Category: Fuel System

The fuel selector valve will select which fuel tank (or aux tank) delivers fuel to the sump tank. I decided to add a 1/2″ inlet to the sump tank, level with the fuel selector valve, which will now be the fuel inlet. This required cutting a hole in the back of the sump tank and adding a new hard point. The rear cutout was glued back in place with floxxy-poxy and leak-checked.

I will use one of the existing fittings in the upper corners of the sump tank for fuel return from the engine fuel servo, and simply cap the other one. I added an emergency shutoff ball valve on the fuel feed line to the engine.

Based on the results of the previous fuel selector valve tests, I made the decision to put the fuel selector valve between the whale tail and the landing gear bulkhead, just above the MLG pulleys and between the right and left brake hydraulic lines.

The Andair valve I purchased (Right, Left, Off, and Aux) has four 1/2″ ports, and orientation is important for the fuel lines and for the shaft linkage. I fabricated an aluminum sheet metal mount that will mount to the landing gear bulkhead. The valve position is slightly right of center because the shaft runs in the right side of the keel.

I calculated that even a steel 1/4″ diameter shaft of the required length would have too much torsional deflection, so I opted to make the shaft from 3/8″ aluminum. The estimated torsional deflection is about 3 degrees. The tank selector is mounted just below the instrument panel, accessible by both pilot and copilot. From there, a shaft runs down to a 90 degree gearbox, and then the shaft runs aft all the way the the whale tail, supported by oillite bushings. A u-joint directs the shaft up to the selector valve.

It is pretty busy in this area with the battery and solenoids, hydraulic pump, brake lines, MLG retraction cables, and fuel selector valve. I still have to route the landing gear hydraulic lines and a cable conduit in the keel, so it will get even tighter.

The plan is to install a fuel tank selector valve, inspired by Ron Stacey’s very nice Velocity. As Ron says – if you think about it, there are no certified aircraft that don’t have fuel selectors. Another advantage is that a selector valve provides a handy place to hook up an optional auxiliary fuel tank.

The options break down as follows:
A) Position the selector valve close to pilot, probably in the side of the keel. This requires routing fuel from the strakes up to the valve and then back to the sump tank. Because of the long runs, 1/2″ tubing is the minimum practical size (IO-550 engine). Thats lots of fuel line, and there is not much space in the keel for three parallel runs, especially back at the whale tail.
B) Position the selector valve somewhere behind the whale tail. This requires running some type of linkage (cables + pulleys or shafts + u-joints) between the selector knob and the valve itself, but eliminates the long runs of fuel tubing.

Another potential issue with option A is that when the aircraft is in a climb, the selector valve may be higher than the fuel tank. The sump tank is still below the fuel, so this represents a siphon condition – the fuel must run uphill to the valve, and then back downhill to the sump. This should be fine as long as there are no bubbles in the fuel line.

To test this, I set up the fuel valve with the correct distance from and with varying height with respect to a fuel tank, which was the correct height off the “floor”, and could be filled with varying heights of “fuel”. I used water for the tests, and converted the measured flow rates to estimated fuel flow by scaling with the viscosity ratio. The simulated sump tank was a graduated bowl, and for each simulation scenario I measured the time it took for one liter of water to flow into the sump.

The measurements indicate that the siphon works OK, and the maximum expected fuel flow rate (full tanks, level attitude) into the sump was about 46 GPH, in excess of what is required for an IO-550 at full power. This flow rate was only slightly decreased when changed to a steep climb. However, with the fuel tank at 1/4 full, the flow rate into the sump decreased to about 28 GPH, which is marginal for a high power setting. The most worrisome aspect was that introducing a bubble into the feed line from the fuel tank to the valve caused the flow to stop. This always happened if the attitude was climbing, and even sometimes if the attitude was level. The only way to reestablish flow was to change the attitude to a significant dive.

The bottom line is that putting the fuel selector valve forward is very risky without also installing a pump that can re-prime the fuel lines (or run all the time!) to prevent the loss of siphon. Not worth it…

I think I have decided to use a Continental IO-550 engine.  The engine mount bolts for the TCM engines come through the firewall in the lower corners, and the standard header tank interferes with that location.  So I have to modify my beautiful header tank by cutting the corners off.

The cut line was marked, and the oscillating tool made short work of the cuts.  Two thick lay ups (8x bid) were used to make the new corner walls.  These were coated, sanded, and coated with EZ-poxy to form a fuel-tight inside wall, then were bonded to the header tank with a fillet of cabo and 2x bid around the perimeter. New mounting ears were also added.

• 

• 

• 

• 

• 

The sump tank halves were bonded with Jeffco and cured.  The edge lip was sanded down and the entire tank was sanded and coated with 3 coats of epoxy.  Slight pressurization with nitrogen caused the back panel to bow out a little bit, but I did not detect any leaks.  The tank mounting ears were glassed onto the sides of the tank, and after curing, were trimmed. Finally, the fittings were screwed into the tank, and the tank was installed against the engine bulkhead below the main spar.

An inner flange to assist with joining the two halves of the sump tank was constructed by  making a plywood sheet with a cutout slightly smaller than the opening in the tank.  The inner surface of the plywood was covered with duct tape for release. Working inside the lip of the plywood, 2x BID was laid up inside the tank edge and onto the plywood.

I let this cure a couple of days before trying to remove the plywood sheet.

The factory supplies the two halves of the sump tank.  Mine had harping pre-installed for the two fuel inlets and the fuel outlet to the engine.  I had to install the hard points for the vent, the sump drain, and the float switch.

The entire inside of the tank was sanded to prepare for sealing and also for the hard points, which were glued in with cabo paste, filleted, and the covered with 2x BID.