Developing full scale unmanned winged tiltrotor vtol

[William_Walker]

Hi, I am developing a full scale unmanned winged tiltrotor vtol similar in function to the Convergence RC model. The vehicle has 4 large wheels and is 15ft long, 8 ft wide and 3.5 feet high. It has two tilt rotors in front and one motor behind the cockpit for verticle flight mode. It has double elevons and double rudders for airplane mode control. It looks like Can Am race car. I am using four Q100-5L Hacker motors 10-11kW 137KV, 32 in x 10.7 in Mejzlik props, and 4 Fatboy 300 A ESCs, and 2-4 Tattu 22Ah 14S batteries. The vehicle weighs 150 lbs and is built out of 1 in foam sheeting with 2 oz fiberglass cloth. I am using some plywood, and carbon fiber for reinforcement. I am using ArduPlane software, Pixhawk Orange Cube flight controllet, Tiranis X9D Plus 2019 transmitter, and frsky rx8r receiver.

The 2 front motors are counter rotating, and the 2 motors in back are coaxial counter rotating and placed back to back. They blow their thrust through a hole in the wing, just behind the cockpit.

I have built and tested several 1/6 scale foamboard versions of the vehicle, and they fly very well. I am from San Diego California, but I am building the full scale model in Sweden.

The full scale model is almost finished and we have started to test it. The vehicle can support it weight and we have taxied it on the street. I experimentally determines the PID values by carefully balancing the vehicle on 2 saw horses at the vehicles CG and firing up the motors so that I can study the vehicle as it tries to use the flight controller to stabilize it.

I used a pull test to determined the thrust generated by the front motors. They generate 64 lbs thrust each, where as 80 lbs was expected.

I tried to hover the vehicle, but it would not take off. The front motors can easily lift the front end of the vehicle. But the back motors will not lift the back end. The coaxial prop motors in back should be able to generate enough thrust to lift the back end. When we lifted the back end with 2 people.while at full throttle, the back end suddenly lifted with great force. It is looking like the pressure is building up in the back because the opening in back may not be big enough for the air to get out fast enough. The opening in back behind the coaxial prop motors is about 6 ft wide and about 3 ft tall. If this is the problem then we can cut the bottom of the side panels housing the wheels, so that the air vents from the back motors better. Currently the side panels are about 6 in from the ground. I have never used coaxial motors, so I am not sure what to expect. The 1/6 scale model uses only one motor in back with a 3 bladed prop, where as the front motors use 2 bladed props. To cancel the torque, the back motor is tilted a little to the side.

When we manually lift the vehicle about a 1.5ft, the vehicle the back end lifts. Cutting 1.5 feet off the bottom of the side panel is doable, but less would be preferred, since we want to use the vehicle on the water, using the side panels as floats. If we cut too much of the side panels, then the front props will touch the water.

Another possibility I am checking into is that we could tilt the motor slightly to help blow the air toward the back, or we could use non-counter rotating motors in the back and tilt the motors a little to the side so that the thrust is blown under one of the side panels.

I must say, I am very surprised about this problem. The 1/6 scale model lifts in VTOL mode very well, but it is not using coaxial motors, and the motor is located above the hole in the wing, and the motor is tilted a little to one side to counter the torque.

Does anyone else have any ideas or comments?

Below are pictures of full scale vehicle. Note cockpit not mounted yet.
20210723_191605|666x500

Link to see 1/6 scale vtol test flight:
https://drive.google.com/file/d/1fSrr22Kh5rR_39nCc20a9UxKJXfvGmig/view?usp=drivesdk

Link to see demo of fill scale vehicle using RealFlight 7.5:
https://drive.google.com/file/d/1AAvC3BLvut7xjlhY81S9EqD8OW0Qw-kS/view?usp=drivesdk

[tridge]

great project!
do you need any help from us at the stage? Would you like to submit the RF model to our models repo? GitHub - ArduPilot/SITL_Models: Models of aircraft for SITL

[William_Walker]

Hi Tridge, Thanks! I have been working on this design for 6 years. We have almost got they vehicle to hover, but we are having trouble getting the back end to lift. The vehicle weighs 150 lbs + 30 lbs for 2 batteries, totaling 180 lbs. The prop motors should generate 80 lb thrust each, but a thrust test shows we are only getting 64 lbs. So for 4 motors we should be getting 256 lbs thrust minus 10% efficiency loss for the coaxial setup, which is more than enough to lift the vehicle. We can see that the front motors easily lift the front end. We can hear the back motors rev up well until about 60% throttle. But as you increase throttle, the sound goes down. We think there is a breathing problem due pressure build up under the back motors, even though there is a 6 ft x 2 ft hole in the back, and 6 in of space under the side panels for the air to escape when it is on the ground. When we lift the back manually, the back end lifts with great force and the back motor pitch increases. We tried placing the vehicle’s wheels on 4 crates 24 cm high, and the back end lifts with the vehicle initially in a horizontal attitude. So one solution would be to cut 24 cm off the bottom of the side panels, but this will probably prevent the vehicle from being amphibious, since the front props will be very close to the water. Also this will be an even bigger problem if we build a manned vehicle which will require bigger and heavier motors and batteries.

So I did some experimenting with a small coaxial motor setup on a stick and observed the change in weight as I lowered the motors into a box with a hole in the top for the motors, and no bottom, and placed the setup on a table. The results show that if the box is only slightly lifted, the motors weight decreases a lot as it approaches the hole and increases as it enters the hole. If the box is lifted a lot, then there is no weight change as the motors are lowered into the box. This experiment tells me that the motors do not work well inside the hole in our setup due to pleasure buildup under the motors. The best setup would be to place the prop motors above the hole, where they can breath better and generate full thrust. Also if one places them just above the hole, one can get extra lift on the ground from ground effect, due to the pressure pushing the prop motors up. We are now looking into how to mount our coaxial setup above the hole. We could just make a platform above the top motor and place the bottom motor there, but this would result in a tall coaxial motor setup, and we would lose some efficiency from the top motor since not all the thrust would go through the hole. A better solution would be to modify our motors so that the shaft of the bottom motors goes through the top motor via bearings in the top motor, and connects to the top most prop. The bottom prop can then be attached to the top most motor. The motors could then remain where they are and the 2 coaxial props would simply be on top. Another option would be to buy a more powerful prop motor. But this would require doubling the voltage and then would require a 4 battery setup, which is much heavier (one 14s battery for each front motor and two14s batteries in series for the back motors).

Do any any of you have any ideas or comments? Does my analysis seem ok? Do you agree that placing the back prop motors above the hole is the best solution?

Thanks, William

[tridge]

Interesting problem! I am not the right person to suggest solutions though - I’m a software engineer and really just appreciate the magic that vehicle designers do to make things work in the face of tricky aerodynamics.
@iampete might be a better person to comment?

[iampete (Peter Hall)]

I see you have ailerons on the back, I wonder if just moving them to the fully up position would be enough. It could be airflow over the back of the vehicle causing a venturi effect sucking the air out from underneath.

A smoke grenade on a stick might be quite informative. Weight the vehicle down so it won’t move suddenly.

[William_Walker]

Hi Peter. Yes, I have tried moving the elevons up. But this did not help much. In the 1/6 scale model, I tried mounting the motor under the wing once and it developed such a high vacuum underneath that I could not pick up the vehicle. Clearly the motor was drawing high speed air from the bottom of the wing and this created a vacuum. You can easily demo this effect by bringing a prop motor neat a wall and the resultant vacuum will suck the motor to the wall.

But when I mounted the motor on top of the hole it worked great. I dont think this is the problem with my current setup. I can hear the the motor pitch go down as I try to increase thrust past 60%. This tells me the motors are not breathing properly. Also the thrust and motor pitch increases as I physically lift the back end. I addition, when place crates 24 cm high under the wheels of the vehicle the back end lifts.

[William_Walker]

I also did another experiment, where I mounted a coaxial motor system to a rod, and lowered the motors into a small box with no bottom, and about the size of the prop.diameter. The setup was placed on a table. When I lifted the box a little, as I lowered the prop motors into the box, I noticed that the setup got very light as I approaches the hole. But as I entered the hole, the setup got heavier. When the box was lifted a lot higher, no weight change was observed as I approached and entered the box. This experiment tells me that if the box cannot vent the thrust coming into the hole, it generates back pressure in the hole which opposes the input thrust. As the motors enter the box the bottom motor develops a breathing problem and which also affects the top motor via coupling. I think the solution is to mount both motor above the hole so they breath properly and generate thier full force into the hole. The resultant back pressure wil also help lift the vehicle.

[William_Walker]

Because of the size of this model, I have been able to get under the vehicle and feel the resultant wind with my hands. I felt no wind under the wing, but felt a strong wind on the floor. Under the props I felt a strong wind directly under the props, especially towards the outer perimeter of the props, but much less wind towards the center of the props. There is of course a lot of wind coming out from under the vehicle and out the hole in the back.

[William_Walker]

Regarding your idea that the coaxial setup is causing high speed air to flow over the top of the wing, I have not felt this, and if it did, this would create a vacuum on top of the wing which lift it, not pull it down as we are seeing. So this cant be the problem.

[William_Walker]

From the basic physics of a prop, it can be seen that If the pressure at the output of a prop motor increases with respect to the input, the thrust will reduce. So if the output side of a prop is too enclosed, this will create an increase in pressure that will reduce the resultant thrust from the motor. Mounting the prop outside the hole enables the motor to develop full thrust before it enters the hole, because the pressure is the same on the input and output of the prop.

[William_Walker]

Picture of coaxial test prone and box setup:

Physics calculation showing increased pressure at output of prop results in reduced thrust:

Reference:
https://wright.nasa.gov/airplane/propth.html

[Swebre (Steve Webre)]

Strongly suggest you make a smoke wand or at least a ribbon on a stick and get a detailed mental image of airflow under the vehicle before make more changes. Might be as simple as several diffuser vanes to break up the tight vertical air column created by that rear fan. A nice transition duct on that rear fan will increase efficiency. Good luck!