VTOL is an abbreviation for Vertical Take-Off and Landing. VTOL describes fixed-wing aircraft that can lift off vertically. This classification includes only a very few aircraft like helicopters, autogyros, jump jets, and tiltrotors. Helium-filled balloons and airships are not normally considered VTOL. The following project was dedicated to our passion for making a functional VTOL design for the hobbyist using conventional components available at multiple vendors.
It seems that some people currently working on a VTOL project are getting hung up on making a functionally true to scale model (like a V22 Osprey) which makes the cost and complexity of the project extremely high. Other successful projects have been contracted by commercial or military organizations where cost is not a major focus.
I have also seen successful VTOL projects that end up looking like spiders from outer space. While I suspect that these “spiders” are good low-cost designs, they do not look or fly like a scale plane.
My focus in this month’s issue of AMP’D will be in making a VTOL design that is a repeatable project for the advanced R/Cer; one that is easy to repair and looks like a real model plane. By using some of the key concepts revealed in this multi-part article, you can create your own VTOL design inspired by your imagination.
Selecting a Host
I initially looked into making a V22 Osprey for my VTOL project. I discovered that this model did not exist in a .60-.90 size kit or ARF so I then planned to use an existing heli body to cover my framework. The model that seemed closest to the V22 Osprey body was the Century Airwolf fuselage that cost around $200?unpainted
I decided that my host airframe must be inexpensive, readily available in ARF form, and, easily repairable. After discovering the full scale Bell/Agusta BA609 Tiltrotor, I realized that my host design could take on the form of a more conventional plane instead of the military heli-like V22 Osprey.
I choose the Multiplex Magister ARF for my host plane due to its size, low cost, and, my previous experience from first reviewing it and then modifying it for brushless power. The Elapor foam can be assembled quickly and repaired with regular CA and kicker (accelerator). The nice thing about using the Magister as a host is that I already know it flies great and it is available from trusted vendors like Hobby Lobby , Horizon Hobby, and Tower Hobbies. The CG was not critical for normal flight and could deviate +- 10mm without issue. l will simply be using twin motors instead of a single one in the nose.
Key Component Selections
When selecting key components, my goal was to design the project so that it may be reproduced easily by others using commonly available parts and low-cost components available through multiple vendors. To date, most designs I have seen are making military-sized VTOL projects, spending thousands of dollars. These projects often used custom machined parts that were not readily available to others.
A computer transmitter with 8 or 9 channels is a must for this project as it makes mixing the spare channels easy to coordinate. I will be using a Futaba 9C transmitter with a dual-conversion receiver. At the very least, an 8 or 9 channel receiver and Futaba 9C transmitter will provide a good start.
When it comes to gyro performance, you basically get what you pay for. The better performing gyros cost more money. To help me select the right gyro for my VTOL project, without breaking the bank, I solicited help from heli experts, Ray and Kyle Stacy.
I also felt that using an inrunner motor would be easier to mount and swing 90 degrees without having to deal with a rotating can. My plan was to use two Jeti Phasor 30-3 inrunner brushless motors, Jeti 40-amp ESCs, and 12″ props on a single 3-cell (20C) 5000mAh LiPo pack. The single 14oz 3-cell 5000mAh pack can deliver 100amps, if needed, and keep the cost and weight of the model down. By using 12″ props, I could theoretically obtain about 8lbs of thrust on my target 5lb plane.
Design and Testing
For initial testing purposes, I saw two independent engineering goals. One is the ability for Vertical Take Off and Landing (VTOL). This testing will lift the plane up and set it down with the props pointed up. There will likely be a need for an additional stabilization tail motor to keep it level on a third axis. The second goal is to simply make a flyable plane when the props are forward. This should be a normal plane design with dual motors. My previous experience with the Multiplex Magister proved it to be an excellent stable flyer. The final goal would be the transition from VTOL to flying (and back) where the plane starts moving forward from hover to flight.
One problem to deal with will be roll control. The electronic speed controllers (ESCs) may not be fast enough to control roll using a gyro. I may need a basic rate gyro to dampen out roll during the hover mode.
Pitch control is another issue. If the aircraft uses gravity for pitch stabilization, it will be become unstable. Perhaps a small motor and prop in the tail pointing up or down might work if the aircraft is made tail heavy. The pivot on the main motors could be made to shift the center of gravity back so that the tail fan can be more effective. When the motors are shifted forward in normal flight, the CG can be moved forward again. These type of design issues were mulled over, again and again. I knew that it would likely take some testing before finding out if a particular technique worked or failed. I also wanted to keep things simple.
Wiring would be simple plug and play. It should allow for easy disconnect of the main motors and various control stages. The Futaba 9C transmitter and receiver will control everything using mixes, knobs, and switches. An inexpensive gyro on the tail fan could be used for pitch dampening. The two main motors could be mixed in the transmitter to provide manual roll control. There could also be gyros on each speed control in the roll direction, if needed.
I decided not to worry too much about regular flight mode since I knew the plane flew well so my initial concentration would be on hovering and stabilization. Did I even have enough thrust to hover? What if the plane weighed more than I had anticipated? These questions were often thought about during the initial design phase. I saw the transition from vertical hover to forward flight as the tricky part.
Here are some key points that I considered to reduce the cost and complexity of the project.
1) Does not need to be functionally scale
2) Use a computer transmitter for mixing functions, 8 (or more) channel receiver
3) Use gyros for hovering stabilization
4) Motors only swing on one servo-driven axis (perhaps only 90 degrees for up and forward
5) Use a tail fan somewhere on aft fuselage for pitch stabilization
6) Use inexpensive rate gyros for low level hovering stabilization
7) Break problem into basic steps
a) Vertical Takeoff and hover
1) Test vertical lift to verify thrust/weight level (correct motor/prop choice)
b) Normal flight with normal gear landing (fly it like a normal plane)
I consider steps 7c and 7d to be the most difficult and will address them in a future part of my article. One possible solution is to rely on advanced piloting skills. Proper hovering stabilization may also prove to be a challenge.
On my first test for VTO (or thrust level) the dual Jeti Phasor 30-3 motors will use the throttle stick for master ESC control and I’ll have a mix set up in the 9C transmitter to use aileron stick to slightly offset the speed of the ESC for simple roll control. This offset mix will help during the test as the gyro will not be set up yet. A simple string tied on the tail will let a human control the pitch for the initial thrust test. I don’t expect to take the plane more than 4′ off the ground on this test. It will qualify my choice of motors, props, and 3-cell 5000 pack as valid components on the Magister…or maybe it won’t qualify them.
According to Motocalc 8, I can expect 4lbs of thrust at 43 amps using an 11×5.5 APC prop on my Jeti 30-3 motor. That’s 8lbs total thrust and 86amps full throttle for both motors on my expected 5lb airframe. Since the generally accepted practice is a minimum thrust to weight ratio of 1.1:1, it was a good start.
The Multiplex Magister host plane and other key components had all been selected. I had also broken the complex VTOL plan into more basic, easier to achieve steps that could be tested independently. It was time to start building!
Building and Stabilization Planning
I decided to try the following techniques for my initial VTOL project stabilization and control. A single JR DS8611 digital servo (260oz/in on 6v) will control both outer wing thirds from not only swinging up and forward but providing additional pitch control in the hover mode. The servo easily mounted in the stock area meant for the standard aileron servo. I will be using smaller HS-55 servos on 6v for each aileron which provides about 1/2 the torque of a standard stock servo on 5v. Just about a perfect torque replacement for the stock requirement. By rotating a large portion of the wing up during hover mode, it keeps the motor thrust very efficient.
The Multiplex Magister design provides a carbon rod and channel meant to hold the two wing halves together. The plane was designed for easy removable wing halves. I will create a one piece wing as the span is only 64″…still very portable.
The stock channel cover can be cut and then sealed with the carbon rod to the outer 1/3 wing half but let the tube swing inside the channel on the center 1/3 portion. This technique allows a single DS8611 servo to control the wing pitch and will perhaps be stabilized by a single gyro.
Since the outer third of the wing on each side will swing up, the ailerons can be used for yaw control as the air from the main motors will flow right over them. A gyro in-line with the aileron channel will provide automatic yaw control. An additional transmitter mix from the rudder stick into the aileron channel (during hover mode only) will allow the pilot to also perform manual yaw compensation.
At ground level, I expect hovering to take place by pilot and gyro control. Higher up, I expect the pilot to control most of the movement. The unique side trim controls on the 9C transmitter may allow for easy mode or gain changes on the gyro. When the wing is swung down for normal flight, I will use leading and trailing edge stops to keep the wing properly positioned. I may also need a servo driven latch to help lock the wing position for normal flight.
My initial flight mode testing would use the following control surfaces.
Vertical Take-Off and Landing
Aileron – moves ailerons for yaw
Aileron – moves ailerons for roll
My Speed 600 motor mount used on the Hobbico Superstar EP can be unscrewed as well as shimmed to vary the thrust line, if needed. I tested the power system using a 12×6 APC e-prop to draw 44amps (440w) at full throttle on a 3s 3200mAh (20C) pack. This should work well with a single 3s 5000mAh (20C) pack on two motors. I expect to hover at around 30-35amps per motor.
The linkage on the JR DS8611 servo used stock Hangar 9 1-1/2″ Titanium Pro-link (HAN3550), 4-40 hardware (HAN3616), and, a 3D servo arm (HAN3578) from Horizon Hobby. Note the carbon rod for locking the leading edge into position. For the trailing edge, I used a small rectangular piece of lite plywood.
I misjudged my control horn position on the carbon bar before CAing it. It was too far forward making it difficult to achieve full vertical position for hover take-offs. The metal bracket “fix” is a Sullivan 5/32″ (#888) wheel pant bracket with one arm removed. It is an attempt to correct the swing range. I will keep these lessons in mind for a second wing build.
I was somewhat concerned about the strength of the 10mm carbon tube for supporting the weight and thrust of the outer wing thirds. I will look into using an aluminum tube for my second wing build. By using a ½” aluminum tube for my second wing build, I could then utilize a regular Hangar 9 (HAN3614) 8-32 Swivel Clevis Horn. The 8-32 screw will easily fit through the tube without weakening it as it would be bolted in place. I should still be able to make this initial version work and gain valuable experience from it.
Since I could not find an APC 12×6 pusher e-prop, the Zinger 12×6 normal and counter-rotating (pusher) wood props (ZINQ1260 and ZINQ4015) were used to cancel the torque of each motor. These props are available at Tower Hobbies. The brushless motor can easily be reversed by swapping any two of the three wires going to the ESC.
A 6v 10-amp Power Force Regulator (VRLI2) from FMA Direct was used to power the receivers, gyros, and servos. With all the electronics, I was concerned that the 3-amp UBEC would not be able to supply sufficient enough current. Further, the 10-amp Power Force Regulator uses noise-free linear regulation for a clean 6v output.
|I found a new use for my old chin-up bar in the garage as it will help me balance the VTOL plane and perform stabilization tests. Initially, I hung the plane from the two prop shafts and later from a single point on the wing chord at the Center of Gravity.|
The photos show that I initially used a LensRC 17t motor from Todd’s Models for pitch stabilization in hover mode. The thrust level was not sufficient and this motor has already been replaced with a higher power 150 watt brushless power system. Although the pitch control cannot be reversed, the pilot has control over the amplitude to make the tail settle where he wants. This seems to work well in conjunction with a second E-flite G90 gyro in-line with the motor’s ESC. The 150 watt motor (BP-12 at BP Hobbies) seemed to provide good pitch stabilization power.
I also experimented with some techniques for transferring the load from the outer wing thirds to the center portion. The results of my experiments convinced me that what I really needed was a stronger aluminum tube to replace the stock carbon tube.
Initial Thrust Test
After all the planning and building, it was time to perform the initial thrust test. My final weight was close to 6lbs instead of my desired 5lb estimate. If the motors couldn’t lift the plane off the ground, I would need to re-think my entire power system or find ways to reduce the unwanted weight.
My hope was that the low voltage drop using a single 3-cell, 20C 5AH pack combined with the efficient transfer of thrust through the tilted wing would prove to be key areas of the design enabling it to work.
The wind was blowing about 10mph that day and it was difficult to lift the plane off the ground without it starting to topple over.
I patiently waited for a calmer day. In the Northeast, the fall season can bring plenty of wind and rain so I had to wait several weeks before my opportunity happened. In the interim, I started to work on my stabilization techniques. When the calm day finally arrived, to my surprise, the experimental plane not only lifted off the ground, but I was only at 2/3 throttle!
“Project VTOL” was off to a great start!
Improving the Wing Design
Sometimes you just have to experiment with a technique to find out what works and what doesn’t work. I did discover several valuable things on my initial attempts. First, the 10mm carbon tube was not sufficiently strong to handle twisting forces when crashing the plane during the thrust and hover tests. Second, the various gyros I tested each had their own good points and bad points. Read on as I reveal my findings.
The 10mm carbon tube, although light, did not hold up to the occasional twisting action that would occur when the plane rolled over during the thrust and hover testing. This type of carbon tube is not designed for twisting forces but has great strength to withstand bending forces. Further, the added weight of twin power systems, wires, and electronics, put a greater load on the wing which it was not designed to handle.
Tower Hobbies sells individual parts for the Multiplex Magister so I built a new wing using a 1/2″ aluminum tube from Home Depot. The aluminum tube is 1/16″ thick and will be able to handle much stronger forces at a penalty of additional weight. The 36″ tube, which I cut to a length of 31″, weighed 3.5oz when finished.
To help strengthen the tube channel and distribute the forces where the wing sections separate, I added some aircraft grade plywood braces and glued them to the center section of the wing.
The 10mm wing channel was also sanded a bit bigger to make room for the thicker ½” (13mm) aluminum tube. This was accomplished by wrapping the 5″ cut off section of tube with #100 and #150 grit sandpaper.
Before mounting the aluminum tube inside the new wing channel, I added a 6″ long hardwood dowel epoxied in the center area where the Hangar 9, 8-32 Swivel Clevis Horn will be mounted. In this manner, drilling a hole through the tube in that section would not weaken it very much.
The entire linkage was now made from Hangar 9 giant scale hardware. It was simple and solid. The mechanical rotation of the new wing control provided about a 130 degree swing.
Gyros and Stabilization
After some research on available gyros, it was time to try one. Basically, what I had learned from several accomplished R/C heli pilots was that you get what you pay for. Of course, I did not want to go out and spend $300 for the top of the line designs so it becomes a compromise area of the design. I decided to order three of the E-flite 9.0 Gram Sub-Micro G90 Heading Lock Gyros for initial stabilization testing. Model Aviation columnist and heli expert, Ray Stacy, had tried one on a small electric and said it worked fine, but they often use heavier and more expensive JR G500T gyros on the larger T-rex brand helis.
The new E-flite G90 weighs 0.3oz and the JR G500T weighs 1.0oz. If it didn’t work the way I needed, I would then try some more expensive JR G500A or Futaba GY401 Rate Gyros.
For my VTOL project, the heading lock mode is not needed. The heading lock mode is actually an angle hold that is prefered operation in helicopter tail control. As I later discovered, the E-flite G90 gyro defaulted to the Heading Lock Mode and could not be changed to the Rate Mode without using a spare channel for control. This limitation would become an issue later on.
Another consideration to keep in mind is the control speed of the ESC. If the ESC cannot keep up with the gyro compensation, it appears to make the gyro look like it has poor performance. I kept this in mind when I was testing my pitch control system.
For pitch control, I decided not to use an EDF motor as it is not meant to run in reverse and the spool-up time may be too long for proper stabilization. The technique I mostly considered was to make the model tail-heavy when the tiltrotors are pointed up so that the aft tail fan can modulate around 50% thrust.
My initial pitch control system was “borrowed” from one of my Icarus Shockflyer models. It was a 50w brushless power system that was meant for a 5oz to 10oz model. Initially, it appeared that the gyro compensation into the ESC for pitch control was rather slow but I eventually determined that the thrust level was not sufficient. I replaced the 50w power system with a 150w brushless power system (BP-12 at BP Hobbies) spinning an 8×4 SF prop. The increase in power provided an increase in pitch response as well as keeping the motor and ESC cooler. The added weight was not an issue.
Up until now, much of my VTOL plane control has been manual using the transmitter sticks through various mixes. The second E-flite G90 Gyro was added to the aileron servo for yaw control. When the tilt rotors were pointed up, the ailerons controlled yaw through the thrust passing over the upright wing sections. Since the gyro was mounted on the outer moving section of the wing, when the tilt rotors faced forward for normal flight (a 90 degree change), the compensation was now on the roll axis. The yaw control seemed to work well when twisting the plane back and forth manually.
The aileron control gyro gave me another idea. I decided to put the third gyro on the JR DS8611A digital servo that controls the wing swing. The E-flite G90 gyro has a digital servo mode for super fast control response. My initial testing showed that this digital servo control mode using the G90 gyro had excellent response time as I could tip the plane forward and backward while keeping the props pointed straight up! One short coming that I ran into was that unless I connected the G90 gyro to a spare channel on the receiver, I could not select the Rate Mode. Further, there was no offset control (or Servo Travel Limit Adjustment) on the gyro so I could not obtain the correct gain I needed when the wing was properly positioned. After wrestling with the limitations, I decided to upgrade my wing stabilization control to a JR G500A Airplane Rate Gyro. The JR G500A (now replaced by the G770 3D) is a more capable (and expensive) ring sensor gyro that has outstanding holding power and drift-free operation. It is designed specifically for airplane use and offers a variety of settings for easy setup. The G500A has a high rate frame selection for digital servos and separate gain and servo travel adjustments. The JR G500A gyro allowed me to set up the VTOL wing swing gain and position as I wanted it. However, the G500A gyro did require using a spare channel for gain control.
Futaba 9C Transmitter Mixes
Mix 1 – Throttle -> Flap (Ch. 3 to Ch. 6 for controlling both main motors)
Mix 2 – Aileron -> Flap (Ch. 1 to Ch. 6 for manual roll control of one motor in up position)
Mix 3 – Elevator -> Gear (Ch. 2 to Ch. 5 for manual pitch control of wing swing in up position)
Mix 4 – Rudder -> Aileron (Ch. 4 to Ch. 1 for rudder stick control of aileron yaw in up position)
Mix 5 – Offset -> Aux2 (Fixed offset to Ch. 8 for turning on/off pitch control motor in up position)
Mix 6 – Throttle -> Aux2 (Fixed curve +100% offset into Ch. 8 for disabling pitch motor when throttle stick is < 10% or all the way down)
Mix 7 – Unused
Receiver Channel Assignments
Channel 1 – Ailerons
Channel 2 – Elevator
Channel 3 – Throttle for main motor 1
Channel 4 – Rudder
Channel 5 – Hover/Flight Mode Switch
Channel 6 – Throttle for main motor 2
Channel 7 – JR G500A Gain adjustment for wing swing servo
Channel 8 – Throttle for 150w pitch control motor
The initial hover testing was done over a soft grassy area in my backyard and at a local park. The transmitter mixes allowed me to manually compensate for all three axis. The ESC gain on the pitch control motor was controlled by a knob on the transmitter which adjusted how the tail sat in relation to the upright wings. After some repeated testing, I discovered that the E-flite G90 gyro would “creep” from its established setting making the tail go higher or lower. Although the amount of “creep” can be minimized by adjusting the sub-trim, it made the pitch control unreliable. When used on a heli rotor control, the effect is minimal as it will slowly make the tail drift in flight. On my VTOL design, the undesired effect would sometimes make the tail flip the plane over its nose onto its back.
My first thought was to replace the E-flite G90 gyro with another JR G500A Airplane Rate Gyro. The JR G500A ring gyro’s drift-free operation could eliminate the servo “creep” experienced with the E-flite G90 gyro. I then realized that I had no free channels to control the gain. I needed another solution so I decided to try a Futaba GY401 Rate Gyro from Tower Hobbies for $140. This gyro uses a Silicon Micro Machine (SMM) sensor that eliminates trim changes in flight as it automatically corrects for a constant offset like cross-wind or other unwanted axis changes.
The Futaba GY401 Rate Gyro eliminated the “creep” in trim that I was experiencing with the E-flite G90 gyro and did not require a spare channel to set the gain or switch to a rate mode. As mentioned earlier, this gyro benefit came at about double the cost but it was a better fit for my project.
The Test of Transition
Once that I was fairly happy with the gyro compensation on the various axis, and, I had some success in low level hovering, it was time to move on to step “b” of my basic VTOL functions. Would my design still fly like a normal plane from take-off to landing even though it was a pound (or 20%) over the stock flying weight?
The main reason for testing normal flight with a tricycle gear take-off and landing is so that I had confidence to fly it like a normal plane if I got into trouble on my transitions in and out of hover mode. As it turned out, my overweight Magister flew just fine! Although I did not perform any aerobatics, it had plenty of power and easily flew at only half throttle. The 5AH (20C) LiPo pack lasted for about 7-8 minutes. The plane also flew tail heavy so I would need to compensate by moving the pack forward from the cockpit area and see how it affected my hovering capability.
Overall, I was pleased with the results of my design and testing of the Magister host plane. The power system had plenty of thrust to hover and the heavier than stock weight did not prevent the Magister from performing like a normal plane.
In the final part, it will be time for the true test of success. Will my VTOL design properly transition from hover to flight and back? Will I need any additional gyros? Assuming my design survives all the testing, what will my final scale appearance emulate? Stay tuned for part 2 of Project VTOL in a future issue of AMP’D.
When you fly electric, fly clean, fly quiet, and fly safe!
Special thanks for contributions by:
“Papa Jeff” Ring and Lynn Bowerman
This section of AMP’D covers some of the questions that our readers have sent in and I thought would be interesting for others.
Jim asks: “Hi Greg,
I lost my manual for the World Models 40S Ultimate and want to know what the recommended controls throws and CG location is?
Greg: Hi Jim,
Andy R. asks: “Hi Greg,
I just came across your Feb 2008 Amp?d article regarding EDFs. I see you were hopping up a twister with the Ammo 28-45-3600. In one of the pictures it shows a 5/32? tube being used as an adapter between the 3.2mm shaft and 4mm shaft adapter. Did this work well and give a nice snug fit? Also, is this setup performing well on 4S? Do you know the amp draw, thrust, etc by any chance?
|Greg: Hi Andy,
So far, I have had no issues with my Twister EDF hop-up. Just rough up the motor shaft and inside of the brass tube before using medium CA to hold it in place. Allow a 24hr dry time before test flying.
With my 4-cell FlightPower 3700mAh pack connected, the current draw was 55amps (about 700w) at full throttle static testing. I never measured the thrust but the performance change is very noticeable over the stock setup.
Note that with the extra battery weight, you need to add.
3/4oz of lead weight to the tail. See the attached photo where I added it below the elevator and painted it silver. I also made an exhaust ring from the bottom of a 20oz plastic bottle of Gatorade.