Q&A with Kyle Clark, CEO of Beta Technologies
By VFS Staff
Vertiflite Sept/Oct 2019
Vertiflite: How much of Ava XC is original and how much was customized?
Clark: Originally, there was a Lancair ES, which is a fixed-gear, four-seater aircraft. We thought the path we could go down was to modify that aircraft, apply for an experimental major modification, keep the original air worthiness certificate, and circumnavigate the airworthiness certification. That didn’t work. The FAA put the brakes on it, because we only used the glass, the doors and the very center section of the fuselage. They said it’s a fresh airplane: you have to do an airworthiness certificate.
We put a new tail on it, starting right behind where the rear seats were. We designed, tested and built the outriggers with Blue Force Technologies. With RDD [Research, Design, and Development, LLC], we took the LX-7 wing molds and tail molds and adapted them for our purposes. The LX-7 is a traditional tail, while this is a T-tail, so all of the structure and all of the rigging internally had to be redesigned.
The wing is only about 250 lb [115 kg] because it doesn’t have to carry all that infrastructure for fuel management and sealing, so it’s really a bare-bones wing. There’s no deicing on the leading edge, like there would be in the electrically heated LX-7.
In the front where the engine mount was, we designed and built a frame out of chromoly that both holds the batteries and supports the front outrigger, which is tilted forward 25° (i.e. 65° from horizontal).
What we call the torque tubes at the ends of the outriggers are populated with motors, gearboxes and belt drives. It’s an epicyclic 3.7:1 reduction gearbox to go from a 4,000 RPM motor to just over 1,000 RPM on the 11-ft (3.4-m) diameter propellers.
Internal to the outriggers, you have two different, independent inverters: one on the outboard and one on the inboard. The inboard one powers the top rotor; outboard, the bottom rotor. Then, a cooling system in the center. Those are electrically and hydraulically completely isolated, so that if the plumbing system cooling the outboard inverter and the lower motor were to fail, you still have the upper motor and the inboard inverter cooled, appropriately managed, and controlled independently from its adjacent motor.
Conceivably, you could lose an entire battery, with two full 800 V batteries that don’t touch each other at all — not galvanically, not through controls power. If you lose the front battery — that controls all of the top props — or the back battery, which controls all of the bottom props, you still have a controllable aircraft.
From a systemic vantage point, you have systemic redundancy. You don’t just look at it and say you have eight motors, can you fail two motors? They may have some type of common mode failure. You separate those completely so that probability of failure on two motors that would put the plane on its back are completely independent from each other. That was the principle of the design.
Then on the articulation of the outriggers, we wrote a series of rules. Number one, they were never allowed to go limp, so that you can still maintain a controllable aircraft. Number two, they had to be synchronous. Number three, they had to move, so we can land safely at any vector angle.
We’re very in tune with the importance of each physical, mechanical or electrical system, as opposed to looking at everything the same. That’s how the articulation system is designed.
As far as the batteries, they are 18-650 cells in this version. They’re packaged inside a phase change composite [PCC] material from AllCell.
The way it works in this particular application is it manages the transient thermal spike of the hover portion of the flight and then allows for cooling. There’s forced convection through each of the batteries during the cruise portion of the flight. Then it also has a phase change material in it that allows it to absorb a huge amount of energy, should you reach the tipping point on any one cell.
The design of the battery system is dominated by the thermal management, both in normal operation and in the event that normal operating temperatures are exceeded.
Vertiflite: Can you explain more about the propeller motors and the gearing?
Clark: Yes. Permanent magnet motors — to get to torques that are measured in the 1000 N-m [700 lb-ft] range — have to be really large. There’s nothing off the shelf that can do that. These are EMRAX motors. They prefer to spin much faster to get the power, which is RPM times torque.
In order to get 70–100 kW out at peak power, depending on the initial condition of the motor, you need to run it at or near 3,000–4,000 RPM. In order to get the disk loading of the aircraft low enough, you need a large disk. In this particular case, the size limit that we’re comfortable with going in edgewise flight in a rigid configuration was 11-ft (3.4-m) props. This puts our RPM down at 900–1,000 for normal operation (up to about 1,200 for peak hour); this requires that we reduce the motor speed by about four.
With gearboxes, you can’t just pick the number four; there are particular ratios that work. This belt drive is a 4:1 reduction — these are for temporary hover testing; our epicyclic gearboxes are a 3.7:1 reduction and will go in for longer-range missions. The gearboxes are already designed; for example, the bearings have a 10,000-hour life. These belts here, we’ve replaced them already after just about 100 hours of time. So, these are just temporary, but they work really well and it’s really simple.
Vertiflite: Who made your propellers?
Clark: We made those. In fact, the first examples we made in my barn to start with. Then when we got a shop at the airport, we started making them there.
These are made out of stacked Vermont Maple with an intermediate fiberglass unidirectional layer between each of them. Then they’re machined, with a scheduled carbon fiber over the top with both unidirectional and weave. You can see the outer layer of weave here, but for UV protection we have to coat the top.
When you’re flying, you actually see them flex. When you’re in the cockpit, you’re eye level with the top [of the propeller] when they’re at 90°. As you modulate the thrust, you see the flex of the rotors moving around. You get about 900 lb [4 kN] on each [prop] picking the aircraft up, and you can actually visualize — like you do with a rotor disk in a helicopter — the disk pulling you around, but here you see the torques change.
Vertiflite: How about the wings?
Clark: For low speed flight, we put double-slotted flaps in the wing that allows us to get very high relative angle of attack on the wing, which helps with the transition profile. Specifically, when you start to get the downwash of the rotor on the wing.
This is a pretty awesome wing. It’s much longer than the original ES wing. This is longer, it has a better airfoil on it, and it has a much more aggressive flap, so it has [great] low-speed flight characteristics.
Vertiflite: The exterior — it’s the standard LX-7 wing?
Clark: It wasn’t the standard LX-7 wing before, but it is now.
It’s an LX-7 length, exactly, with the internal geometry and the leading edge changed. The LX-7 had double-slotted flaps, but we worked with RDD and dropped the flaps further, and added some geometry changes to allow it to get better low-speed characteristics.
They first flew it in July 2017 after we had already started working with them on other parts of the project. We knew the anticipated performance, but when they flew it on a fixed-wing aircraft we said, “That’s the wing we want. That’s the baseline we want.” So, then we started working with them to get this configuration nailed down.
RDD was really good. They’ve done high-performance gliders. They did a lot of conversions on the prop jets for Lancair. They built the Mako, which Lancair dubbed as their new aircraft.
They’re a group of maybe 25 people there that are absolutely obsessed with making high performance experimental GA aircraft. They designed and built the Airbus glider [Perlan 2] that set the altitude record [52,172 ft (15,902 m) in September 2017].
Clark: Let’s look in the cockpit.
This is the flight deck. It’s super sparse for a reason, but very capable.
What you’re looking at here is a fly-by-wire system. Put your hand on that stick and just move it around a little bit. You see how it moves over here? Go ahead and roll it side to side.
You see this display right here? It’s way different than anything you’d see in regular aircraft, because you’re monitoring things like your CAN [Controller Area Network] bus, the battery module, the voltage, your inverter and motors.
If we throw these [water] pumps on here, you see our flow comes alive on the different top and bottom. Remember that their flow is independent on top and bottom. These are the top four rotors, and these are the bottom four rotors; that should all come on now and indicate flow.
Vertiflite: Water flow for cooling?
Clark: Exactly. Then these are the torque gauges. As you’re moving and maneuvering, you see the torques on all the different motors.
Dave Churchill on our team designed these systems for MicroStrain, then they were acquired by LORD. We have the people on the team who actually designed and developed inertial measurement systems.
Touch that video camera on the left side. That’s how we fly right there. You actually get an underbelly view of the aircraft while you’re landing it. For example, if you want to land it on a spot, you see the visibility isn’t awesome out the window; it’s not like in a helicopter. For this prototype we put those cameras in to allow us to really nail the touchdown points.
Vertiflite: Where did the [single] seat come from?
Clark: It’s a racing seat from car racing, so it’s made for sideloading. It’s not made to be comfortable, I guess, it’s made to be safe. This is your collective right here.
Vertiflite: Collective is what — speed or torque?
Clark: Collective is average torque of all eight motors. Then you modulate the distribution of torque to each of the motors with a stick, some with the IMU [inertial measurement unit] that is running through the constructive mathematical model of the plane.
For example, we’ll use feed forward algorithms to quickly stabilize the plane in the case that it gets perturbed out of balance or the intention of the pilot is not matching the attitude of the aircraft. Then we’ll use a PID [proportional–integral–derivative] algorithm to stabilize it over a longer period of time. Obviously, that gets tuned based on the rotational moments of inertia of the plane and roll, pitch and yaw.
Then we have a full telemetry system beaming everything down to the ground over a couple of different methods. One is Wi-Fi, one is LTE and another one is over S-band radios.
We’re doing a lot of different things with making sure we get good data down and up in real time, because in the future when we put eVTOLs in the air, it’s going to require a huge amount of data flow between the ground and the air.
Vertiflite: So, why do you have this configuration?
Clark: Early on we were like, “What is the quickest path between a concept and a flight vehicle that can take the biggest risks in electric propulsion and bring them down?” Everybody knows we can make a wing that flies, an aircraft that’s stable, one that has control harmony, but the big risks were: can we make a battery system that is safe for flight; can we make a flight controller in a Part 23 airplane that is both redundant and low cost; can we communicate to the motors in a difficult electromagnetic environment?
Everybody loves to sit down and sketch up a beautiful aircraft, but that’s not the hard thing, that’s not what’s interesting here. What’s interesting here is utilizing and exploiting the benefits of electric propulsion while managing those hard spots.
Vertiflite: Tell me about your propeller design. You have different sets for different test regimes?
Clark: Depending on your ratio of time spent in hover flight and cruise flight, you have a different optimization point for the rotors. If you have a lot of testing going on where you’re just in hover and you want to do 15-minute hovering, you put on a propeller like this that allows you to do whatever test objective you have for a longer period of time. This one is just under 9° [twist] at the tip. Whereas if you’re going to be moving fast quickly, you want a higher advance ratio propeller, so this is a higher pitch propeller. This one is almost 11° at the tip. That allows you to move very quickly without putting the tip into a negative angle of attack. Because this rotor is in the downwash of the top rotor, this one always sees more airflow than this one, because it has its own ingestion or induced flow and then the downwash from this rotor.
As we get going over 100 kt [185 km/h] or to 120 kt [222 km/h], this one actually goes to the top, and there’s an even higher advanced ratio propeller that goes on the bottom. You can still hover the aircraft, but the optimization point is to spend more time in cruise.
This particular [blade] is set up for a particular test where we’re in between [hover and cruise]. We’re at like 70 kt [130 km/h] going forward, and that allows this propeller to still be flying and not go to negative angles of attack and it allows this thing to be at a more optimal point.
Vertiflite: That’s how far you’ve gone, about 70 kt [as of the end of March 2019]?
Clark: Our peak so far is 74 kt [134 km/h]. We’re just flying on the wing with the torques pretty low, to establish pitch margins between the front and the back. Probably two weeks from now, we’ll up that and we’ll hit about 100 kt [185 km/h]. This thing will articulate forward in the air and then come back just a little bit. Just nurse it forward and come back.
Vertiflite: What kind of angles have you flown so far?
Clark: Up to 65°. This is our max angle so far. That gets us 70+ kt [130 km/h]. Once we’re on the wing, the torque — it’s very nonlinear, because you go from here to here and you’re just ripping forward, and then you turn torque way down and you only go to about 100 or 105 kt [185 to 195 km/h] as you go horizontal for optimal range. You can obviously go 150 kt [278 km/h] if you wanted to, but for optimal range you turn that down, and then it’s you’re like swimming through the air with very low torques.
The highest risk is as you transition from the rotor onto the wing, and that happens between about 75° to 60°. At 60°, you’re going fast enough so then you grossly reduce the torque, so you’re not powering through this portion. You get it flying on the wing and then you kind of nurse it through. It’s fun!
For more on the Ava XC, see “Electric VTOL for Organs on Demand,” Vertiflite, Mar/Apr 2019 and www.eVTOL.news. High resolution photos are available on the VFS Photo Gallery: gallery.vtol.org/album/Pv6x.