New Products Flash Sale
Thread Tools
Old Jun 24, 2014, 11:50 AM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Build Log
Doak VZ-4 (Model 16)

The Doak VZ-4 was a Vertical Takeoff and Landing (VTOL) airplane built as a prototype for the US Army. The plane was designated as the Doak Model 16 by the Doak Company. The VZ-4DA was eventually donated to the U.S. Army Transportation Museum at Fort Eustis, where it is on display with other aircraft in the museum's outdoor Aviation Pavilion.

Our model was built using measurement taken from the plane. After many crashes and rebuilds, the model was made to fly reliably in forward flight with conventional takeoffs and landings. Vertical takeoffs and hover were not possible with manual control. We are embarking on an effort to incorporate the KK2.1 board for control in VTOL and hovering.
jb753 is offline Find More Posts by jb753
Last edited by jb753; Jun 24, 2014 at 12:17 PM.
Reply With Quote
Sign up now
to remove ads between posts
Old Jun 24, 2014, 11:52 AM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Full Scale Doak

Only a single prototype was built and it was assigned tail number 56-9642. The aircraft was originally powered by an 840 shp (630 kW) Lycoming YT53 turboprop engine mounted in the fuselage, later replaced by a 1,000 shp (750 kW) Lycoming T53-L-1 turbine. The engine drove two wingtip-mounted 5-foot- (1.5-meter)-diameter fiberglass tilting ducted fan propellers through a "T" box on the engine that transmitted power to the propellers via a 4-inch (102-mm) aluminum tubular shaft and two smaller shafts. The fans were positioned vertically for takeoff and landing with a rotation speed of 4,800 rpm required for liftoff and rotated to a horizontal orientation for horizontal flight, the first time this VTOL propulsion concept was tested successfully. It accommodated a two-person crew, with a pilot and observer seated in tandem in the cockpit. The pilot used a standard stick and rudder to control the aircraft. Its landing gear was taken from a Cessna 182, its seats from a North American F-51 Mustang, and its duct actuators from a Lockheed T-33 Shooting Star. Yaw and Pitch control in hover was accomplished with vanes in the jet exhaust stream. Roll control in hover was accomplished using variable pitch vanes in the duct inlet. They were controlled with a mechanical linkage from the control stick. As the nacelles rotated the amount of control of the inlet vanes was reduced with none in forward flight.

Flight testing began at Torrance Municipal Airport, and Doak completed several tests by 1958. The Doak Model 16 hovered for the first time on 25 February 1958, and the first transition from vertical to horizontal flight (and back again) took place on 5 May 1959. Although the prototype generally was successful, its short takeoff and landing performance less than hoped for and it displayed a tendency to nose up while making the transition from vertical to horizontal flight.
Doak's engineers believed that they could solve the prototype's problems, and after taxiing testing, 32 hours of flight testing in a test stand, and 18 hours of tethered hovering, the aircraft was transferred to Edwards Air Force Base, California, in October 1958. It underwent another 50 hours of testing, in which it proved capable with the turbine engine of achieving a maximum speed of 230 mph (370 km.hr), a cruise speed of 175 mph (282 km/hr), a range of 250 miles (403 km), an endurance of one hours, and a service ceiling of 12,000 feet (3,658 meters).

Doak VZ-4 (1 min 52 sec)
jb753 is offline Find More Posts by jb753
Reply With Quote
Old Jun 24, 2014, 11:59 AM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Full Scale Doak Pictures

The pictures are of the full scale Doak during testing.

Variable pitch vanes in the inlet duct were used to control roll in hover.

Vanes in the jet exhaust were used to control pitch and yaw in hover.

The plane was originally all green and more orange was added during the testing program.
jb753 is offline Find More Posts by jb753
Reply With Quote
Old Jun 24, 2014, 12:07 PM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Our model

Our model was built by Tex Harrison using measurements that he made of the plane at the Army Transportation Museum in Fort Eustis.

We have made 57 attempts to fly. We had 14 crashes on take off, 15 good take offs and flights but crashed on landing and 27 successful flights. Most of the successful flights were with the ducts locked forward.

Doak 5 22 11 Flight 3 (4 min 5 sec)


There have been many lesson learned during our flight program. One lesson is what happens when you have a bad battery with a big loss of capacity. Even though the motors and ESCs were identical, one shut off early.
Doak 5 22 11 Flight 4 Bad Battery (1 min 9 sec)
jb753 is offline Find More Posts by jb753
Reply With Quote
Old Jun 24, 2014, 12:11 PM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Vertical Flight

Our attempt at vertical flight (tethered) showed that the plane was uncontrollable. Therefore we are going to add the KK2.1 board using the OpenAero2 VTOL software.

Thanks HappySundays and Ran D St. Clair for your work on the software and documentation.
jb753 is offline Find More Posts by jb753
Reply With Quote
Old Jun 24, 2014, 12:22 PM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Reserved

Reserved for future information
jb753 is offline Find More Posts by jb753
Reply With Quote
Old Jun 25, 2014, 10:49 AM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Assumptions and channel assignments

Assumptions
1. A 3 position switch will be used to set the flight modes
A) Hover
B) Slow Forward Flight (SFF) (20%?)
C) Fast Forward Flight (FFF)
2. Auto level will be used in Hover Mode
3. Stability will be used in FFF Mode (probably pitch only)
4. Aileron stick will control ailerons in FFF and Lift Motors in Hover Mode. Some use of nacelle tilt may be used in FFF.
5. Rudder stick will control the rudder in both Hover and FFF but may be at different percentages. Rudder stick will control nacelle tilt (opposite directions) in Hover. Nacelle tilt will only be used if the rudder vane in the ducted fan airstream is not effective enough.
6. Elevator stick will control the Elevator in both Hover and FFF but may be at different percentages.
Elevator vane will be the main pitch control in Hover.
Elevator stick will control nacelle tilt (same direction) in Hover. (Need to think about this)
7. Tail motor (ducted fan) will be on 100% in Hover and off in FFF.
8. Nacelles will be full up in hover and full forward (possibly +2 or 3 degrees) in FFF.
9. CG will be on the nacelle rotation axis in hover and slightly forward in FFF. More forward CG may be helpful.
10. Pilot will control Yaw in Hover (Possibly some stabilization)
jb753 is offline Find More Posts by jb753
Reply With Quote
Old Jul 01, 2014, 03:01 PM
VTOLicious is offline
Find More Posts by VTOLicious
Registered User
VTOLicious's Avatar
Austria
Joined Oct 2007
281 Posts
I always loved the Doak! Keen to see it hovering with the KK-board.

All the best, Michael
VTOLicious is offline Find More Posts by VTOLicious
Reply With Quote
Old Jul 14, 2014, 08:06 PM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Baby Steps #1

Since we do not have any experience with the KK2 board, we decided to start simple and make a flat plate form that holds servos that represent all of the servos and motors that will be used on the flight model. We can give commands as well as tilting/moving the form and observe the outputs. We can test many different programming values and see what change they make.
jb753 is offline Find More Posts by jb753
Last edited by jb753; Jul 15, 2014 at 05:11 PM.
Reply With Quote
Old Jul 14, 2014, 08:40 PM
Ran D. St. Clair is offline
Find More Posts by Ran D. St. Clair
Registered User
Ran D. St. Clair's Avatar
United States, CA, Sunnyvale
Joined Dec 2006
1,701 Posts
Quote:
Originally Posted by jb753 View Post
Since we do not have any experience with the KK2 board, we decided to start simple and make a flat plate form at holds servos that represent all of the servos and motors that will be used on the flight model. We can give commands a as well as tilting/moving the form and observe the outputs. We can test many different programming values and see what change they make.
Good plan. It should serve you well.
Ran D. St. Clair is offline Find More Posts by Ran D. St. Clair
Reply With Quote
Old Jul 19, 2014, 08:27 PM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Great Progress!

If you are not well versed on the KK2 VTOL, I highly recommend building the flat plate airplane and doing the experimentation with that. We were able to get everything set up today and confirm that all of the inputs and feedback were working correctly. When I would originally talk to Tex about setting up the board he would just get a blank stare. By the end of the day, he was making all of the changes.

We are now working on Baby Steps #2 which is a much simplified Bi-copter that will only be used to set up the hovering control. It will have two lift motors without nacelles and they will be fixed upward. It will also have the EDF motor blowing air over the vanes in the tail. I should have a picture of that next week.
jb753 is offline Find More Posts by jb753
Reply With Quote
Old Jul 30, 2014, 07:42 PM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Baby Steps #2

We were able to learn much using Baby Steps #1. We tuned all of the values and were able to get all of the responses that we desired. For many of the values we had to put in big values to be able to see all of the responses and know that they will need to be dialed back for flight.

Baby Steps #2 was built to be able to determine and tune the values for hover only since FFF already works without a KK2. It will also be much easier to repair as we go through "learning opportunities", otherwise known as crashes. In the attached picture the KK2 has not been hooked up yet. BS #2 only has the components needed to determine the hover settings - the two lift motors, the EDF and the vanes for pitch and yaw control.

We welcome comments from experience OpenAero VTOL users on the values in the attached file.
jb753 is offline Find More Posts by jb753
Reply With Quote
Old Jul 30, 2014, 09:17 PM
Ran D. St. Clair is offline
Find More Posts by Ran D. St. Clair
Registered User
Ran D. St. Clair's Avatar
United States, CA, Sunnyvale
Joined Dec 2006
1,701 Posts
BS2 Strikes me as a really good plan.

The first thing I notice is that the two lift motors don’t look like they tilt. That is fine for hovering but the thrust from the EDF will push it forward and you will have to hover 5 to 10 degrees nose high. The AL (AutoLevel) will then be fighting to push the nose down and it will want to scoot forward on takeoff, so it will be awkward. I recommend you shim the main motors to tilt them back 5 to 10 degrees and adjust the angle until the body hovers about level. It should make your life easier in general.

The next thing I note is that both of your lift props turn CCW when viewed from the top. You need to replace one of them with a CW prop. I recommend the right one be CW, that way when you give left roll you will get a slight amount of left yaw in the bargain.

I assume that the CG is directly in line with the 2 lift motors…

As for your parameters…

The Profile 1 values look a little high in Roll, I would guess that you will need to increase all of your Profile 1 Pitch values. Without the aerodamping of a wing and tail these values are going to be quite different between BS2 and the final model. In general the final model will be able to use higher values without any tendency to oscillate. The Yaw P value in P1 looks fine. The yaw stability will be weak, but you don’t want to crank it up too much for fear of compromising your pitch control authority. The final aircraft will have a strong tendency to weathervane and you don’t want to fight that. BS2 doesn’t have a vertical stabilizer to weathervane so it will be different in that respect as well. The Acc Vert P value of 10 in P1 is fine, because it is optional, but you can crank it up to 50 or more without problems.

You might consider gluing on some depron scraps to add wings and a tail just for aerodamping. It also looks like the “wing span” on BS2 is shorter than the real thing. There isn’t much room for wings, and you don’t want to block the prop blast. The tail looks shorter as well. All of this will make the tuning values somewhat different when you get there, but it doesn’t mean you can’t have some fun and learn some things.

Your pitch P of 5 in P2 (FFF) is next to nothing, which is probably fine. Unless you are dealing with a tail heavy situation all zeroes should be fine as well. I assume BS2 is never intended to be flown in anything other than hover mode.

Starting with OUT1, your tail EDF. As soon as you crack the throttle beyond 2.5% it is going to jump to 100% throttle in hover mode. You will have no throttle control other than the transition switch. If that is what you want then great. I would probably use the P1 and P2 throttle volume in combination with the various offset values to make it come on a little more gently, but it is your choice.

Consider that if everything is in balance at half throttle in a hover, and you increase the lift throttle to climb it will want to go backwards. Likewise it will want to go forwards at lower throttle settings. You might want to think about how you can better balance those forces.

OUT2 is an aileron servo, but I don’t see any such servo in BS2, so I will skip it for now.

OUT3 is the Elevator servo which I assume controls the horizontal surface behind the EDF. Your settings look fine.

OUT4 is Rudder, which I assume controls the vertical surface behind the EDF. Your settings look fine there as well.

OUT5-6 are the Right and Left lift motors respectively. +15 and -15 Aileron volume input sounds about right. Turning on the Roll gyro and AL sounds good.

OUT7-8 are the Right and Left tilt servos which appear not to apply to BS2. Using the offset values as you have done should work fine when the time comes. You can also use differential tilt for improved yaw control in P1 if you want.

All values are subject to tuning but it all looks pretty straight forward. It looks like an interesting take on a tri-copter.

Be sure to set the MPU6060 LPF to 21Hz and balance your props.


Good luck!
Ran D. St. Clair is offline Find More Posts by Ran D. St. Clair
Last edited by Ran D. St. Clair; Jul 31, 2014 at 10:28 AM.
Reply With Quote
Old Jul 30, 2014, 09:57 PM
jb753 is offline
Find More Posts by jb753
Registered User
Williamsburg, Va, USA
Joined Jul 2001
343 Posts
Ran D
Thanks for the input. SB#2 is only about 1/2 the size of the final plane. We wanted to start with it simple and plan to add a fin and possibly other surfaces later to see how much it weather vanes. We also anticipate that the much lower mass and lack of aerodamping will make the tuning more critical than in the final model.

We anticipate that the EDF will push the plane forward and are planning on shimming the motors as part of the tuning but wanted to see how big of a factor that is.

We have the CW prop on order but again we want to try it both ways to see how much difference it makes, especially to see if the vane is powerful enough to stop the rotation.

The P2 pitch input was added for landing. The plane flies fine in FFF but is slightly pitch sensitive on landing at the hover CG. It works well with a more forward CG but we were testing flight with the CG in line with the motors.

I did not understand your comment on OUT1, the EDF. We are having it being controlled by the profile switch to make it simpler and not require a separate throttle on a slider switch. The throttle will is not mixed in so it will only be be controlled when the profile switch is moved into hover or SFF. At startup, the plane will be on the ground and heavy enough that I do not think that the sudden EDF power will have a big effect since it is relatively small. I did not include the entire worksheet but have a value of 5 for the transition. I was hoping that would work for the inbound transition. Is there another way to make it come on gradually?
jb753 is offline Find More Posts by jb753
Reply With Quote
Old Jul 31, 2014, 12:15 AM
Ran D. St. Clair is offline
Find More Posts by Ran D. St. Clair
Registered User
Ran D. St. Clair's Avatar
United States, CA, Sunnyvale
Joined Dec 2006
1,701 Posts
Quote:
Originally Posted by jb753 View Post
I did not understand your comment on OUT1, the EDF. We are having it being controlled by the profile switch to make it simpler and not require a separate throttle on a slider switch. The throttle will is not mixed in so it will only be be controlled when the profile switch is moved into hover or SFF. At startup, the plane will be on the ground and heavy enough that I do not think that the sudden EDF power will have a big effect since it is relatively small. I did not include the entire worksheet but have a value of 5 for the transition. I was hoping that would work for the inbound transition. Is there another way to make it come on gradually?
When the throttle stick is below 2.5%, all motor outputs are driven to a 1ms pulse. This prevents all motors from running and is an important safety feature. This is true regardless of whether you are in P1, P2, or anywhere in between. If you are in P1, hover mode, and you crack the throttle, then the EDF will immediately go to full throttle. This may not be enough to move it on the ground at such a low throttle setting on the lift motors, but that is only part of the issue.

If we assume that the aircraft hovers at 50% throttle, then the lift motors will need to be angled back to produce a rearward thrust that balances the EDF forward thrust. If you increase the throttle to climb, or even to arrest a descent, then this rearward thrust will increase, overcome the EDF thrust, and push the aircraft backwards. The opposite will happen if you decrease the throttle.

If the EDF is throttled in proportion to the lift motors then everything will balance at all throttle settings. The aircraft would only go up or down with throttle changes, which is what you want. This does not require a 2nd slider for the EDF throttle. It can be driven from the same throttle stick as the lift motors. As a matter of fact, that is exactly what you want, so they will always work together.

This implies that when hovering both the lift motors and EDF would be running at 50% throttle. That implies that the tilt back angle of the lift rotors would be less and the control authority in pitch and yaw would be less as compared to when the EDF was running at full throttle. All of this is probably an acceptable price to pay for keeping everything in balance. Since we are talking about an imbalance of longitudinal thrust, and not pitch, roll, or yaw, the FC won't be able to auto stabilize it for you. I strongly recommend you keep everything in balance. There will be minor thrust imballances created when you apply pitch or yaw control, but those can probably be ignored.

A minor side benefit to this strategy is that the EDF comes up slowly in proportion to the lift motors as you advance throttle so you won't be abusing it by slamming it with instant full throttle all the time.

I assume you will want to turn off the EDF when in forward flight to save power. I suggest you use the sine squared transition curve. That will give you good pitch and yaw control until you are well into forward flight and the regular control surfaces become effective.

As for using two lift props turning the same direction, it can theoretically be done by tilting one forward and the other back to counter the net torque, but I do not recommend it. You will be creating an asymmetric situation that will have secondary effects as you begin to translate forward. It is also very unnecessary. With the popularity of multicopters, propellers are now regularly available in both CW and CCW matched sets. If you try to cancel the torque with your vertical surface behind the EDF you will use up much of your available control authority just holding heading. There is no benefit, only down side to the idea.

I suggest you buy several sets of lift props and balance them ahead of time. You will probably break several before you figure it all out.

All of this comes from understanding the physics of the situation. You don't need to take my word for it, you just need to think it through. Failing that, the aircraft will teach you what it needs as you try to get it flying. It is just easier if you understand what to expect.
Ran D. St. Clair is offline Find More Posts by Ran D. St. Clair
Last edited by Ran D. St. Clair; Jul 31, 2014 at 10:53 AM.
Reply With Quote


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Sold Vintage Combat Models F-16 NIB rare kit. nero1156 Aircraft - Fuel - Airplanes (FS/W) 1 Jan 25, 2015 07:56 PM
Sold ★☆ OS vz-b and RB S7 ☆★ Blacktiger355 Cars - Accessories (FS/W) 11 Dec 30, 2013 07:47 AM
Discussion Looking for Doak VZ-4DA plans. IWC VTOLs 0 Apr 06, 2013 09:18 PM
Discussion Looking for Doak VZ-4DA plans. IWC Electric Ducted Fan Jet Talk 2 Apr 06, 2013 09:36 AM
Doak Model 16 - VZ-4DA JonJohnston Electric Ducted Fan Jet Talk 7 Sep 26, 2006 06:47 AM