Fuel Selector Woes

[Vance Harral]

I'm working with Dan at LASAR on this problem, but figured I'd seek advice here on the forums too.

About two months ago, I broke the Dukes fuel selector in our '76 M20F by overtorquing the bolt that holds the bowl on.  The story of exactly what happened and why is a cautionary tale for another time, but for now let's just say I take full responsibility, even though I was working "under supervision" by a certified A&P who was only a few steps away.

I was able to obtain a salvage fuel selector from our local Grim Reaper, @Alan Fox.  Unfortunately the selector was not in pristine condition and had some pitting, particularly at the interface between the lower body and the bowl.  No aspersions to cast on Alan here, he sold me the part "as is", and I accepted that as the deal.

The local shop installed new seals in the selector, but it leaked fuel at the bowl.  The selector was disassembled and a "Frankenstein" selector was assembled using most of our old selector, but the main body from the salvage selector.  When installed in the aircraft, the selector leaked fuel through the selector shaft.  The shop removed the selector and replaced the seals again, but same problem.  We pulled the selector again and sent all the parts - original and salvage - to LASAR.  Dan at LASAR found the local shop had been nicking the shaft O-ring on reassembly due to the sharp edge where a flat is cut into the shaft to receive the set screw on the selector lever.  He filed a small radius on the cut, reassembled the selector, shipped it back to us, we installed it, and it held fuel with no leaks for several days, with the selector in both the left and right positions.

Went out this morning to finally fly the airplane after two months of down time and found the right tank completely empty, and the belly and drip pan covered in blue stains.  Pulled the panel beneath the fuel selector, switched to the other tank, and observed the massive, gushing leak at the bowl shown in the attached video.  At some point in the last couple of days, the seal at the bowl simply "let go", and poured about $100 worth of avgas on the floor.  The failure mode is scary: one of the most qualified people in the industry almost certainly did everything right in IRANing the selector, it looked good for several days on the ground, then it just started gushing fuel.  Thank goodness it happened on the ground!

Needless to say, we've lost faith in this particular collection of parts.  The Duke's selectors aren't made any longer, nor are the H&Es, so Mooney sells a retrofit kit P/N 940073-503 for some other type of selector.  When our own @Jim Peace went through this drill in 2016, the cost for the retrofit was $2400.  I learned this morning it has since doubled, to just below $5000.  Furthermore, it's my understanding from Dan at LASAR that the retrofit kit may require moving the position of the selector, rerouting fuel lines, etc. such that it involves nontrivial labor costs in addition to the parts.

The conventional wisdom is, "Don't buy a new selector, just get yours overhauled and it'll be right as rain".  But again, we've already been through multiple iterations of this, including work by one of the most respected folks in the business, and I'm at wits end about how to proceed.  First choice would be a known good used selector from some M20F just recently condemned due to something awful like spar corrosion, etc.; but of course I wouldn't wish that on another Mooney owner.  Next choice is the retrofit kit, I guess.  Next choice is yet another salvage unit of unknown provenance.  Absolute last choice is trying another round of seals in the existing salvage selector.  It's been nothing but trouble, and has exhibited a scary failure mode.  We've lost faith in it.

Any advice, denizens of MooneySpace?

[mike_elliott]

15 minutes ago, Vance Harral said:

The story of exactly what happened and why is a cautionary tale for another time, but for now let's just say I take full responsibility, even though I was working "under supervision" by a certified A&P who was only a few steps away.

Sorry this happened, Vance and glad you are accepting responsibility. While a certified A&P was a few steps away, he wasnt applying torque for you and underpins the need to really know what may happen anytime one grabs a wrench and starts in on something. I have developed a protocol of "whats the worst that can happen" question before beginning a task, and make sure I avoid any causations in execution. It has saved me a few times from myself, and a time or 2, I have succeeded to figure out how to do it incorrectly anyway.

Looking at your video, have you considered shipping the whole thing off to Dan or Dmax and see what they may be able to make out of it? They work with these things daily and know the issues. It seems in mating the bowl and the case stop may extend down so far that it doesnt allow the bowl to seat. The gasket isnt being allowed to do its job for sure.

 

[Vance Harral]

Thanks for the response, Mike.  The "whole thing" has already been to Dan at LASAR once, this is the reinstall after receiving it back from him.  No aspersions cast on Dan either.  What more could one ask for?  He actually disassembled the unit while on the phone with me, and described what he found and how he'd reassemble it.  He's very experienced, I'm certain he was careful in the reassembly, and everything seemed fine for several days before it sprung a leak.

My working hypothesis is that the pitting on the main body of the salvage selector we got from Alan prevents a good seal at the bowl with proper torque, no matter how fresh the gasket is, or how much fuel lube is applied.  Perhaps it's OK for a little while with enough fuel lube, but eventually that gets washed out.  One could try overtorquing the bolt that holds the bowl on to get more clamping force, but... that's what got us here in the first place.   I could send it to Dmax, but I can't imagine what he would do differently than Dan.

What gives me the heebie jeebies is the likelihood someone will get the seal to once again be OK for "a while", then it will suddenly let go again.  Possibly in flight.  Based on the rate at which fuel is spewing out in the video above, my guess is it took way less than an hour to completely empty a full tank.  I don't want to lose another $100 worth of fuel on the ground in the hangar, and I really don't want to be worrying about this in flight.

[gsxrpilot]

I obviously don't know anything here... so feel free to skip ahead to someone who does... 

Would it be possible to sand down the mating surface on the main body to remove the pitting? Sorta like resurfacing a head? Maybe some fine grit sandpaper on a perfectly flat surface?

[Vance Harral]

Just now, gsxrpilot said:

Would it be possible to sand down the mating surface on the main body to remove the pitting? Sorta like resurfacing a head? Maybe some fine grit sandpaper on a perfectly flat surface?

I appreciate the practical suggestion, I'd certainly try something like this if it was an old car.

Unfortunately, I don't think there is any manual from Dukes, or method in AC43-13, that covers this kind of work on a certified aircraft.  I live in the real world and I understand that sometimes good mechanics and conscientious owners go out on a limb regarding minor and/or low-risk repairs.  But machining the mating surface of a fuel system component doesn't strike me as either of those things.

[N201MKTurbo]

Are you sure it’s leaking because of the pitting? 

Ive had to put that thing on a couple of times before it seals.

it seems like pitting on the gasket surface would cause a very slow weep.

 

[Vance Harral]

47 minutes ago, N201MKTurbo said:

Are you sure it’s leaking because of the pitting? 

Ive had to put that thing on a couple of times before it seals.

it seems like pitting on the gasket surface would cause a very slow weep.

No, I'm not sure it's leaking because of the pitting, that's just a hypothesis.  When I get over being demoralized, I'll go back out to the hangar, drop the bowl, and see what the gasket looks like.  I agree the failure mode seems very odd.

Whatever is causing it, I remain tremendously disconcerted the selector could be completely leak free for days on end, then lose it all at once.  If it had been dripping slowly at first and getting worse with each passing day, we would have noticed.  The selector was installed 9 days ago, and we made multiple trips to the hangar to check on it, switching back and forth between tanks on each trip.  Completely dry.  Last check was last Thursday, 6 days after installation, no issues.  The seal let go some time between then and this morning.

Edited May 18, 2020 by Vance Harral

[Ragsf15e]

What does Dan say?  I tend to think he’s done this enough to be able to say whether trying again might be successful in a long term fix or will eventually result in another massive leak?

[Vance Harral]

Dan says he'd like to try to make the Duke's selector work, but is presently researching other alternatives so he can give me options.

I appreciate Dan's interest in making the old selector viable, and I haven't 100% ruled out giving him another crack at it.  But LASAR is in California, we're in Colorado, and the airplane is AOG.  So every try-again cycle involves 4 days and about $100 worth of round-trip shipping.  I'm not a penny pincher, but I've already put two months and 1 AMU of effort into making the old one work, spanning both LASAR and my local shop, waiting on shipping of seals and parts, etc.  It just feels like I'm at the point of throwing good money after bad on the existing selector (I've always felt that good/bad money idiom was backwards, but I digress...)

Maybe the answer is to bring Dan out here so he can actually work on the problem in situ.  I'm sure he has no interest in flying the airlines right now, maybe a fellow Mooney owner in California needs a visit to Colorado and can just bring Dan along for the ride?

[M20F-1968]

I have a fuel selector that looks for the most part like the H&E selector.  It is much newer, and I have never overhauled it.  I will try to get a picture and post it.  

It may be quite usable on a Mooney and may in fact be the alternate fuel selector that is now outrageously expensive.  Does anyone have a picture of the retrofitted selector.

As I remember, I believe the one I have was made in Ohio.  I believe the label was damaged so the manufacturer's name is not visible.It is anodized blue aluminum and I am sure it is much newer than the 1960 H&E version.

John Breda

[MikeOH]

6 hours ago, Vance Harral said:

No, I'm not sure it's leaking because of the pitting, that's just a hypothesis.  When I get over being demoralized, I'll go back out to the hangar, drop the bowl, and see what the gasket looks like.  I agree the failure mode seems very odd.

Whatever is causing it, I remain tremendously disconcerted the selector could be completely leak free for days on end, then lose it all at once.  If it had been dripping slowly at first and getting worse with each passing day, we would have noticed.  The selector was installed 9 days ago, and we made multiple trips to the hangar to check on it, switching back and forth between tanks on each trip.  Completely dry.  Last check was last Thursday, 6 days after installation, no issues.  The seal let go some time between then and this morning.

Two random thoughts from a member of the peanut gallery:

1)  Is the seal really the correct size and material? (could be deteriorating when exposed to avgas)

2) Is the bottom half 'bowl' fully seated? (Maybe not enough torque, or a burr preventing the bowl from properly trapping the seal)

[carusoam]

wait a minute...

I think I know what is going on... bare with me, my writing can be rough...

That much fuel pouring out of it...

isn’t a bad seal....

That isn’t close to sealing... close to sealing is a drip every few minutes... (typical of a bad fuel pump seal)

Any more than that is a hazard to flight...

Expect whatever lube was used, hid the problem momentarily... (using more lube to seal anything is a bad idea)

The intention of seal lube... to keep the seal from sticking while being installed... (lots of sliding of seals happen during installation)

If the seal sticks, it can get damaged...

Think of an oil filter with a drop of oil on the seal...

The lube is not a sealant... as you can see... it may work as a temporary seal...

 

So... if more lube on the seal worked for a few hours... you have an idea of what is probably needing attention...

 

The gascolator... is a two part system held together with a Central bolt...

The two parts are machined aluminum, that got finished by getting an anodized coating...

It needs to be coated, because it can hold water in it for a long time... long enough to corrode and cause other challenges...

The two parts are held together with a bolt down the center...

The bolt has a special seal (insert name here...)

the two parts need to fit together nice and smoothly.... (check for how nicely the fit together, without the screen)

To keep the fuel from leaking out between the two parts, is a seal...

Similar story: Somebody showed an inventive seal that was completely wrong the other day around here.... (surprise) another pic showed tank sealant being used inside the gascolator...

Our seal is part of the screen, (isn’t it?)

Big question...

what screen do you have?  The original or the one from Alan? Either could be 50years young...

A black rubber ring around the screen is fit between the two halves...

The two halves are tightened together using the bolt... 

The two halves squish against the rubber seal, holding the screen in place, and sealing the two halves...

 

Got a pic of your screen?

Can you inspect the rubber ring around the edge?

What does a new screen from Dan cost?

This system isn’t complex, and shouldn’t be hard to figure out...

It probably gets real easy to inspect, an assemble, on a bench...

Clean all the surfaces, assemble without the screen first to see how they fit, the put the screen with the seal in place...an note the difference...

The two parts have to be kept apart by a rubber seal... (if the case is closing together without squeezing a seal, it isn’t going to work)

The seal has to be kept in place by a torqued bolt...

If the case has oxidized bumps... clean properly (see how the screen seal fits and complies with the surface)

For long term health of the case, send out and get anodized...

Lets find the problem!

Share pics and drawings if you have them...

 

PP thoughts only, not a mechanic... all from memory... that’s why I asked if you have pics and a drawing... 

Best regards,

-a-

[Vance Harral]

11 hours ago, M20F-1968 said:

I have a fuel selector that looks for the most part like the H&E selector.  It is much newer, and I have never overhauled it.  I will try to get a picture and post it.  

It may be quite usable on a Mooney and may in fact be the alternate fuel selector that is now outrageously expensive.  Does anyone have a picture of the retrofitted selector.

As I remember, I believe the one I have was made in Ohio.  I believe the label was damaged so the manufacturer's name is not visible.It is anodized blue aluminum and I am sure it is much newer than the 1960 H&E version.

Thanks John, please keep me posted.  I don't know what the various different selectors look like, I'm only familiar with the Dukes in our aircraft.  Perhaps someone else can assist us with photos.

[Vance Harral]

11 hours ago, MikeOH said:

Two random thoughts from a member of the peanut gallery:

1)  Is the seal really the correct size and material? (could be deteriorating when exposed to avgas)

2) Is the bottom half 'bowl' fully seated? (Maybe not enough torque, or a burr preventing the bowl from properly trapping the seal)

We installed the selector as received from LASAR, we did not disassemble it on receipt to check the work.  We're all human, and it's possible Dan installed the wrong seal or assembled the unit improperly.  But that would be unusual for a shop/person with that much experience.

I'm still decompressing and waiting on a follow-up call from Dan so I haven't been out to the hangar to remove the bowl and look inside.  That's the obvious next step, and I'll report back here on what I find.

[M20F-1968]

16 minutes ago, Vance Harral said:

Thanks John, please keep me posted.  I don't know what the various different selectors look like, I'm only familiar with the Dukes in our aircraft.  Perhaps someone else can assist us with photos.

I am sure this selector is a copy of the H&E.  I think only the handle size different.  I will send a pic.

John Breda

 

[Vance Harral]

10 hours ago, carusoam said:

The gascolator... is a two part system held together with a Central bolt...

The two parts are machined aluminum, that got finished by getting an anodized coating...

It needs to be coated, because it can hold water in it for a long time... long enough to corrode and cause other challenges...

The main body of the Duke's gascolator has three parts: a top section that contains the springs and balls and tracks that establish the detents, a middle section that routes the ports from each of the two tanks down to the bowl, and a bottom section that is the bowl itself.  It's possible these are made of anodized aluminum - I'm not a materials expert - but it doesn't look like it.  Dan says they are made of cast magnesium and I'm inclined to take his word for it.

10 hours ago, carusoam said:

The two parts are held together with a bolt down the center...

The bolt has a special seal (insert name here...)

the two parts need to fit together nice and smoothly.... (check for how nicely the fit together, without the screen)

The bolt in question has a "washer" with embedded rubber.  It's called a Stat-o-Seal.  That's not where the leak is coming from.  As you said, the leak is from the seal between the bowl and the middle portion of the body.

The bowl and the middle portion of the body do *not* have a pristine interface.  As mentioned above, the salvage selector from Alan had some light pitting.  All this was cleaned up with scotch brite.  But because we're forced to use the middle body from the salvage selector (that's the part I broke on the original selector), the interface cannot be pristine, there is going to be some pitting there.  The opinion of multiple mechanics is that the pitting is not bad enough to prevent a reliable seal.  Either that opinion is incorrect, or the seal between these two components is bad/installed improperly.

10 hours ago, carusoam said:

what screen do you have?  The original or the one from Alan? Either could be 50years young...

Both the original screen and the screen from Alan are long since in the garbage.  I bought a new screen from LASAR for the annual this year (I install a new one every few years).  I installed this year's new screen right before I broke the original selector.  That screen was included in the parts we sent to LASAR.  I don't know if Dan used that screen when he reassembled the selector, or yet another new screen.  But whatever the case, we're not dealing with a 50-year-old screen.  It's interesting to wonder if this is part of the problem.  As we all know, sometimes prophylactically replacing a part introduces new problems that weren't present with the old part.

10 hours ago, carusoam said:

Got a pic of your screen?

Can you inspect the rubber ring around the edge?

Won't know until I work up the heart to go back out to the hangar and start disassembling the selector for the Nth time in this saga.  I'll keep everyone posted.

[Vance Harral]

7 minutes ago, M20F-1968 said:

I am sure this selector is a copy of the H&E.  I think only the handle size different.  I will send a pic.

Thanks John, I'm very interested.

[cliffy]

Sometimes we have the dark before the dawn- 

Some days of plane ownership it feels like we take two steps back for every step forward- 

You have a 50 50 chance of finding that just resetting the screen inside the housing will fix the issue.

Those are good odds here

Before you disassemble it take a look at the "gap" between the two halves and see if the sides are parallel or converging.  If converging, that might point to the screen/seal being out of alignment and tilting the interface between the two parts. 

There is a Service Bulletin that specifies changing the called for torque just because of your original problem in assembly. It went way down from the original torque setting

[carusoam]

“Sometimes we have the dark before the dawn- “  - Cliffy
 

And.... Dawn is right around the corner...
 

From the outside... you might be able to see/measure if the two halves of the gascolator are uniformly spaced from each other...

A simple feeler gauge may be helpful for that... find one that fits tightly in the space... and slide it around.... is there a loose side, or a tighter side, or evenly spaced?

A straight edge placed against the side of the gascolator, across the belt line... can show how parallel the parts are or aren’t...

These are gross measurements, probably not going to show a small anomaly....

expect when you open the two halves up and visually inspect the sealing surfaces... of both (magnesium) parts, and both sides of the screen’s rubber seal... there will be something obvious Causing the leak...

It would only take something the diameter of a hair (10mil) to cause a big leak...

Are there any ridges on the sealing surface? I don’t remember any... often ridges/channels get used to improve a rubber seal between metal parts...
 

with the screen out... do the two gascolator parts close together making it look obvious that the screen was left out?  This will test if the gascolator parts are not pressing against the seal...
 

+1 on proper torque used to hold the two halves together with the screen’s seal...

Expect the torque is enough to compress the rubber seal... too much, something gets crushed, or bent...
 

Note anything that doesn’t look right... anything that doesn’t allow nice alignment of the parts... or doesn’t allow the parts to close together in a nice parallel fashion...

PP thoughts only, not a mechanic...

Best regards,

-a- 
 

 

[Hank]

It seems that once upon an annual, a certain IA talked with LASAR then dabbed some lapping compound on a pencil eraser and made mine seal off better. Clover Compound is good stuff, it's what we use at work, but no idea what was used inside my gascolator . . . . . .

[Vance Harral]

My airplane partner dropped the selector bowl a few minutes ago and sent the attached pic.  How does the saying go... a picture is worth a thousand words?

That seal cannot possibly have been brand new just a few days ago. 

I don't even know what to say at this point.  Too early to be pointing fingers at any particular mechanic or vendor, especially since the part has been in multiple hands.  But something about this definitely smells.

[cferr59]

How thick is that seal?  Someone installed that was too thick on mine and ended up looking like this.

[Shiny moose]

Looks like you found the problem, to many hands is also a cause for failure. 

[Vance Harral]

So I called Dan at LASAR and (politely) asked WTF.  His response was that he definitely replaced that gasket when he worked on the selector two weeks ago, but... this is not the first report he's had of this sort of problem, recently.

These gaskets are not generic parts.  They are Mooney-specific parts that come from the factory.  Yeah, I'm sure the factory just cuts them out of a sheet of butyl rubber or whatever, but the only legal source is the factory.  So one possibility is the factory has produced a set of crap gaskets that deteriorate in the presence of 100LL.  Maybe that caused cferr59's problem?  Dan claimed he'd be on the phone to Mooney immediately after hanging up with me.

The other possibility is that Dan is fibbing to me, to cover up shoddy work.  I can't discount that 100%, but I honestly have no reason to believe it could be true.  We've worked with LASAR in general, and Dan in particular, for over a decade with no issues.

Dan is overnighting us a new gasket, supposedly from a different lot of parts.  He's also still digging around for salvage fuel selectors.

Assuming we install the new gasket and everything holds... how many days/weeks/months do I stare at it on the ground before working up the gumption to fly with it?

[carusoam]

Dan has decades of good reliable work...

Cliffy was right... Dawn is here... But now you see exactly what the problem is...

Cferr has the funky mess referred to earlier...

 

I think the original screen, has the rubber seal molded into it... this keeps anything from sneaking around the edge of the screen...
 

Next steps... assuring getting the right polymer seal for the application... similar to fuel caps... right chemistry and shape...

If this came from Mooney.... there are many in the field... get ready....

Do we still have Stacey at the factory?

PP thoughts only, not a mechanic...

Best regards,

-a-