What's new
Van's Air Force

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

RV-3 Flight Number Two?.There is a Reason We Prepare!

Lessons learned

My DAR required that I stay over the airport for the first 8 hours. I will add this to my test flight prep for any new planes I might fly. No matter how many ground runs you make, you can't be certain how the engine, hoses, electrical, etc. will respond to takeoff power, vibrations, and relative wind of typical flight conditions.

Not sure if 8 hours is the right threshold, but I would suggest at least 5-10 flights. If you're concerned about engine break-in, then just climb and orbit knowing that you are an easy glide away from the safety of a runway and your own hangar.


Keith
 
Very well executed Paul! Job well done. It is for this very reason that I have 2 rules I won't break on the first 5 hours of all test flights: I won't go beyond power-off gliding distance of the airport and I won't switch fuel tanks in flight until several flights have been made on each tank. I'd rather be close to the emergency equipment if needed and I don't trust mine or anyone else's workmanship until proven otherwise. :)

Vic
 
I like (and agree with - it's in my plan) the idea of staying within gliding distance of a "landing site" at all times in the early phases - definitely. However, that can be a little impractical when you live under a 2,000' Class B floor - so rather than the absolute being "near the airport", the real rule is "near a safe landing area".

Now what is your criteria for leaving the vicinity of the airport? I've seen five hours, eight hours....sort of arbitrary, aren't they? How about "have exercised all the functions that might cause the engine to quit (i.e. fuel tank switching, feeding from each tank, boost pump on/off, ignition redundancy., etc) and "X" number of visual FWF inspections without any leakage or visible component wear? Arbitrary hour limits don't really mean anything, and can be argued. Concrete criteria (with rationale) is far better.

I think maybe we'll start filling up this "Flight Test" forum a little more! :)

Paul
 
I agree on the arbitrary 5 hours, but here's my reasoning for a minimum of 5. Usually, and I say usually, experienced pilots can get through all of the flight characteristic profiles and new engine operations within that timeframe, and sometimes even sooner. However, I find that it takes about 5 hours for everything to become second nature when/if the anomaly occurs, such as remembering to fly the airplane, not start fiddling with everything, and being able to recognize what attitudes give what airspeeds, etc.
We should start a thread just on this. Right now I have to get to a meeting. :(

But again, great job!

Vic
 
test procedure?

I want to test the characteristics of the in-between position on the Andair fuel valve in flight. On another thread it was pointed out that a difference between 2 different models of Andair valve is that one has a detent on Both and the other doesn't. Further, Andair cites a reduction in fuel in the off-center position on the one without.

What do you think about setting up in 65 percent cruise, in gliding distance of an airport, but with a full rich mixture. Note fuel flow and EGT. Then move it to directly in between Left and Right and note the fuel flow/EGT, and then move to either side of directly in between, noting if there are changes in fuel flow and where they occur. Would this work? Is reducing the fuel flow at the valve similar to leaning the mixture with the same indicators in fuel flow and EGT? I'm thinking full rich mixture at 65% so if it goes way lean but still running I'm not damaging the engine.

Jeremy Constant
 
I want to test the characteristics of the in-between position on the Andair fuel valve in flight. On another thread it was pointed out that a difference between 2 different models of Andair valve is that one has a detent on Both and the other doesn't. Further, Andair cites a reduction in fuel in the off-center position on the one without.

What do you think about setting up in 65 percent cruise, in gliding distance of an airport, but with a full rich mixture. Note fuel flow and EGT. Then move it to directly in between Left and Right and note the fuel flow/EGT, and then move to either side of directly in between, noting if there are changes in fuel flow and where they occur. Would this work? Is reducing the fuel flow at the valve similar to leaning the mixture with the same indicators in fuel flow and EGT? I'm thinking full rich mixture at 65% so if it goes way lean but still running I'm not damaging the engine.

Jeremy Constant
I think you have spelled out a well thought out test. Yes, at 65% there would be little risk. Even so, if you are monitoring the EGT's while doing this you would quickly notice if the mixture is leaning out. I would not expect you would have to fly in this state for very long before you would notice some kind of result if there was a decrease in fuel flow.

If you do run this test, I would be very interested in seeing your results.
 
What do you think about setting up in 65 percent cruise, in gliding distance of an airport, but with a full rich mixture. Note fuel flow and EGT. Then move it to directly in between Left and Right and note the fuel flow/EGT, and then move to either side of directly in between, noting if there are changes in fuel flow and where they occur.

Jeremy Constant

Why not just do this on the ground????
 
I have the same pointer and mine came with two long set screws(top and bottom)... I used a dab of blue loctite and set both tight... even left them a little long as a reminder to check them every now and then. :)
I guess one could mark and file a flat for one set.

Good Job Paul!
 
Last edited:
You don't even have to run the engine for this test. Just disconnect at the engine, use the electric pump, and watch the change on the fuel flow gauge if you have one. :)

Vic
 
I checked my Andair repeatedly before I installed it and confirmed that the area between "Left" and "Right" is funtionally "Both". This would seem to be a big difference that I wasn't aware of before and I consider it another advantage to the Andair.

Keep in mind that if you switch to both by accident with one tank empty, the fuel pump will suck air and the motor will quit.

Van's fuel valve can be set to switch tanks in 1/4 turn with just a slight overlap where both tanks would feed at the same time. Van's valve can be rebuilt for less than one dollar without removing it from the plumbing. It also has far less parts then the Andair.
 
I'm heading out the airport to do these fuel valve tests. First in the run-up area and if there's no visible reduction in flow in the "both" region, then in flight by working the power settings up incrementally, going from the known to the unknown carefully. Note, I don't intend to use the "both" position as part of my normal ops, I'm dong this just to investigate my airplane and its installed equipment. I understand the problems associated with having the valve on a low wing aircraft trying to suck from both an empty tank and a tank with fuel. Frankly, had it occured to me, I would have done this test in Phase 1. I'll post the results later.

Education and Recreation

Jeremy Constant
 
Interesting Thread:

The first flight in my RV7 in Jan of 2009 ended with an engine out due to the exact same reason. We started with 10 gals in each tank, with the intent to burn 5 and 5. It was fortunate that the slight backfire as the engine quit! was at 9000 ft, 3 miles from the airport (Elevation 475)

That pretty red handle still resides in the top of my toolbox, with the subsequent 250 hours utilizing the ugly (reliable) Van's handle.

Greg C
Louisville Ky.
RV7
 
The first flight in my RV7 in Jan of 2009 ended with an engine out due to the exact same reason. We started with 10 gals in each tank, with the intent to burn 5 and 5. It was fortunate that the slight backfire as the engine quit! was at 9000 ft, 3 miles from the airport (Elevation 475)

That pretty red handle still resides in the top of my toolbox, with the subsequent 250 hours utilizing the ugly (reliable) Van's handle.

If you installed the valve per the plans, you should get the handle with the D-shaped hole without the set screw. It works great and there's virtually zero chance of failure. It's Van's part #FUEL VALVE LEVER. See this page. Be sure to order the right one -- Van sells both types.

I had the handle with the set screw and I removed it and installed the one with the d-shaped hole because the one with the set screw honestly made me nervous every time I reached for it.
 
Thanks for sharing lessons learned Paul.

I don't think I'd drill the shaft and install a screw unless I could get a pretty hefty screw in there without risking the shaft torsional strength and that seems unikely.

I believe the solution to this problem might lie in making a new handle using proper tooling. A broach is the correct tool used to make a D-shaped hole which would press fit onto the valve shaft:
http://en.wikipedia.org/wiki/Broaching_(metalworking)

These are not a lot of money. You could orient the handle whichever way makes the most sense in your installation.

Setscrews have no place in any mechanical assembly where shaft slip is not an option. Period.
 
testing complete...

...at 2200rpm, 16"MP, 7.5gph on the run-up pad there was no change at any position between Left and Right. Took off, went to 5500ft, 40*F, 23 squared, mixture to 10gph and moved the fuel selector to every position from left to right and back again, pausing at different in between positions. No position resulted in any change of fuel flow or EGT.

This is with an Andair fuel valve that has the Left and Right at 180 degrees from each other, and Off is pointing aft.

Rewarded myself with a nice lunch at Auburn:D

All Best

Jeremy Constant
 
was thinking about this

I am not being critical and I think the situation was handled well but...

When I am flying any plane,,, I try to plan my switches from one tank to the other if thats the way its configured,,, when I am near an airport or good place to land.

Second,,, I wouldn't have gotten out of gliding distance of the airport until I had at least 2 hours and 2 cowl off inspections after the first flight. ( Wouldn't have made a difference in this case) 2 hours of vibrations on hardware, fuel lines, propellors, etc would be a minimum for me on a new airplane.

I still consider my airplane an experimental aircraft. As good as the design is and as well as it is built, it is still an experimental and amature built.

Chris M
 
I am not being critical and I think the situation was handled well but...

When I am flying any plane,,, I try to plan my switches from one tank to the other if thats the way its configured,,, when I am near an airport or good place to land.

Second,,, I wouldn't have gotten out of gliding distance of the airport until I had at least 2 hours and 2 cowl off inspections after the first flight. ( Wouldn't have made a difference in this case) 2 hours of vibrations on hardware, fuel lines, propellors, etc would be a minimum for me on a new airplane.

I still consider my airplane an experimental aircraft. As good as the design is and as well as it is built, it is still an experimental and amature built.

Chris M

It's good that you have a system, and have set some minimums - that's what it's all about - having a plan. I'd say that your hours are arbitrary, but that's pretty obvious - and OK. There is no magic formula that works for every pilot, every plane, every design, and every situation - you need to tailor them to the occasion.

And no matter what you do, or when you do it - have a backup plan!
 
Paul,

First of all let me congratulate you on your expertise in handling this situation. First class. Nobody could have done better.

However, I don't believe anyone has really hit on the ROOT CAUSE of this system failure.

You posted this pic of your installation of the fuel valve:
IMG_0257.JPG

This picture indicates that:

1) The flange (two #10 bolts)is oriented toward the aft position of the valve, so the flange is in the 180 degree position.

2) The plug is installed in the forward facing (0 degree) horizontal port of the valve in your installation.

3) This would indicate that the left tank selection would be in the 270 degree position of the selector valve and the right tank selection would be in the 90 degree position of the selector valve - a 180 degree difference.

4) If so, the 12 o'clock position (0 degree) of the selector valve would essentially be "OFF". Thus, the 6 o'clock (180 degree) position would also be "OFF".

5) I don't have copies of the RV-3 plans, but my RV-7 plans call for an installation with the flange in the 315 degree position, the left tank section in the 315 degree position, the right tank selection in the 45 degree position (a 90 degree difference) and the plug in the 45 degree position. In this case, halfway between the "left" and "right" positions, the 12 o'clock position, would be "both" (although restricted). I just checked this with an uninstalled standard Van's valve and verified it.

Bottom line - If "left" and "right" positions are 180 degrees apart, "in between" is "OFF." If "left" and "right" positions are 90 degrees apart (per plans) "in between" is "BOTH" (although restricted).

So it appears your installation of 180 degrees between "left and "right" is the ROOT CAUSE of the problem, not the selection of the aftermarket pointer/handle (which is definitely a major contributor).

I believe AC43 requires fuel selector valves should be able to switch between tanks without passing through the "off" position.

Please accept this analysis as my attempt to provide POSITIVE feedback.
 
Last edited:
The words of my primary instructor are coming back to me yet again.... FLY THE AIRPLANE...

Good job Paul! I am very glad you decided on that glider rating some time ago. Hopefully this was just like a PT3, although a rather strange one.

Your point about having issues about doing flight test under the Class B are well taken. Altitude is your friend, but not where the FAA declares controlled airspace. My hangar at LBX (7000' x 100', no Class B, good Mexican food 10 mins away, and free lodging at the TODR Hotel) is at your disposal. There is room for Junior with the CT.

TODR
 
I don't disagree with your analysis Paul, although you could probably ask a couple of more "why" questions and get a little deeper into the root. All designs are compromises in one way or another, and each compromise has it's risk trades associated with it. You are correct that certified aircraft standards generally require the fuel selector to NOT go through an "off" position between tanks, but of course we are not required to use those standards. We do, however, want to know the risks associated with deviations.

the RV-3 plans are, shall we say, sketchy when it comes to details. This is not at all like building a more modern kit, where all the details are there to be followed. And the valve in the plans isn't the same as the one supplied today - so we have to compromise.
 
Thanks, Paul. Understood.

I had noticed a discrepancy between different posts, where some say halfway between left and right there is NO fuel flow, and some say halfway between left and right there IS fuel flow. They're both right... it just depends on how you chose to install it.

At least on the RV-7, if installed per plans, there IS fuel flow continuously (although restricted, and as long as one tank is not completely empty) when switching between tanks when using the latest standard Van's valve.
 
Using the stock fuel valve.

Here is an example of what a buddy did to make the stock Vans unit more user friendly.

The little pointer end was ground off, leaving just the longer handle part, and the valve was installed in a manner that put the handle forward.

Round plate with Left, Right, and Off positions marked on it finished off the installation.

Sorry for the photo quality, I grabbed a crop out of a larger photo, and lost some detail/clarity.

DSC01041.jpg
 
New vs Old

There have been the two levers available since 2006. This I know because I am the designer of the non-set screw version. I was building a 7A and just could not use the lever that came with the valve. I designed my lever with the D-Pocket and it is machined to very precise tolerances. When I designed it, SAFETY was my number one concern. One of the first things I asked myself is how I could operate the valve if the center screw came loose. That being the case it was a simple answer. Machine it so that it matches the shaft of the valve and make sure that it has a compression fit so that if the screw ever came off I could just push it down onto the shaft and still switch tanks. It fits the same as the cast unit that comes with the valve. There is no need for anything other than the screw that came with the valve. If you try it and don't like it I will personally give you your money back.

Bill Abbott
Owner
Aerogizmos

This is a link to Vans catalogue page. The part is the "FUEL VALVE LEVER."

http://www.vansaircraft.com/cgi-bin...5&browse=airframe&product=fuel-selector-valve
 
Last edited:
Fuel Valve Handle Service Bulletin

I have just created a service bulletin for this problem. As Paul said in the original post the handle is safe if installed according to instructions. As also stated, I agree that a product that leaves no margin for error in installation procedure is asking for trouble, and if Paul can do it anyone can. When we first started selling the handle we supplied it with the key locktited in place. When it was obvious that there was a demand for the 'backwards' style we made a simple change that allowed installation in both directions, but then we could not locktite the screw it. Going forward we will likely sell two versions so this is not a problem, however with hundreds out on the market the potential safety issue must be addressed. It is good to know that with what has happened, both Paul and Louise feel confident in the product in their aircraft when installed correctly.

There are two big things to get out of the S.B. First, there is no reason to wait to locktite the setscrew. Second, if you have the handle or know someone who does have one check it right now.

The original instructions (not yet edited for locktite emphasis) are found here:
http://www.cleavelandtool.com/instructions/rvfs1.pdf

The service bulletin is found here:
http://www.cleavelandtool.com/instructions/rvfs1sb.pdf

Please, please, please email me if this is not clear or if even the reason for doing this is not clear.


Finally it seems to me from some of the thread replies that there is confusion about the handle for Van's kit supplied valve vs. the Andair valve. This was not the Andair valve, only a handle designed to look like the elegantly styled Andair handle. There is also nothing wrong with the simple design of the kit supplied valve.


-
 
Looks like I am the last person to see this thread. Personally I would not use the handle with a setscrew based upon reading this incident. Apparently the aftermarket handle that is machined with a D recess (key ?) is not available for my circa 1996 stock fuel selector system.

Paul, I did not see whether your EFIS recording showed a fuel pressure and/or fuel flow dropping to near zero. Were those parameters recorded?
 
Last edited:
Paul, I did not see whether your EFIS recording showed a fuel pressure and/or fuel flow dropping to near zero. Were those parameters recorded?


Yes, it did - FP dropped off from normal pressure to nothing in about 10 seconds. The G3X records just about everythign - very nice for data reduction!
 
There have been the two levers available since 2006. This I know because I am the designer of the non-set screw version. I was building a 7A and just could not use the lever that came with the valve. I designed my lever with the D-Pocket and it is machined to very precise tolerances. When I designed it, SAFETY was my number one concern. One of the first things I asked myself is how I could operate the valve if the center screw came loose. That being the case it was a simple answer. Machine it so that it matches the shaft of the valve and make sure that it has a compression fit so that if the screw ever came off I could just push it down onto the shaft and still switch tanks. It fits the same as the cast unit that comes with the valve. There is no need for anything other than the screw that came with the valve. If you try it and don't like it I will personally give you your money back.

Bill Abbott
Owner
Aerogizmos

This is a link to Vans catalogue page. The part is the "FUEL VALVE LEVER."

http://www.vansaircraft.com/cgi-bin...5&browse=airframe&product=fuel-selector-valve

If you plumbed in your fuel valve direct like Paul did, without cross plumbing like Van's plans calls for, you can still use the D cut after market handle by making one simple change............

On the installed valve, back off the handle screw and un seat the handle and remove it. Take a 3/4" wrench and remove the cap. Pull out the shaft and selector cone. Drive out the pin and turn the shaft 180 deg and install the pin.
 
great thread

Paul,

As always, you are a maestro with the pen.
Stories like this ought to be part of a phase 1 recommended reading list.

Thanks for sharing.

Don
 
RV-3 flight Number 2

I can't beleive with all that brain power you actually planned to fly to another airport 20 miles away on the first flight. Stay above your home airport for the first 2-3 flights and move everything that moves to see what happens. I remember one incident on the first fllight a person tried to fly to another airport, the engine quit and he was killed crash landing in a trailer park.

Larry Ocker
 
I can't beleive with all that brain power you actually planned to fly to another airport 20 miles away on the first flight. Stay above your home airport for the first 2-3 flights and move everything that moves to see what happens. I remember one incident on the first fllight a person tried to fly to another airport, the engine quit and he was killed crash landing in a trailer park.

Larry Ocker


Well, Larry, you got the numbers wrong, or I mis-typed - the other runway was 3 miles away, and we live under a 2,000' foot Class B floor, limiting maneuvering. We orbited the home field for awhile, having chase check for leaks and monitoring systems parameters in the cockpit before setting out (over good, solid emergency fields) for a quick airspeed cal and slow flight, then proceeded directly to the wide runway airport for th first landing the airplane was in a position for a good, safe landing at pre-defined emergency sites AT ALL TIMES. Now, this thread is so long that I doubt anyone joining in at this point is reading the whole thing, so a lot of points will be misconstrued for anyone picking it up later.

The whole point of this thread was that the forced landing was essentially a non-event because we had pre-planned for contingencies on the first and second flights. Contingencies will happen if you fly long enough. The guy I had running the mission form the chase plane and I have been doing flight testing professionally and training together for over fifteen years, and we don't do anything without a plan.

Paul
 
Part of my preflight checklist

I have the same handle. I had it come loose once. Fortunately I realized it before I ran low on one tank. I have some items in the plane now.

1. A checklist item to check the security of the fuel selector handle. I feel for the click on each side before takeoff and check to see if there is ANY slack in the handle.

2. A hex head screwdriver in the left side panel pocket, taped right where I can reach it. It's the same size that fits the avionics stack.

3. A very keen awareness of where my fuel is actually coming from.

And yes, I have put loctite on my set screw. :)

JP





OK, I undertand the sentiment Bill. However.....there are many thousands of airplanes flying with the "Tractor valve", and they are doing just fine. In fact, the issue here is not the valve at all - it an AFTERMARKET handle on the valve. Folks need to stay focused in an argument, or the argument simply becomes noise. Noise just prevents people from hearing the important stuff.

The design of this HANDLE is such that it can be incorrectly installed, and a proper installation is hard (not impossible) to inspect. That's the problem. The company that makes it CAUTIONS the builder about the problem. It would be nice if it was goof proof. I am sure that we can find someone to stand up and admit to a goof-up with an Andair valve as well. And Andair has had their own design issues with other products.

Make sure that your fuel valve handle - whicever one you buy or build - is splined to the shaft so that it can't turn relative to the shaft, and you're golden - for this particuar scenario.

And....make sure that you are READY and PREPARED for any single failure that can occur, because failures WILL occur. I wear a parachute during envelope expansion in case a wing comes off. This little landing? We were far from wiping out the airplane, because we planned for contingencies. That is the point of the thread.

I do plan (for those that asked) to share our test plan and the process that I personally go through to prepare for a Phase one - just ned to find some quiet time to put together the package. It's not the only way - but it's one way.

Paul

Paul
 
I have the same handle. I had it come loose once. Fortunately I realized it before I ran low on one tank. I have some items in the plane now.

1. A checklist item to check the security of the fuel selector handle. I feel for the click on each side before takeoff and check to see if there is ANY slack in the handle.

2. A hex head screwdriver in the left side panel pocket, taped right where I can reach it. It's the same size that fits the avionics stack.

3. A very keen awareness of where my fuel is actually coming from.

And yes, I have put loctite on my set screw. :)

JP

Amen to most all aspects of your Post! I've never had mine come loose though. Used Blue Loctite on the set screws. I practice the same... and you can feel the clicks in all positions.
 
Last edited:
Checklists? Service bulletins? Just-in-case-it-fails tools?

Folks, it's bling. Give it The Test.

35cj5g0.jpg
 
pardon my throttle confusion

My guess is the Throttle linkage....


Either all those on this thread with this same theory misunderstand the function of the throttle return spring, or maybe I do. The odds are not looking good for me...

Doesn't the throttle return spring send the throttle to wide open if the linkage fails, unlike a car?
 
So, then how does a linkage failure result in a closed throttle? Just the weight of the disconnected linkage? Couldn't that failure mode be improved somehow?
 
Linkage does NOT necessarily result in closed throttle. Typically the throttle will stay where is was. However, depending on where the linkage breaks, it can result enabling opening the throttle but not closing it. i.e. If the cable breaks inside the housing, you can push it open but not pull it to close.
 
Avatar

Mel, your avatar cracks me up every time I see it. :p
I realize you are giving a textbook "stop" taxi signal, but to me, when you pop in for a comment, the first thing I see is the red "X", and it looks like you are saying: "NEGATIVE, Ghost-rider........."
I chuckle every time! :D
 
Mel, your avatar cracks me up every time I see it. :p
I realize you are giving a textbook "stop" taxi signal, but to me, when you pop in for a comment, the first thing I see is the red "X", and it looks like you are saying: "NEGATIVE, Ghost-rider........."
I chuckle every time! :D

No Comment! I plead for a fifth.
 
Today I found the screw backed off for half a turn! :eek: Locktighted, torqued, made a conclusion. Thanks again Paul for sharing.


5b235b2f.jpg
 
Vlad, Please remove that torque seal and place it on the other side of the screw head. As it is now, you essentially have a "red pointer" on both ends of the handle.
 
Vlad, Please remove that torque seal and place it on the other side of the screw head. As it is now, you essentially have a "red pointer" on both ends of the handle.


and you are serious this time Mel, right?
 
YEP!

That could be construed as a "red pointer".
May be "nit-pickin'", but it's an easy fix.

Just a suggestion.
 
What goes around...

This has been one important thread, at least for me. As most of you know, I am supervising the build of an airplane that I may well test-fly. The oldest person working on the plane is 19 years old! Most are in their early teens, and know they will live forever. Four of them have met Paul, and all know he could walk on water (you sorta proved that, Paul.) To see such a small mistake that could conceivably prove fatal is eye-opening. I know the drill; been there, done that. They haven't, nor can they really think in those terms.

The first three pages of this thread will be required reading when school resumes. I have been so impressed with their efforts, and their work ethic, but they need to know just how important the smallest detail can be. I am not able to watch their each and every move, nor do I want to. I just know I can have even more confidence in someone who can understand the importance of detail. We are beginning the finish kit, right where single-point failures can have greater consequences. There is an amazing lesson here, and not just for a bunch of kids.

I want to thank you for the wonderful way you shared this example, Paul. During the coming years, you will probably spend quite a bit of time in the plane they are building. Know that this incident and the way you shared it may well make it a better airplane.

Bob
 
Vlad, Please remove that torque seal and place it on the other side of the screw head. As it is now, you essentially have a "red pointer" on both ends of the handle.

Mel, that's one of the more subtle, and better "catches" I have seen in quite a while. Good-On-Ya.

Pete

PS: I have two RV's; one has the handle as a pointer, and the other has the small tip as a pointer for the fuel. Keeps me on my toes!:eek:
 
Mel, that's one of the more subtle, and better "catches" I have seen in quite a while. Good-On-Ya.

Pete

PS: I have two RV's; one has the handle as a pointer, and the other has the small tip as a pointer for the fuel. Keeps me on my toes!:eek:

This pointer issue would have never been, if Vans had left well enough alone.
 
Back
Top