do not. I wonder if the Bonanza cowl design creates conditions that, on average, fall just outside where this can happen, but there?s enough airplane-to-airplane variability that a small percentage of them do. 🤷🏼

Entirely possible. I have no idea about Bonanza cowl pressures.

I?ve been wrong many times before, though...

Me too.

This is interesting!
 
My hangar mate next door has a Mooney with GAMI injectors. He noted blue staining on the cylinders around his injectors. It was the pressure differential n the cowl as compared to the injectors. He learned to reduce his manifold pressure very slightly did indeed stop it from happening.
Deep magic happening here but 100% repeatable on his Lyc. IO360.
Turns out it is a well known phenomenon on the Mooney Owners club site.
He didn?t mention the smell, just saw the blue stains.
 
He noted blue staining on the cylinders around his injectors. It was the pressure differential n the cowl as compared to the injectors. He learned to reduce his manifold pressure very slightly did indeed stop it from happening.
.
From what I have been told and learned, this has much to do with the plenum and cowl design which has rarely seen in the RV.
 
I experienced this in an RV6 with an IO360. Similar to RVbySDI, if I leaned aggressively and quickly, I smelled a touch of fuel. By moderating the leaning, I rarely encountered this.
 
Maybe a dumb question...

Bill, is it possible the fuel vent line from the pump is either too short or routed different from the plans? Honestly, I don?t know if fuel should ever come out there unless there?s a problem. Guess I need to read up on that, but mine seems bone-dry and no evidence of blue dye.

I was looking at my plane trying to figure out what fuel system component would allow vapor into the vents or elsewhere and I?m at a loss as well.

I have been doing a good bit of LOP, especially using the Garmin lean assist feature, and have never smelled fuel. I could probably get to my LOP setting faster than I am, but I?m still figuring out some power settings. Are you using a Bendix or AFP servo? If AFP maybe a call to Don might provide some insight.

That?s all the wild guesses I have.
 
Several pilots here report they only get the fuel smell momentarily when leaning is described as too quickly, or aggressively, or a similar description. Question for you...Does this rapid leaning with fuel smell go so far as to cause the engine to stumble, or more specifically, include a loss of RPM?

I'm looking for some phenomenon which would explain a momentary reversal of nozzle deltaP due to rapid leaning.

Break.

I imagine a few readers are wondering how a "turbo injector" is different from the standard.

Actually, the nozzle body and restrictor is the same for both standard and turbo. The difference is in the shroud assemblies. A standard nozzle has a simple bell-shaped shield covering a brass screen. A turbo shroud is sealed to the nozzle body with o-rings, and the bleed air is supplied via a pipe and hose.

This is the nozzle itself. Note the air bleed hole in the side of the nozzle body. Air enters there and entrains with the stream of fuel from the small diameter passage in the restrictor, mixing as fuel and air pass through the larger diameter passage and into the intake tract.

Nozzle.jpg


Here are the shrouds, standard on the left, turbo on the right.

Std%20and%20Turbo%20Shrouds.jpg


So why the piped bleed air with a turbocharger? Manifold pressure is higher than plenum pressure by a large margin, so plenum pressure won't work as a bleed air source. Instead, bleed air is supplied from a tap downstream of the turbocharger.

Now to the EAB application. We can install turbo shroud assemblies on our normally aspirated engines, and supply the feed tube (the "rail") with a ram inlet to ensure high bleed pressure. There are two theories in play here. One is that higher bleed pressure results in better mixing and atomization in the nozzle, resulting in less cycle-to-cycle variation when deeply LOP. The other is that the ram bleed pressure source is consistent, rather then suffering from random turbulence and variations in local pressure at each nozzle location on top of the engine. I don't know that anyone has absolutely proved either theory, but a lot of users have reported better overall engine behavior, including myself. FWIW, it's also impossible to vent fuel from the nozzle bleed into the upper plenum, as seen in the video link of the Bonanza engine. If there is any reversal, the fuel is trapped in the shroud and returned into the nozzle.

This is the current installation on my RV-8. Don Rivera at Airflow Performance can supply all the major parts.

Frontal%20View.jpg


Components.jpg


Another partial installation, customer photo courtesy of AirFlow Performance. The stainless steel rail (the long tube), short hoses, shrouds, etc are kit parts. Builder adds a custom ram inlet tube from the blue fitting forward, bent to suit the cowl inlet.

Install%201%20from%20DR.jpg
 
Last edited:
Stock IO-390 engine driven fuel pump with a Vans provided tube that dumps below the cowl.

The 1/4? line on the fuel pump is not a vent, it is a telltale for a failed fuel pump diaphragm. In any case this feature is not affected by what fuel mixture you are running.

If the fuel pump telltale drain has fuel coming out, replace your fuel pump before next flight.

Here is one of dozens of articles about this: https://www.tempestplus.com/Products/FuelProducts/FuelPumps/FuelPumpFAQs.aspx

If you do not have fuel coming out of the fuel pump drain, I would start to look for the fuel leak on the cabin side of the firewall.

Carl
 
If you do not have fuel coming out of the fuel pump drain, I would start to look for the fuel leak on the cabin side of the firewall.

I have, no leaks (no blue stains). But this phenomena only happens when I've gone LoP.

I'm going to do some more testing.

Thanks for the explanation about fuel pump vent!
 
Bill, is it possible the fuel vent line from the pump is either too short or routed different from the plans? Honestly, I don’t know if fuel should ever come out there unless there’s a problem.

You're correct James. No fuel will appear in the fuel pump vent line unless the pump is failing internally. It is a telltale device, as well as a path to get the leaking fuel safely out of the engine compartment.

I was looking at my plane trying to figure out what fuel system component would allow vapor into the vents or elsewhere and I’m at a loss as well.

The cowl seams leak air, and the cabin air inlets on the side of the fuselage will pick up some of it, feeding it to the cabin. If I get a little oil dropping on a hot exhaust pipe, for example, I get a whiff of it in the cabin of the -8. Fuel smell could follow the same path, and probably does in Bill's case. The issue here is the source of the fuel. Given the "only when going LOP" clue, I can think of no source other than fuel venting from the injector bleeds, and I'm very curious as to why.
 
Last edited:
There are two theories in play here. One is that higher bleed pressure results in better mixing and atomization in the nozzle, resulting in less cycle-to-cycle variation when deeply LOP. The other is that the ram bleed pressure source is consistent, rather then suffering from random turbulence and variations in local pressure at each nozzle location on top of the engine. I don't know that anyone has absolutely proved either theory, but a lot of users have reported better overall engine behavior, including myself. FWIW, it's also impossible to vent fuel from the nozzle bleed into the upper plenum, as seen in the video link of the Bonanza engine. If there is any reversal, the fuel is trapped in the shroud and returned into the nozzle.
Dan,

Thanks to your explanation and education on this subject in earlier posts, I started looking into this as a possible improvement as I had suspected the same phenomena at higher altitude (above 17K) and planned to see if I could benefit from it.
A bit more research and learning, I end up changing my restrictors in the nozzle to a smaller size for better atomization especially at higher altitude and indeed it helped very much. I took a flight yesterday up to FL210 and still could do LOP though the speed suffered very much. For the same altitude (17.5) a gain of .5G of fuel by going LOP and losing only 1 knot was recorded.
From what I have learned, one of the tale tell of low plenum pressure that will cause this issue is the blue stain at the bleed air passage on the nuzzle.

Though I can imagine the fuel smell makes it to the cabin if the source is at the nuzzle.
 
... I end up changing my restrictors in the nozzle to a smaller size for better atomization especially at higher altitude and indeed it helped very much.

I've considered it, but not tried it. What size, and any change in fuel flow at high power?

From what I have learned, one of the tale tell of low plenum pressure that will cause this issue is the blue stain at the bleed air passage on the nuzzle.

Keep in mind...blue stain can also indicate the nozzles are clocked incorrectly, i.e. the stamped figure "A" on one of the flats is not facing down. When it faces down, the bleed hole is on top. Reverse it, and fuel can boil out the bleed hole after shutdown. I kinda think some might boil out even with the bleed facing up.
 
I've considered it, but not tried it. What size, and any change in fuel flow at high power?



Keep in mind...blue stain can also indicate the nozzles are clocked incorrectly, i.e. the stamped figure "A" on one of the flats is not facing down. When it faces down, the bleed hole is on top. Reverse it, and fuel can boil out the bleed hole after shutdown. I kinda think some might boil out even with the bleed facing up.

Unfortunately I did not make a note of my FF at high power after the change and it is raining hard today to go for a second flight . I should be able to get that info tomorrow. I changed from .28 to a .24 size restrictor.

And you are, again, correct in regards to the orientation of the hole on the nuzzle.
 
You're correct James. No fuel will appear in the fuel pump vent line unless the pump is failing internally. It is a telltale device, as well as a path to get the leaking fuel safely out of the engine compartment.

Very interesting thread! To expand on Dan?s comment, a close look at the vent line exit is an excellent preflight check item. If you see blue or liquid fuel, the lower diaphram has a problem. If you see oil, the upper diaphragm is goners. In either case, fix before further flight. If you don?t have an over-boarding vent line(I?ve seen several homebuilts without!), install one properly and promptly, making sure it exits the cowl in a way that is easy to preflight.
- Otis
 
This is a new anomaly.
You smelled it quickly.
Pax did not.
I vote back pressure from leaning creates a slight seepage? Fuel selector/pump/filter area.
Does it happen using both tanks?

Use a white cotton glove from Home Depot to pick up trace.

When doesn?t it do it?

R
 
Thanks for all the feedback!

Went to the hangar at lunch ... a few notes:

1. No blue stain around vent line
2. No blue stain around any of the injectors
3. I have yet to re-sealed the two firewall wire penetrations from my pmag install (on purpose)

Assuming #3 is where the fumes are coming from, the question still remains from where the fuel originated.

If the vent line isn't an overflow line, and no visible signs of leaks, then just fuel overflow/excess blow back from the servo?

At this point assuming it's just leaning too far, too fast. As soon as the temp drops below 10 million degrees I'll go test some more :D
 
Last edited:
Previously I wrote:
The detailed record shows one very short period of pressure reversal, and the amplitude is quite small. I very much doubt there is any significant fuel escaping from the nozzle bleed.
The data is on my laptop, which is at the office today. I'll try to post a plot later.

So much for memory. Here's a data sample from 10500 ft @ WOT, a likely setup for LOP cruise. I have not yet grasped a connection to fuel flow, but it describes the pressures.

overlay.jpg


You're looking at the difference between bleed air supply pressure (here in a turbo rail and shroud) and the intake port at the primer fitting. Green is bleed pressure exceeding port pressure, while red is port pressure exceed bleed pressure. Put another way, red has the potential to blow fuel out the nozzle bleed hole.

The 23 BTDC notation is where I used a second channel to put a timing marker on the plot...it's #1 plug firing. There are 720 crank degrees between 23 BTDC markers; at 2400 RPM each division represents 0.00133 seconds.

With the throttle wide open, there are three periods of pressure reversal, not one as I remembered this AM. The periods are short. The longest averages about 3.5 divisions (0.00466 sec, 67 crank degrees), just after the intake valve closes on the compression stroke. All three added together come to 0.0117 seconds, or about 168 degrees of crank rotation. 168/720 means there is the potential for reverse flow about 23% of the time. Hmmm.

Here is the effect of throttling at 6500 feet. Assuming the same forward airspeed, bleed pressure would remain the same, while the partially closed throttle plate reduces pressure in the intake port. DeltaP shifts about 1" Hg in the desired direction, but the time periods (in seconds or crank degrees) don't change much.

6500%20WOT%20vs%2065.jpg
 
Last edited:
1. No blue stain around vent line
2. No blue stain around any of the injectors
3. I have yet to re-sealed the two firewall wire penetrations from my pmag install (on purpose)
Assuming #3 is where the fumes are coming from, the question still remains from where the fuel originated.
If the vent line isn't an overflow line, and no visible signs of leaks, then just fuel overflow/excess blow back from the servo?

You may not have spent enough time in the fuel smell condition to stain significantly.

There is no "overflow/excess blow back" from the servo. Line pressure between the engine driven pump and the servo remains stable, regardless of mixture knob position, a fact you can see on the cockpit fuel pressure gauge. The only pressure change due to leaning is detailed in the IO-390 manual...nozzle pressure.

Nozzle%20Pressure.jpg


Note that a shift from 11.5 GPH (a typical best power mixture at 65%) to 8 GPH (about 50 LOP) changes nozzle pressure by about 2 psi.
 
Last edited:
Several pilots here report they only get the fuel smell momentarily when leaning is described as too quickly, or aggressively, or a similar description. Question for you...Does this rapid leaning with fuel smell go so far as to cause the engine to stumble, or more specifically, include a loss of RPM?

I'm looking for some phenomenon which would explain a momentary reversal of nozzle deltaP due to rapid leaning.
Yes, the engine will stumble in that scenario and yes the RPM drops off rather dramatically, although I do not recall how much. It has been several years since I experienced this phenomenon.
 
In my case, I had over leaned and there was a reduction in power when I got a whiff of fuel. The engine ran a little rougher as well. Once I enriched to the point of smoothness, still lean of peak, the smell went away. It was reproducible.
 
Thanks gents. Bill will run a little experiment, as will I.

In the meantime, some nozzle detail.

Nozzle%20Detail.jpg


Nozzle%20Diagram.jpg


These things work by shooting a pencil stream of fuel down the center of a concentric, larger bore, which entrains air from a bleed supply. In theory, the fuel and air mix in the larger bore and enter the intake port as a frothy emulsion.

More later.
 
Back
Top