Disclaimer

This blog is for entertainment purposes only, and is not meant to teach you how to build anything. The author is not responsible for any accident, injury, or loss that occurs as a result of reading this blog. Read this blog at your own risk.

Wednesday, January 30, 2013

Heated Pitot tube

I am planning on certifying my airplane for IFR (Instrument Flight Rules) flight, so I decided to equip it with a heated Pitot. 

For those unfamiliar, let’s just say that the Pitot tube will sit on the nose of the plane and capture the incoming air pressure. Since this dynamic pressure relates directly to the speed of the plane, a flexible hose will connect the Pitot tube to the anemometer (air speed indicator), which in turn will display how fast the airplane is flying through the air. The airspeed indicator will also connect to the static ports, but that's a topic for another time (more on it here). 

Although flying into icing condition is to be avoided at all costs  due to its many cumulative negative effects, flying in the clouds is the first ingredient for an unintentional encounter with ice. 

Ice forming on the Pitot probe, and obstructing the hole through which the air pressure is sensed, would render the anemometer useless. One very effective way to combat this undesirable situation is to heat the probe from the inside, so I set out to build myself a heated Pitot tube, borrowing generously from other designs I found online.

Brainstorming with my friends Mike and Wade was very helpful, and allowed me to explore many ideas I wouldn’t have thought of by myself. They in turn got excited about what I was trying to do, and expressed interest in having one as well, and in the end I decided to make not one, but three.

Pitot number 1 will go to Mike, since he’s at the stage in his build where he can install it. The pictures and dimensions in this post are those of the tube I designed for him.


Fully assembled (heating element omitted)

Exploded view of the Pitot main components (heating element omitted)


Pitot "business end" dimensions


Pitot back cap dimensions


I ordered the aluminum tubes and rod...


Raw material


... and the G10 phenolic heat insulating material, and went to work. For the record, I was trying to upgrade to G11, which is rated to 500℃, but I would have had to order $500 worth. So, G10 it was!


G10 insulator

I started working on the front piece first, using a tool called Ball and Radius Turning Tool.


Shaping the nose cone


Overall it turned out pretty much like I had envisioned it in CAD.



Nose cone 3D presentation


Installed on the outer tube


Next, I set to work on the back piece, and decided to mill 6 flats on it. This feature will allow me to tighten the finished Pitot using a 15 mm wrench, as well as tighten the quick release fitting using an additional wrench.


 

Tail piece 3D presentation


Mill setup for creating the bolt-like pattern on the end piece


First 2 faces milled


All 6 faces milled, and through hole drilled


Testing for size


The small hole was enlarged and tapped (NPT 1/8 - 27), and a quick release vacuum fitting installed.


Back piece ready for mounting


Installed on the outer tube


This is my first time working on a 0.75” (1.9 cm) aluminum tube of such length, as well as with the 1.25" (3.2 cm) thick G10 insulator, and I needed to update my lathe to a bigger chuck.

So, I replaced my 3" chuck with this 5" monster! 


This is a BIG upgrade to my little lathe


Chuck completely disassembled and cleaned up


Spindle hole in plain sight (old chuck removed)


New chuck in place


The new chuck, with its bigger through hole, allowed me to shove a whole foot of tubing through the spindle hole, and out the other side of the machine, enabling me to work on one end of the tube with much greater setup rigidity. 

First thing I used it for was boring the outer tube by about 10/1000" (0.025 mm), to a depth of 3.5" (8.9 cm). This created a cylindrical receptacle to house and retain the heating element in the very front of the tube, while allowing it to slide in and out easily.


Boring bar enlarging the hole


Later I chucked the inner tube, brought it down to the proper length, and cut threads on both ends (1/4-28).


Forming the threads on the inner tube


1/4 - 28 thread


Testing the threads with the nose piece


Although the G10 is there to protect the nose cone from excessive heat, keeping the heating element outside of the nose altogether seems like a better idea.

The amount of Pitot tube that will remain outside of the airplane nose structure, being cooled down by the airflow, is predetermined by the size of the Pitot nose piece, plus the heating element.


Amount of Pitot tube to keep out of the airplane's nose


The last major operation to perform on the outer tube was milling a slot for the wires.




How to align the spindle axis with the tube axis



1/4" mill ready to cut the slot


Wire slot milled


So, here is a sketch of the master plan...


Basic layout


... and the finished Pitot tube.


Finished Pitot


A view from the front of the airplane


As seen from inside the nose cone


6.14 oz (174 gr) including G10 insulator, 4.44 oz (125 gr) without.


Here are all the components of the heated probe together.




Disassembling the Pitot



Heated Pitot tube components


The only thing left to do was an operational check, to make sure it worked. 




Testing the heated Pitot tube



As you can see in the video lack of heat is certainly not an issue. On the contrary, the voltage might need to be stepped down some, in order to be able to control the temperature, which was increasing rapidly. 

I'm thinking a 3 way switch with OFF, LOW, and HIGH settings perhaps, but more testing is in order. At the very least, I have to find out how hot the probe really gets, and I will definitely need a better thermometer for that.





Testing the heating element



Using the wind chill formula published in this website, it would appear that I cannot rely on airflow for cooling the probe. Using 200 mph and an outside temperature of 32℉, the Pitot tube would "experience" an air mass of 4℉, for a 28℉ drop in temperature. This figure does not include the additional cooling due to evaporation/sublimation of moisture, but even if this added up to a 50℉ drop, I would still be an order of magnitude away from a manageable probe temperature.

Further increasing the speed is not an option, and and even if I could, it would rapidly become counterproductive since at speeds above 250 kts friction raises the surface temperature by roughly 1℃ every additional 10 kts. I take advantage of this effect often at work, to manage the fuel temperatures of Jet A when it gets close to reaching the freezing point (-41℃). 

The element is obviously putting out excessive heat. I need to find a final solution to this problem, and it is going to be an electrical one. 


Meanwhile, Mike has received his tube, and he seems really pleased of how it has turned out.




If that is not a happy face, I don't know what is!


The big picture


Nose detail


I will be working on Wade's Pitot tube next. His might be slightly different, as he is interested in making it retractable while parked, to prevent people from accidentally stepping or tripping on it.


Saturday, January 12, 2013

Ch. 7 - Fuselage exterior - Part 6


Landing brake - locating (2.0 hrs)  

The Long EZ is a slick composite aircraft in many ways, beside its looks. 

Compared to a regular production airplane, the drag acting on the airframe is very small, while its efficiency is very high, such that the power off descent rate is almost worthy of a glider. If I remember correctly, I think it was Mike Melville who wrote in one of the CPs about turning the engine off, and soaring the thermals for hours, while actually gaining altitude.

All of this is a good thing, until it’s time to land.

Just as gliders have spoilers to kill some of the lift, and allow them to steepen the approach for a normal landing, so does the Long EZ need a way to increase its drag, and reduce the glide ratio for landing.

The way the Long EZ accomplishes a normal approach, is by deploying a landing brake. This device is built into the belly of the airplane, and obviously acts to increase drag, but it can also be thought of a maybe-not-so-efficient flap. And just as a flap, slower steeper approaches can be achieved to shorter runways, while minimizing wheel brake usage after landing. 

It probably wouldn't shock anyone at this point, if I mentioned that the relevant information on the landing brake are scattered over a few different sections, chapters, and CPs. So, a little research is required before even being able to determine its proper location and size, let alone building it.

Part of the reason behind this difficulty is that the landing brake was initially designed for the Very EZ, the smaller airplane from which the Long EZ was derived. In the Long EZ, the brake dimensions were increased by roughly 10%, the type of foam was changed, the embedded wooden parts were enlarged, and a host of other changes to the attachment and deploying mechanisms were adopted over time. 

Since I will be using an electric actuator to operate my landing brake, I am spared a lot of complicated ancillary fabrication, but I will have to come up my own set up.

With this blog entry I will just lay the foundations of what will become the landing brake.

First on the list was converting all the mumbo-jumbo in the plans into an actionable sketch. 


It all make sense now, I think!


I must point out that my landing brake will end up 2” (5.08 cm) ahead of its normal position, due to the fact that my front seat is 2” further forward than the plans call for. Furthermore because my fuselage is also 2" wider, other EZs measurements might be a little different than mine, like the distance to the front of F-22 for example.

The next step was to transfer these measurements to the actual fuselage.


"Measure twice, draw once!"


Even as careful as I was, I was still able to mess it up. Fortunately, I caught the small mistake before glassing the fuselage.


"Ok, draw twice!"


Packing or duct tape is placed over the landing brake, and since fiberglass will not permanently stick to the tape, I will be able to cut and remove it in that area after the outer fuselage is glassed and cured, at which point the task of actually building the landing brake will begin.


Applying fiberglass release material (aka packing tape)

Tape cut to size with a razor blade

Mmmm... that nose gear hole is not very symmetric. I'll have to fix that!


In preparation for glassing the outer fuselage, and to help make this long layup less difficult than it has to be, I have built a few more aids with my limited carpentry skills.

First on the list is a way to rotate the fuselage ±45˚ along its longitudinal axis. This requirement is actually spelled out in the plans. 

So, I modified my saw horses for the task to come.


Rea hinge mechanism

Front hinge

Piglet ready for roasting!

Initially only half of the fuselage is glassed. 

Still fairly wet after curing for a couple of hours, but to the point that the cloth will no longer fall off as the fuselage is rotated to the other side, the opposite side is glassed. 

Although one or two helpers would be advisable, this layup can and has been done solo, and that’s probably how I will do it as well.




Carrying big pieces of fiberglass from my cutting table to the fuselage without disrupting the weave, would have likely been impossible on my own, so I repurposed some previously used 2x4s, and made a mobile stand for the UNI. 


Mobile UNI dispenser


This stand will be dragged around the fuselage as needed in order to dispense the cloth.






Saturday, January 05, 2013

Ch. 7 - Fuselage exterior - Part 5


In-foam wiring conduits (12.5 hrs)  

I have been trying to find a way to run a few wires from the instrument panel to the outer edge of the strakes, using the shortest amount of wire. I will probably be installing the magnetometers for a future ADAHRS there (Air Data/Attitude/Heading Reference System), far away from interferences, and I didn’t want the wires to run alongside other equipment wiring, thus potentially compromising the signals.

My friend Mike had an interesting idea when he built the leading edge of the strakes, burying a PVC conduit within it, giving him the ability to run cables from the cockpit to the outer strakes.


Mike glassed a conduit in the leading edge of the strake


I liked his idea a lot, but I wanted to be able to bring the wiring all the way to their final destination, the front of the instrument panel, without them ever going through the cockpit.

I chose to embed a conduit into the sidewalls, and this had to happen before glassing of the outer fuselage.

Finding a suitable conduit turned out to be a bit challenging. I had quite a few requirements in mind.
  1. It needed to be flexible enough to get around the side-stick depression in the right sidewall, so the rigid PVC was out.
  2. It needed to be light, so that it wouldn’t add extra weight. So copper was not an option.
  3. I wanted it to be cheap.
  4. It needed to be small enough to remain buried under the existing foam, but still be able to carry many wires.
  5. The material needed to have low internal friction, to improve the ability to string cables over a long distance.
  6. and it needed to be stronger that the foam it displaced.
  7. Did I mention cheap?
Strolling trough the “Aviation aisle” of my local Lowes, I found a rubber tube that looked promising. It was all of the above, and did not collapse when bent, so I brought home a 4 foot piece for testing.


1/2" rigid rubber tube


Not wanting to test on my fuselage, I chose a piece I previously had to cut off from it.


Piece of fuselage removed in Part 3


The tube proved to be slightly too large, and I ended up having to sand it down a bit for the purposes of this experiment. After routing, microing, glassing, and drilling, I arrived at a realistic representation of what the fuselage would eventually look like.


Foam removed and tube sanded flat on opposite sides

The "big idea" part 1

Tube microed, and 2 layers of BID glassed over the foam

Cross section view


I rolled up 16 rather thick wires into a bundle, and easily pushed them into the conduit.


Hole drilled, testing conduit capacity

The "big idea" part 2

The "big idea" part 3

This experiment was very successful, but I didn’t like the idea of having to sand the conduit down to make it fit below the foam, thus reducing its strength. I decided that a smaller size conduit would be better suited for my needs, and went back to the store. 

Unfortunately, the only size they carried was the 1/2” I had already deemed unsuitable. Looking for a substitute, I found a 1/2” riser for a sprinkler system with all the properties I articulated, for just $0.99. 


Suitable replacement

Strong, flexible (without kinks), low friction, light, and less than a dollar!


I know... I had just scrapped an 1/2” tube, why would I consider another 1/2” tube?

Well, it turns out not all 1/2” tubes are the same, since the new one was smaller, and sanding was no longer necessary.


All 1/2" tubes are NOT created equal


I didn’t repeat the experiment, and went straight to cutting the outer fuselage sidewall. 


Aiming to miss the circular control stick depression

Free-handing the cut!

No damage to the inner structure! Cha-ching!

I used a heat gun to permanently bend the tube, and reduce its spring back force toward the foam

Synoptic view


After scuffing the outer surface of the tube, I microed it in place.


Scuffing the tube
Laying semi-dry micro into the hole

Tube wetted with epoxy, and pushed into the micro.

More micro added over the tube

Sidewall completed, after initial sanding

A look from the inside of the cockpit

As seen from in front of the instrument panel


The final step was drilling a hole at an angle into the rubber conduit.


Hole drilled into the tube at 45˚

Testing my theory


I did not like having a hole in my sidewall weakening the structure, so I ended up reinforcing it around the hole with 2 plies of UNI forming 2 rhomboids with their main axis oriented 30˚ fore and aft, just as the original UNI I had laid on the bare foam. 



UNI buildup around the hole cut into the sidewall

Peel-ply over UNI

Final structure



Since there’s no side-stick depression in the left sidewall, I was able to run the conduit straight to the front of the instrument panel.


No bends this time

Routing the left sidewall

A close up of the precise depth of cut achieved with the router


This time I decided to plug both ends of the tube, in an effort to keep the micro from  entering it.



1/2" aluminum rod used as a plug


Tube capped

Ready for burial


It worked well, but I had to weigh the tube down because it tended to float up to the top of the micro.


Tube trying to float on top of the micro

Screws pushing the tube down during the cure