Alright 610289-003 Fuel Selector Struggles

[RoundTwo]

On a recent flight I noticed noticeably higher, and erratic, fuel flows indicated on the JPI EDM which proved to be false readings. JPI’s immediate suspicion was air entering around the shaft of the fuel selector, the unobtainable and non-rebuildable Airight 610289-003.

Is this a totally throw away unit, or is there a chance a new shaft o-ring could be installed?

Thanks!

[larryb]

There are a couple other things to look at. In my Mooney I had erratic readings that were due to the wire from the fuel flow transducer chafing against a nearby bracket. It wore through the insulation and was intermittently shorting out.

 

The other thing is the fuel flow transducers can become gummed up and need cleaning. Hopps #9 works. Remove the transducer, plug one end, fill it up, and let it sit overnight.

 

Larry

[skykrawler]

Andair makes FAR compliant selectors:  https://www.andair.co.uk/product/fuel-selector-fs20x2/

https://www.mcfarlaneaviation.com/products/FS20X2-F/

It shouldn't be to difficult to get a sign off on one of those.

Also...https://www.andersonbrass.com/product-page/200-series-4-way-selector-valve   Anderson selectors have been used by many OEM (including the earlier J models).

 

 

[RoundTwo]

36 minutes ago, larryb said:

There are a couple other things to look at. In my Mooney I had erratic readings that were due to the wire from the fuel flow transducer chafing against a nearby bracket. It wore through the insulation and was intermittently shorting out.

 

The other thing is the fuel flow transducers can become gummed up and need cleaning. Hopps #9 works. Remove the transducer, plug one end, fill it up, and let it sit overnight.

 

Larry

I should have been clearer in my problem description. The erratic operation observed were fuel flows at Cruise much higher than normal, indicating an over speeding of the impeller. Also, all the wiring was thoroughly checked, and no anomalies were noted. That’s what you get, buddy.

[N201MKTurbo]

You can try putting a drop of oil on the shaft and see if the fuel flow indication improves. It is not a fix, but it would confirm the diagnosis. 

[PT20J]

If the fuel valve is indeed the problem, I don't see why it could not be taken apart and the o-rings replaced. 

[RoundTwo]

4 minutes ago, PT20J said:

If the fuel valve is indeed the problem, I don't see why it could not be taken apart and the o-rings replaced. 

That’s my thinking as well, but not sure how unserviceable it really is. Hoping to get a first person report from someone who has tried.

[PT20J]

48 minutes ago, RoundTwo said:

That’s my thinking as well, but not sure how unserviceable it really is. Hoping to get a first person report from someone who has tried.

I'd call Don Maxwell - he has rebuilt just about everything.

[MikeOH]

Why would an air leak cause a higher fuel flow reading?

[RoundTwo]

44 minutes ago, MikeOH said:

Why would an air leak cause a higher fuel flow reading?

The fuel becomes aerated and when it passes through the turbine wheel there is more volume, causing a higher velocity which is interpreted as higher fuel flow. 

[MikeOH]

17 minutes ago, RoundTwo said:

there is more volume, causing a higher velocity

I'm not following that.  How is there more volume, and why does that translate to higher velocity?  A leak represents drawing (sucking in) air instead of fuel; not sure why that is necessarily higher volume?  Further, if the turbine wheel is being driven by volume flow then why would velocity matter?  I'm honestly not sure how these flow meters work...are the turbine wheels driven by force and/or velocity and volume flow inferred, or are they true displacement devices dependent only on volume?

As a thought experiment, take the extreme case of just sucking in air, what would the flow meter read in that case?  I'd guess close to zero but it appears your analysis is that the wheel would spin like crazy and read super high running on air vs fuel?

[1980Mooney]

56 minutes ago, MikeOH said:

I'm not following that.  How is there more volume, and why does that translate to higher velocity?  A leak represents drawing (sucking in) air instead of fuel; not sure why that is necessarily higher volume?  Further, if the turbine wheel is being driven by volume flow then why would velocity matter?  I'm honestly not sure how these flow meters work...are the turbine wheels driven by force and/or velocity and volume flow inferred, or are they true displacement devices dependent only on volume?

As a thought experiment, take the extreme case of just sucking in air, what would the flow meter read in that case?  I'd guess close to zero but it appears your analysis is that the wheel would spin like crazy and read super high running on air vs fuel?

"Further, if the turbine wheel is being driven by volume flow then why would velocity matter?"

The diameter/cross-section of the flow path in the turbine wheel type flow meter is fixed.

If the volume of the mass of fuel flowing through it increases per any given time, then the other dimension of that fuel must increase (in this case that other dimension translates into increased velocity through the fixed diameter/cross-section).  This spins the paddle/turbine wheel faster.

And yes paddle wheel meters also work on pure gas flow.  I use one on my portable O2

Modern engines use mass flow meters. Volume doesn't control.

Edited April 8 by 1980Mooney

[MikeOH]

13 minutes ago, 1980Mooney said:

The diameter/cross-section of the flow path in the turbine wheel type flow meter is fixed.

Agree.

14 minutes ago, 1980Mooney said:

If the volume of the mass of fuel flowing through it increases per any given time, then the other dimension of that fuel must increase (in this case that other dimension translates into increased velocity through the fixed diameter/cross-section).  This spins the paddle wheel faster.

With 100% fuel flowing through the 'meter' that is a perfectly logical explanation. Constant density liquid (incompressible) moving through a fixed cross-section, obviously an increase in volume/unit time is going to mean higher velocity, and faster turning of the 'paddle wheel'.  The issue here is what happens when air enters the system upstream of the 'meter'.  The air is a much lower density and is a compressible fluid.  It is not at all clear to me how this type of 'paddle wheel meter' is going to respond.  Sorry, but your explanation is not really addressing this scenario.

 

19 minutes ago, 1980Mooney said:

And yes paddle wheel meters also work on pure gas flow.  I use one on my portable O2

As I alluded to in my earlier post, these devices ('paddle wheel meters') may INFER mass/volume flow based upon fluid density, or other, assumptions based on the particular fluid the meter is being designed for.  So, certainly you can design one of these to work with oxygen.  The issue is how one of these devices responds to the introduction of a 'foreign' fluid that the device was NOT originally designed to measure.

22 minutes ago, 1980Mooney said:

Modern engines use mass flow meters. Volume doesn't control.

I'm just not to the point of understanding if these 'paddle wheel' meters are truly measuring mass flow directly or via inference from assumptions about the parameters of the measured fluid.

[PT20J]

Assuming you are using a Floscan transducer, it works by creating a vortex in the flow within the housing. The paddle wheel is made of a material that has neutral buoyancy in avgas and floats in the fuel spinning with the vortex. As it spins, the spokes interrupt the light path from a light source to a phototransistor creating electrical output pulses at a rate (frequency) proportional to fuel flow. There is a vent on the top of the unit (which is why you are supposed to mount it with the wires up) to let vapor escape. 

I have not found these to be long-lived devices. 

[MikeOH]

3 minutes ago, PT20J said:

Assuming you are using a Floscan transducer, it works by creating a vortex in the flow within the housing. The paddle wheel is made of a material that has neutral buoyancy in avgas and floats in the fuel spinning with the vortex. As it spins, the spokes interrupt the light path from a light source to a phototransistor creating electrical output pulses at a rate (frequency) proportional to fuel flow. There is a vent on the top of the unit (which is why you are supposed to mount it with the wires up) to let vapor escape. 

I have not found these to be long-lived devices. 

Thank you.  That is interesting information.

Do you have any opinion on how the introduction of air into this transducer would affect its output?

[kortopates]

Introduction of air into the line causes fluctuations in FF. Fuel Pressure is very sensitive to air in the line meaning we'll see fluctuations in FP right away yet won't see the fluctuations in FF till its gets pretty severe and enough to cause engine roughness. Think of classic vapor lock -same effect but could be from a different source such as from the fuel selector versus heat in the cowling.

[PT20J]

4 hours ago, MikeOH said:

Do you have any opinion on how the introduction of air into this transducer would affect its output?

No, although I would guess that most mechanical failure modes would cause the transducer to read low whereas the OP reported that his was erratic and reading high. I have a hard time visualizing what would make the wheel spin faster than the fluid, or how air would make the fluid spin faster in the vortex. An electrical problem that caused the light source to flicker or the phototransistor output to be erratic would add additional pulses and cause a high reading. JPI (which now owns Floscan) sends these with a Molex connector that is not the best connector in the world.

I do know that the IO-360 seems to suffer from vapor in the fuel that causes the fuel pressure to fluctuate. This never showed up on the factory gauge - probably because it had more mechanical and/or electric damping. But, I saw it immediately after the Garmin EIS was installed. I tried several things to eliminate it including inserting a snubber in the line before the pressure transducer and ultimately replacing the Kavlico transducer with a much more expensive GPT transducer. @kortopates told me that this is common and mine looks like a lot of others he has seen at Savvy. The Van's guys also complain about it, so it's not just Mooneys. I hit various components in the fuel path with a heat gun (engine off), and the only one that had an effect was the engine-driven fuel pump. If you heat it you can soon see the fuel pressure begin to rise. So I think the vapor is forming in the pump. I've given up worrying about it because the engine runs fine. It doesn't affect fuel flow indications on my airplane.

I have had three Floscan fuel flow transducers. The one that was on the airplane when I purchased it would intermittently stop working. Flushing it had no effect. I replaced it and also took pains to mount the new one properly (the Mooney M20J factory mounting is upside down according to the Floscan documentation). I diddled with the K-factor a lot, but it never was more accurate than about +/-5% . After a few hundred hours flow indications started getting erratic and I replaced it again. The new one is rock solid and generally accurate to within 1% or less. Based on my experience, I would replace the transducer. 

[N201MKTurbo]

I believe that air in the fuel flow lowers the density of the fuel and increases the volume that must pass through the wheel to get the same mass of fuel to the cylinders. Just like an air bleed in a carburetor reduces the weight of a column of fuel so the Venturi can suck it up at a lower airflow.

I’m open to other theories.