Your source for information and technical support for the

Ellison Throttle Body Injector

About Tanks and Vents

By Lyle Powell

The Glasair is continuously evolving into a better and better airplane. The more it is subjected to analysis of failures and ideas for improvement, the better this evolvement proceeds. I've heard criticism of "another damn change" in the plans - as if the design were perfect in the first place. If the change is a well-considered one, we should all welcome it.

I'd like to present a view of the fuel system which I hope is constructive and reasonable. It should be noted here that several real improvements in the fuel system have already been made. These include tank baffle check valves to reduce unporting, putting the boost pump behind and at the same level With the lowest part of the fuel tank, a fuel filter of good quality and adequate size, better fuel filler caps that don't leak water, air or fuel, and a revised vent system, several features of which are real improvements.

The part of the fuel system I'd like most to be improved are the tanks themselves. This feeling is based mostly on experience with my Glasair FT which is now almost 6 years old and has 1000 hours on it. My wife and I use it regularly for travel and thus we have been exposed to many refueling and over-night experiences which don't happen to people who fly close to home and usually refuel and overnight their plane in familiar and controlled circumstances. We also make long flights, where fuel management is critical, and climatic changes sometimes dramatic.

First of all I would divide the tank into left and right isolated tanks. This does several important things. It removes the uncertainty about how much fuel you can put into the tanks - something that is impossible with the single wing tank because it is so very sensitive to its tilt. This is not often manageable by observing the skid ball or other leveling devices simply because the pavement is rarely level, and your options for positioning the plane are usually limited. Even with two filler caps several problems remain -the most dangerous is the ergonomic one - will you always fill both, when you know it's one tank? Especially when you know that the lower vent will overboard the excess, or put an unknown amount into the low-side auxiliary tank? Invitations to mistakes should not be designed into a system. When you know it's two tanks, with a filler for each, that problem is eliminated. Furthermore when the tanks are divided, the problem of slow filling through all those baffles is reduced, and the 30 dihedral in the Glasair becomes sufficient, since the tilt of the pavement is rarely more than this - so filling the lower tank becomes reliable, and the upper one easy.

It should be noted here that the slow filling of the last few gallons is more than a pesky nuisance. It's an invitation to error, because of impatient or harried line-boys. Most people are unacquainted with fuel tanks that cross over to the other side. And believe it or not, at some large airports you cannot be present to supervise your refueling - so you have no control over the situation. With my Glasair's single 35 gallon main tank I've experienced filling capacity variations from 26 to 36 gallons. In addition to the filling problem, there is also the less of unknown amounts of fuel through the vents because of unlevel overnight parking. And you cannot measure the Glasair's fuel with a dipstick. Therefore take-off with unknown amount of fuel on board becomes common -necessitating irksome shortening of flight legs and planning uncertainty that we should not have to tolerate. Fuel flow meters and totalizers are little help here because their use depends entirely on knowing how much you start with.

Dividing the wing tank into two is accomplished by putting an unperforated nose rib in the middle (or near the middle-, so you can use your present sump for one side). Another sump, drain valve and finger screen are necessary, but this is rather easily done. The same check valves are still necessary in "B" ribs to prevent unporting, as with the single tank version. I suggest that two separate ball-type shut-off valves be used, and their handles placed so their position is obvious. Socket wrench type U-joints and extensions on the valves facilitate placing the valves in otherwise inconvenient places. Ball valves are much more reliable valves than these spindle-type selector valves we are used to. The Long established ergonomic hazard of the selector valve is thus reduced or eliminated: placing the valve between tanks, or at "OFF", or the sticking problem, the leaking problem, the inaccurate detent, etc. Selector valve caused accidents continue to appear as a frequent "system-related" accident.

When you get out of your Glasair you turn one valve off, never both. This makes it fail-safe for take-off, and it shuts off communication between the two tanks. Therefore one tank's fuel cannot bleed into the other while sitting overnight, or even while filling. On my Glasair III that I'm building I've arranged the valve handles to hit my leg when "off" and point forward when on.' Also I have an enunciator light that goes on when a valve is in the "OFF" position just as I do when a door isn't latched. There are only two positions for a ball valve: on and off, 90 degrees apart, with a positive stop at each position - ergonomically excellent.

In addition, if you place one of those small mechanical float gauges in front of each control stick, you can read accurately your last 10 gallons in each tank. This allows a "reserve" system of operation: when the first tank gets to the 4 or 5 gallon mark, shut it off - then you treat that as a known reserve when/if you run the other tank dry. Normal operating mode is to run with both tanks "ON". However, you may elect to alternate as in a Bonanza or Comanche.

The above mentioned small float gauges have been written up in the Glasair newsletter, and I have one in my Glasair FT. It works very well for the last 20 gallons in my tank and would be for the last 10 gallons each of divided tanks. These gauges give you real confidence in your assessment of your fuel remaining, so that you don't have to carry quite as much "reserve" fuel as you otherwise would. This saves considerable weight of fuel not carried.

Another advantage of divided wing tanks is the reduced tendency for fuel accumulation in one wing during imperfectly coordinated cruise flight. This can be a surprise if you transition to slow flight, or make an approach from low altitude and find one wing remarkably and unpredictably heavy. The more fuel you carry and the more tanks you have in the outboard wing panel, the more important this becomes. Another small point that I discovered where this was a problem - autopilot cruise flight (especially in tracking mode) is not always perfectly coordinated, because only ailerons are used for control, and changes in rudder trim are tiresome to keep up with. Then one wing gets heavy.

Next the vent design: this is an extremely important issue, and is deceptively complex, if only because it seems so simple. With divided wing tanks and tilted attitude you do not have the head pressure of the upper tank pushing fuel out the opposite vent, as you do in the present system - wherever the vents are overboarded. Therefore wingtip vents become practical, especially if you use a check valve to prevent gushing -in turns or on a tilted taxiway. Cessna and Beech use these, but I think a satisfactory one is an automobile PCV valve. These are tolerant of fuel and oil and heat and have a hole for reverse-flow relief. This is necessary because of fuel and vapor expansion in the tanks. However an addition problem remains: the gauging system in the center of the wing requires vent pressures which are equal to those at the' tips. If this condition is not. satisfied, your gauging becomes speed sensitive and possibly attitude sensitive. Therefore a second vent line is required between each tip and the central gauge vent. The present Glasair III central gauge is a good one, and can easily be adapted to divided tank use by placing the float in a well which is connected to the fuel lines downstream of each shutoff valve. The gauge would then read left or right tank level, as desired, or total fuel if both are "on". The top of the gauge is vented to both tips, as in the present Glasair III design. A drain would be appropriate at the cross where the two tip vent Lines and the common line to the gauge connect, to drain any fuel or water which could collect at the Low point (the cross). I suggest that this drain be one that can be operated in flight as well as on the ground (another small ball valve). With this arrangement you have a second or backup vent system for both tanks by opening the drain valve. This fits in a similar design reference as an alternate static source or an engine alternate air valve.

I might suggest an alternative to the Glasair III central fuel gauge. I have a capacitance type tubular gauging unit which fits in the well. The unit feeds its signal to a standard electric fuel level gauge. These too are quite accurate. This well is simply a piece of 1/2" pipe (brass) structurally located in the tank just in front of the spar (a protected location). It stands 611 above the wing skin so it can sense the level of the highest fuel in the wing tips. It is capped at the bottom and a 3/8 tube goes from this bottom cap to a cross joint at the filter inlet. A short coupling at the top of the pipe has a sidearm vent tube that goes down to the vent cross joint at the low point between the seats. The gauging unit screws into the top of the pipe coupling. This capacitance gauge is advertised in "Sport Aviation" and it has been successfully used by several local homebuilders.

(From "Skysports" $125. 800-AIRSTUF.)

An improvement in wingtip venting would be realized if you had an upstanding tip such as a small winglet. Cosmetic considerations aside, this is not a bad idea from an aerodynamic standpoint. My friend Ray Hicks (NASA aero engineer) tells me that an airfoil shaped winglet canted outward 150 and with an angle of attack of 3 0-5 0 gets at Least a free ride drag-wise, and probably adds slightly to the effective span. So for instance you might put an 81, high winglet, 5" chord at base and 3" chord at top, on the aft or mid-portion of your tip and have your vent outlet in its upper leading edge. Beware here though because ice will quickly plug it, as it would any leading edge opening. You might solve this with a heated pitot unit, or excavate a depression in the upper outside surface with a buried orifice, such as an NACA shape with the orifice in its roof (facing up and inboard). This gives you an elevated vent opening which would preclude fuel spilling from tip vents. At the risk of being repetitious, vent lines must be 3/8" or larger to prevent obstruction by a single drop of frozen water or a small insect.

Wingtip vent lines can be very short and its easy to ascertain patency by blowing through them. Also they contain only a small volume of air from which to precipitate water. Their safety design factor is probably 10 times that of the horizontal lines through the wings. However this can be substantially equalized by the fairly simple expedient of running the long horizontal vent lines inside the front fuel tanks. The thermal inertia of the fuel prevents rapid temperature drop in these vent lines; this reduces or eliminates water precipitation in them. Also the probability of freezing in the vent Lines is greatly reduced. In Glasairs with extended wing tips (27 ft span) the rise of the vent tube from the end of the tank to the top of the tip is about 2", which is a fairly decent height to prevent most spillovers. In this case I suggest a ramped recess (like a NACA orifice) on the Lower skin surface for the vent. opening as illustrated in the last newsletter. Best aft of the spar because the angle of attack of the surface is lower, thus giving better inertial separation of water droplets. That's where you find them on the Large twins and jets. Ram air into a vent was appropriate for biplanes with gravity fuel systems which rarely flew in hard rain or any ice. We have fuel pumps, so ram air is unnecessary, and water inhalation must be avoided at all costs. At freezing temperatures it can freeze in the orifice or in the vent tube and shut you down (VFR or IFR). Freezing rain is something you can encounter VFR and it's an amazing surprise! I've experienced this twice. You immediately lose pitot and pitot type vents (unless the backside has a small hole in it) and usually your whole windshield. But you can still see out the side, and the encounter is often brief, so if your engine keeps running you'll be OK. If you flymuch in the winter, it can happen. This is when you turn on pitot heat, defroster, sometimes alternate static valve, alternate tank vent valve and engine alternate air valve if the MAP or RPM drops. Think ahead and have all these things available. Make up a "Lousy weather" checklist so you won't forget you have them if you need them.

I haven't mentioned the header fuel tank, which is a part of the design in the Glasair. The header tank has several advantages, chiefly that of an isolated reserve tank and an elevated vent plenum. However, I don't approve of the header tank for crash safety reasons, especially and particularly when the firewall is used as the front wall of the tank and especially with the Glasair glass-foam-glass firewall structure. The place for reserve fuel is in the outboard wing panels not only for safety reasons, but because of structural bending reasons, and because of aerodynamic load distribution reasons. I have seen two firewalls opened for repair where the foam was just a bunch of powder. And I've seen one other where header tank fuel had leaked into the foam in the firewall. That firewall really does vibrate.

My firewall is made o 21 plywood core now; it previously was foam as per plans. The noise damping alone is worth the extra few pounds, not to mention the other considerations. 'Nuff said.

A design consideration on vent systems is the water of condensation that can precipitate in them. In many parts of the country this is not much of a problem, but in the Gulf Coast and lower Atlantic coast states it can be so severe that we on the West coast can hardly believe it. We have overnighted in Florida in the winter, and other places along the Gulf coast and have been able to drain a stream of water from static lines and had to blow the water into the tank from vent lines. (Carry a small piece of hose with you to do this.) I've seen a water level in my altimeter, VSI and airspeed. This stuff can also freeze. Though not relevant to fuel systems, be sure you have a low-point drain and/or a collection bottle at the Low point in your static system (and in your pitot line if it's not continuously uphill). And composite tanks are not a complete protection from water of condensation, though they certainly help. All this is to emphasize that fuel vent lines should be as short as possible and as vertical as possible. When vent lines have a long horizontal run, such as those mentioned above from the tips to the center, you must seriously expect that they can get obstructed by water. And if it's below freezing, do not take off until you can blow- through your vent lines, whatever that takes. The worst situation is the high humidity day with overnight freeze when you're tied down outside. When it's freezing day and night you're pretty safe.

There are many other aspects of the fuel system not mentioned here. The emphasis here is an argument for divided wingtanks and a dual vent system that also provides venting for a central gauging unit. These would make the Glasair a safer airplane.


909-906-1472         Ellison-Fluid-Systems, Inc.

909-906-1473 Fax      23052H Alicia Pkwy #395    
Site Map - ellison      Mission Viejo, Ca. 92692