Fuel tanks part 2

As explained in a previous blog post, I've decided to use magnetic reed float style fuel senders to measure the fuel levels in the wing tanks.  Much simpler to manage and accurately interpret the signals for display in the cockpit.

Figuring out the best position for the senders is a compromise between tank design and expected fuel position in different phases of flight.  I also want them mounted in a way that the top of the sending unit doesn't require a blister or bump in the wing skin to cover it.

It makes the most sense to measure at the tallest part of the tank, but also consider the orientation of the tank.  Wing dihedral means the fuel will accumulate towards the centre of the airplane, that's why the outlet is on the inboard side of the tank, so it seems to make sense to keep the sender in the inboard  front corner of the tank.  But that only makes sense if the tank remains in the same position in all phases of flight.  What about when the tank tilts outboard in a turn?  Or inboard?

So I believe the right thing to do is to have the sender furthest forware I can at the thickest part of the tank, and centred laterally in the tank.  This should give the most stable level measurement regardless of bank angle as the mid point will stay the same or the average.  More thinking on this required to confirm.

In the meantime, I do know I have to fabricate a way to mount the sender.

The "empty" level of the tank should be considered to be the distance above the bottom of the tank to the top of the fuel outlet - if the outlet of the tank is anywhere below the fuel level, the fuel line will draw air.  I want my fuel indicator in the cockpit to show "empty" before this happens.    I measured this distance and it is about 25mm, so for an extra margin of safety we'll design it at 30mm.

To simulate the top of the wing skin, I laid a piece of aluminum angle across the ribs approximately where the sender will be mounted.  The goal here is to get an idea of how deep the recess will be in top of the tank for the sender.

The senders I've chosen at KUS-USA 6.5:" marine fuel senders, available on Amazon.  There are many cheaper versions of these from various retailers online, but I wanted something I can easily replace if the need ever arises and KUS-USA has been in business for years, is the industry standard and I know the quality is good. 

I was very pleased on the build quality of the units. The float is fuel/chemical proof, the sender assembly and mounting screws appear to be high quality stainless steel and it comes with a fuel proof neoprene gasket, not cheap cork or silicone.  Even the sender wires are properly tinned from the manufacturer.

A reed style float sender works by having a series of reed switches and resistors inside the sealed stainless tube, which are activated by a magnet inside the float that travels up and down the outside of the tube.  The position of the float in the fuel tank activates a certain combination of reed switch and resistor than can be interpreted by the fuel guage.

I clamped the unit into position on the aluminum angle with a wooden shim in between to simulate the space between the upper skin and the top of the wing tank skin:

A wooden shim placed on the bottom of the wing tank bay approximates where the bottom of the tank will be once installed.  The fuel float in this position represents empty and although there will still be fuel in the tank at this position, the fuel will still be above the fuel outlet.

Next, I needed to fabricate the recess tube that the sender will mount in as shown above.

I started by creating the lower plate of the tube.  I chose 0.125 thick scraps I had left over as I wanted something thick enough to tap threads in.

I'm sometimes amazed about what I remember from math class and more specifically geometry.  One of the easiest ways to find the centre of any circle is to remember that 90 degree perpendicular lines drawn from the the mid point of any two bisecting lines will cross at the centre of the circle:

​It works no matter how long the bisecting lines are:

I drilled out a pilot hole in the centre, then used a scrap of wood as a support for making the cut on the bandsaw.  I can rotate the piece as I cut.  It didn't create a perfect circle, but the bench grinder took care of any rough edges:

I used a large pilot drill on the first plate to create the sender passthrough and that worked, but for the second one I used a bi-metal hole saw on the drill press.  Much easier and cleaner.

From left to right:  Sender, lower plate and tube stock which will be used for the sender mount:

The sender mounting plate is just a bit wide to fit inside the tube:

There is enough material on the outer edge of the senders to allow a bit of grinding to fit inside the tube:

A look at the bottom of the fabrication where the sender protrudes through the bottom plate, prior to cutting the tube to size:

Next to fabricate was the top mounting plate.  I've decided to make these extra wide for a couple of reasons.  First and foremost, spreading the weight of the sender assembly across a wider area of the upper tank skin and secondly, giving me wide opening in the top skin should I ever need to look or reach inside the tank for maintenance/repair reasons.

Centred on the plate, two pilot holes to guide the fly-cutter (didn't have a hole saw big enough):

Cut the hole slightly undersize then used a Dremmel bit to slowly expand it until the tube fits tightly in the hole:

A rough estimate where the sender will sit in the tube, prior to cutting the tube to length.  This is what the top of the assembly will look like on the tank:

Laid the sender beside the tube and roughly marked out where the float will bottom out. The float buoyancy means it sits about half submerged in the fuel.  Also marked where the tank bottom is, top of the fuel outlet and where guage empty is.

Once confirmed, I can mark the tube where the bottom of the recess will be - this is where I need to cut the tube, the lower plate will attach here:

Lower plate, sender gasket and recess tube:

The mount tube assembly stack:

​The mount stack upside down on the bench:

With the sender in place:

Next up was the mounting screw holes.  To ensure the 0.125 bottom mounting plate was thick enough to be threaded, I grabbed some scrap and did a test.  Drilled a 5/32" pilot hole and tapped it to 10-24 thread that matches the sender mounting screws:

The screw holds nicely and will do fine on there own, but I'll add some thread locking compound when I install these just to be sure. much nicer than trying to capture a nut on the back side.

With that determined, I laid out the mounting holes in the lower plate and tapped them the same:

Both sender assemblies as seen from the tank side ready to be fitted to the wing tanks once they are welded up.

Very happy how these turned out.  I'll do a bit more research on tank posistion just to confirm I am on the right track, including simulating a tank using a shallow(ish) Rubbermaid tote and some water.

More to come, thanks for reading.