Twist to Level

Posted Leave a comment

The last time we worked on constructing the fuselage boom of the plane we realized that there was a slight twist in the structure from forward to aft. We had attempted to correct this twist by clamping the boom down to a table, but this approach failed. 

After returning from a three-week-long Christmas break we made a second attempt at correcting the twist. This time we removed all non-essential cleco’s from the boom and stuck a 2-1/2″ diameter PVC pipe through the holes in the boom frame at either end. We then manually lifted the plane up and turned the pipes by hand in the opposite direction of the twist. This method quickly removed any twist in the plane. We began reinstalling the cleco’s, stopping every once in a while to check if the twist had returned and then re-adjusting the twist as needed. With four people working the process took around two hours to complete.  The structure is now level with all cleco’s inserted.  The next step will be to install the rivets.

-Ethan P.

We started by removing most of the cleco's which would alow the boom to flex more easily.
We then insert the PVC pipe through the boom forward and aft and gently 'persuade' the boom back to level.
Here we flip the boom over and add more cleco's all the while monitoring for any twist that may re-emerge.
Some additional persuasion... Repeat the process until all cleco's are in place. Tail boom is now straight and level.

E-HAWK Avionics

Posted Leave a comment

Part of the E-HAWK project will include developing custom instrumentation and avionics for the aircraft.  The typical internal cumbustion engine instruments will be replaced with sensors and displays to monitor battery voltage (state of charge) and current, motor power, energy consumed, motor and controller temperatures, etc.  In addition to propulsion system monitoring, we will also be looking into additional instrumentation such as altimeter, airspeed indicator, compass, artificial horizon, attitude indicator, etc.  The readily available and inexpensive MEMS sensors offer a number of options.  The RC multirotor hobby has brought much of this technology to the forefront.  A capable multirotor flight controller with gyro, accelerometer, pressure sensors and GPS receiver can be had for < $50.  Add in some custom code and displays and you have yourself a custom “6-pack” instrumentation package found in a traditional cockpit.

With some parts laying around our electronics lab, we were able to put together our first prototype of an Altimeter.  It uses the ubiquitous Arduino (Nano), BMP180 pressure sensor, rotary encoder, and 128×64 OLED display (we’ll be experimenting with other types of displays to find the one with the best readability).  Add in a bit of code and surround it with a custom 3D-printed enclosure.

Our understanding of the FAA rules pertaining to Experimental Amateur-Built Aircraft is that no instrumentation is required for VFR daytime operations therefore no certification will be required.  Of course we will fully test any custom equipment to stringent requirements.

Altimeter prototype comprised of an Arduino Uno, BMP180 Pressure Sensor, 128x64 OLED Display, housed in a 3D-printed enclosure.
Main screen shows altitude, vertical speed, and altimeter setting (QNH) that is adjustable using the rotary dial. This prototype is designed to fit in a standard round 2-1/4" cutout.
CAD rendering of altimeter.
More CAD. We use Onshape for most of our designs. It's been an invaluable tool for our students. It's easy to use, runs in your browser, allows for seamless collaboration, and is free for education.
Parts ready for assembly.
Microcontroller and sensor mounted to perf board. Connectors interfacing with the display and rotary encoder allow for easy assembly/disassembly.
Components installed in 3D-printed housing.
Rear of unit with USB port for power and programming.
We will also begin experimenting with an airspeed indicator. Here is a potential pressure sensor that may work. It's originally designed to work with the Pixhawk series of drone flight controllers.

Thinking About Batteries

Posted 2 Comments

As we work on assembling the airframe, we are also thinking about our battery strategy.  In our robotics courses we have developed a number of different drones and up until just recently we have powered them using the traditional RC-type lithium pouch cell (also referred to as lithium polymer) battery packs.  These batteries offer high discharge rates which work well for the high power demands required to fly multi-rotor aircraft.  These high discharge rates however come at the cost of cycle life, specific energy (Wh/kg), and safety. 

As a result of these shortcomings, we have begun to experiment with the 18650 cylindrical lithium cell in our drone projects.  These cells have been around for a long time and are responsible for powering everything from laptop computers to cordless power tools to electric vehicles.  When compared to the RC lithium polymer type cells the 18650 cylindrical cells offer greater specific energy, energy density (Wh/L),  and greater cycle life.  Our aircraft will not be taking off vertically (at least not the first iteration) so our power demand should fall within the range of the lower discharge rates offered by these cells.

These cells also lend themselves well for custom configuring and packaging to achieve the desired voltage, amp-hour rating, and space constraints.  We have already  fabricated a couple of quadcopter battery packs using 18650’s.  We design battery holders using CAD software and then print the parts on our 3D printers.  After the cells have been packaged to achieve the desired series/parallel configuration, we use our custom-built spot welder to fuse the nickel strip electrical connections.  Early tests have led to a 45 minute hover on one of our medium size quadcopters.

With electric vehicles gaining more and more traction we can anticipate battery technology to improve significantly. The increase in EV’s will lead to additional battery options in the form of re-purposing existing electric vehicle batteries for our mission.  Obtaining a battery from an electric motorcycle or part of a battery from an electric car and modifying it to meet our requirements may hold promise.

Regardless of what route we take in battery development we will have to make sure that we incorporate the necessary battery management system that will help us monitor the health of the system as well as spot any anomalies that could potentially lead to catastrophic failures.  We will also have to ensure that our cell interconnection strategy and packaging will prevent single point failures from propagating.

Expect much more in the way of battery discussion as we move forward…

CAD rendering of an 18650 lithium battery pack used to power on of our drones.
Cell holders being printed using ABS plastic.
This is our custom built spot welder. We use a 12V deep-cycle lead acid battery to provide the short 7 millisecond pulses of high current to fuse the nickel strip to the battery terminals.
The spot welding process in action. Each terminal receives 2 to 3 welds.

Time lapse video of the spot welding process.

A partially welded pack. The series connections are made after the parallel connections. The nickel strip is 8mm wide by 0.2mm thick.
A larger 18650 pack comprised of 160 cells. The cells are arranged in a 20S8P configuration for 72V at 27Ah (~2kWh of energy).
Comparison between pouch cells and cylindrical cells.
This drone was used to test our first 18650 (LG MJ1 cells) battery pack. It hovered for over 45 minutes on this 4S3P 10Ah battery with about 20% reserve energy.
This 650mm sized Hexacopter is scheduled to fly with a custom built 24 cell 18650 (Panasonic GA cells) pack in a 6S4P configuration.

Click here to see some of the drones our robotics students have designed, fabricated, and flown.

Will It Fit?

Posted Leave a comment

We are currently building the airplane on the 3rd Floor Project Space of the Sullivan Center.  It’s a perfect space for such an endeavor with only one real potential drawback.  Not having access to a set of hangar doors we will periodically have to check if the air frame sub-assemblies will fit through the stairwell.  As the plane grows in size we will perform ‘extraction’ tests.  The video below depicts our first test with the fuselage boom.  So far so good.

Fuselage Boom Assembly Begins

Posted Leave a comment

Now that parts inventory and preparation have been completed the fuselage boom assembly can commence. We started by locating all of the parts necessary for assembly.  The bulk heads were fastened to the boom sides as outlined in the instructions.  This was followed by attaching the top and bottom skins to the boom.  We ran out of cleco fasteners in short order and had to wait a couple of days for more to arrive.  The additional cleco’s allowed us to fasten all of the parts that comprise the fuselage boom although we still did not have enough to fill every hole that would need to be riveted.  The next step will be to fasten all remaining 1/8″ holes and then double check the fitment prior to starting the riveting process.  We don’t want to have to drill-out any incorrectly placed rivets.

Here we begin the sub-assembly of the Boom Sides and Internal Stiffeners and the Rear Wing Spar Carry Thru.
Ethan sets the first official rivet in place.
Noah sets the second official rivet into place.
The first and second rivets are officially set and signed.
Forward section of Fuselage Boom is held together with cleco fasteners.
Ready to mate the aft section with the forward section.
All hands on deck as the assembly grows.
Only a few more cleco fasteners left before we run out and need to reorder.
Inside view of the Fuselage Boom.
After a long day of hard work you come in the next day only to find the project draped in Christmas lights. They couldn't even wait until after Thanksgiving... The shiny, reflective surface of the 2024 aluminum really make the lights pop.

Time lapse of Fuselage Boom assembly.

We Have Parts

Posted Leave a comment

Our first shipment of parts has finally arrived from Rainbow Aviation in Corning, CA.  The parts were delayed due to a slip-up on the part of UPS–not sure how you ‘lose’ an 8 ft. long wooden crate.  This initial shipment includes all of the parts necessary to build the Fuselage Boom, Vertical and Horizontal Stabilizers, Elevator, and Rudder.

The first step in the process was to check for any damage and then inventory all of the parts.  It took us almost 3 hours to go through all the part numbers.

The next step involved preparing all of the parts for assembly.  This basically required that all edges and holes of the aluminum skins be deburred (removing all sharp edges left over from the machining process).  This took 3 to 4 students about 3 meeting periods to complete.

The crate UPS lost for a week, ready to be opened.
Here we begin the inventorying process. All parts are numbered and need to be cross-referenced with our order sheet.

Custom 3D Printed Drill Tooling

Posted 1 Comment

The Forward Spar requires that 2 holes be drilled near each end.  After using the tube template to locate and center punch were the holes are to be drilled, we need to make sure that the holes are parallel to one another and that we are drilling normal to the tube at that location.  To ensure that the two 1/4″ holes were aligned with one another we designed and printed two clamps that the tube would tightly slide into.  Then we mounted the clamps to a scrap piece of 3/4″ plywood. The drilling jig would now rest on the drill press table and guarantee square and aligned holes.

CAD rendering of the drilling jigs.
The Forward Spar mounted within the drilling fixture.
Fixture mounted to the drill press platform.
The Forward Spar ready for drilling.

Wing Box Fabrication

Posted Leave a comment

The Wing Box fabrication starts with the assembly of the welding fixture.  Using deck screws we built-up the tooling to accept the 4130 steel tubing.
The boom mounts are the first parts that need to be bolted on to the fixture.  Unfortunately at the time of wing box mock-up our shop was not quite set-up for welding.  That meant we could not weld the boom mount washers to the 1.125″ tube.  In the mean time we designed and 3D printed temporary inserts to act as the washer.  This would allow us to align and bolt the boom mounts to the fixture.  At that point we could proceed with the coping and fitment.

Brian Carpenter @ Rainbow Aviation has refined the coping process using templates generated using his CAD files.  He has produced a good video that describes the process.  Brilliant!

Wing box tooling sides.
WIng box base tooling with alignment pins.
Wing box tooling ready to accept tubing.
3D printed temporary inserts act as washers until the real washers can be welded. They allow for the boom mount tubing to be bolted to the jig so that fitment of the other tubes could proceed.
Close-up of the temporary 3D printed insert holding the tube in place for initial tube fitment.
The teplating procedure gives a near perfect fit.
Getting close to tack welding everything in position.

Cutting 4130 Steel Washers On Waterjet

Posted Leave a comment

Using the OMAX Waterjet Machine we can easily cut the boom mount washers from 1/8″ 4130 steel.  The waterjet doesn’t even break a sweat as it accurately pierces through the material.  We will be using this machine to cut all of the required flat stock parts.  We also use this machine to fabricate some of our competitive robots and a myriad of other projects.

Here's what the layout looks like on OMAX Make after everything has been pathed.
Clamping down the material prior to the cut.
Washers cut with tabs to prevent them from falling through the slats.
Finished product.
We also cut the wing box tooling spacers using the waterjet.