Build Log F-35B Lightning II VTOL - Scale composite build!!!

[G-force237]

Hey guys,

Just wanted to post some progress on a build my brother and I have been working on this past summer. This is actualy a continuation off of a project my brother started last summer, however with a slightly new plan.. So here we are in the VTOL forum.

For those interested, take a look a the original project here: http://www.rcgroups.com/forums/showthread.php?t=891668
That thread has more of the details on the fabrication of the plane its self. we will be keeping that thread up to date as well.

The Plane:

The hover platform will consists of a lift fan, and 3 bearing swivel duct to vector thrust down. The two main fans will also be used as pitch controll by varying fan RPM. Roll controll will be accomplished by two edf 40 'puffers' placed at the wing roots.


The plane its self will be aprox 40 inches nose to tail, and is expected to weigh in some where around 75 oz. Very heavy! but the power source has been sized accordingly (heh heh ). For the lift fan and main engine, we will be using two LSJS1400's installed in a Minifan pro, and a WM400 respectively. We chose to use the WM400 for the main engine as its advertised ability to create a large pressure rise would help minimize losses through the 90 deg bend of the 3bsd. the minifan pro was chosen for its looks (and dynamicaly balanced rotor). The 'puffers' will be 2 edf 40's paired up with a set of Hyperion 5800kv motors. The lot is to be powered by 4 4s 2100mAh Flight Power Evo Light V2 packs.

For controll of the pitch axis we will be varying the RPM's of the lift fan and main fan. The plan is to gyro stabalize the pitch axis with a GY 240, and then use an SM2 mixer to vtail with the throttle chanel. Roll axis will be done by starting each edf 40 at 50% throttle at the centered aileron position and vary rpm up or down based on aileron input. The roll axis will also be gyro stabalized witha GY240 and may also be vtail mixed with a master throttle input. This may aid in the activation of the speed controllers and will also allow the fans to be shut off upon landing or transition. The rudder axis is to be controlled by swiviling the 3bsd left and right just as the real one does. This axis will be gyro stabalized with a telebee zoom.

As for thrust numbers. well.. we hope to get more than 75 oz. More specificaly, our goal is to acheive a 20% margin which comes out to be about ~90 oz of thrust. the way we expect/hope our thrust breakdown to be is as follows. 53oz from lift fan, 28 oz from main fan, 9 oz from 2 edf 40's @ 50% throttle. The 28 oz from the main fan was estimated based off of an assumed worse case 50% loss through the 90 degree bend in the duct.

well I think thats enough talk for now. here are some renders of the overall layout to provide a better idea of what our plan is.

[bwalt822]

subscribed, good luck...you'll probably need.

How do you plan to actually move the swivel duct?

[G-force237]

For the swivel duct, the priority was placed on actuating the segments in a way that is as simple as possible. We figure making a 3 piece swiveling duct that is strong enough and efficient enough will be complicated as is even before worying about actuating it. so the simpler we can make the actuating mechanism, the more feasable it will be.

The plan is to have a built up frame constructed from laser cut ply parts that would provide the 'bearing' surfaces, aswell as structural rigidiy of the over all duct. The way the duct will be actuated is by servos located along the axis of the duct placed inside a fiberglass tube to maintain clean airflow. we will then have stators within the duct to link the outer shell of the duct to this core aswell as to help direct air flow through the 90 degree bend. hopefully this core will introduce minimal impedence to the air flow as it should continue straight back from the motor on the fan. we will also maintain 100% FSA through the bend of the duct, tapering down to 90% through the last segment. Weight wise it is looking like the duct all together with supporting electronics (servos mixers ect.) will come out to be about 5 oz.

here are some pics of the design aswell as some photos of the completed duct segments.

Assembley and testing will start now! More to come soon!

[bwalt822]

Impressive. Will you post videos when the full swivel is complete? Do you plan on testing it under thrust before continuing the rest of the plane?

[G-force237]

Originally Posted by bwalt822

Impressive. Will you post videos when the full swivel is complete? Do you plan on testing it under thrust before continuing the rest of the plane?

Here ya go

yesterday the duct was constructed and after about 10 minuts of ironing out the mixing in the radio we had a fully actuating 3bsd! Bellow is a video showing how the duct can swivel for vertical and conventional flight, aswell as rudder controll. Overall weight 4.8 oz. This includes the bulkhead that mounts the duct.

swivel duct (0 min 14 sec)

This next video is the thrust test. This was a 100% throttle run powered by 3 4s 2100mAh packs. From the video you can see we managed 2lb 9 oz of thrust!! actualy outperforming our expectations by about 13 oz!!

Swivel duct2 (0 min 19 sec)

our next step will be to get all power sources running together and make sure we are are going to get the thrust we expect out of the others, and make sure we dont get excessive voltage sagging as two LSJS's is going to be a whole new story.

[bwalt822]

That is awesome! Have you swiveled it at full thrust? In the video it sounds like you dont go to full throttle until its swiveled and then you throttle down right before you swivel flat but I could be wrong. You will also lose some thrust from the intake.

[G-force237]

We have not attempted to swivel the nosel under full power yet. for that we will need to modify our stand and most likely change locations as to not rearange the furnature in the basement. (we are unseating cealing tiles as is ) so far we have just tried swiveling the nosel up and down under about half power. at the moment it works well when converting from full down to straight, however it has trouble going from straight config to down. it seems the thrust from the fan makes the duct naturaly want to straighten, so the servos do not quite have the torque to fight that.

full throttle tests are definately in the future though because if the duct cannot atleast straighten under full power, transition would be unlikely.

Good point on the intakes. we will be expecting some more losses when the fan finaly gets installed. we are keeping our fingers crossed that it is not much.

[cray2602]

G-force that looks awesome.

I am currently working on my 3rd attempt at a VTOL and i am also using the GWS EDF-40 fans in the wings with Hyperion 5800kv motors. I have the assembly slopped together on a plank so i can move fans around and tinker with everything before comitting it to an airframe. My plan is very similar to yours but rotateing the rear fan itself rather than having the ducting rotate. A few things i have learned so far - With the EDF-40 fans close to the wing roots they needed to much change in RPM to get a drastic difference, it was VERY difficult to control. I plan to move them farther out on the wings to see if the extra leverage gives them more effect. I also started with the front lift fan closer to the CG than the rear fan, that didnt work for obvious reasons. The fore and aft down thrust points definately have to be equal distance from the CG. My rear fan itself pivots left and right for rudder control and it works excellent, better than i hoped.

Not trying to steal your thread here, just thought maybe i could add a bit of what i have ran into so far. Good luck, looks like you guys put alot of sweat and headache into this. Looks amazeing.

[RCvertt]

Originally Posted by G-force237

Bellow is a video showing how the duct can swivel for vertical and conventional flight, aswell as rudder controll...

Nicely done

I would be curious to see a thrust test comparison between this and a regular straight duct of the same size.

[Sky-walker]

my only thought is the loss of airflow due to the un-smooth swivel parts. or perhaps it aint affecting much ?

[mnfiero]

very nice work!

[member2006]

Awesome I thought that was the real thing

As far as airflow loss, it shouldn't be much of a problem. I've messed around with 90 degree tube bends (1 lb model w/vasa fan 65mm and mega motor 2:1 ) and there doesn't seem to be much airflow turbulence. Just make sure you have the the thrust which these guys seem to have.

[VTOLicious]

Great! The swivel-duct is a masterpiece of engineering (like the real one) Looks promising, also for other VTOL applications

You will definitely have some weight issues with all the EDFs, ESCs, Batterys, Servos, Gyros, etc. ...anyways it will be a great plane and at least STOVL capable - scale

Regarding your control system: Have a look at Likais 4 EDF-VTOL Machine (inkl. Video). His setup might work for your plane too.

Rgds Michael

[G-force237]

Cray thanks for the input. As neither of us have done this before any insight is welcome. For now we are probably going to stick with the fans at the roots as bulkheads have already been lasercut and designs have been finalized (for the moment atleast!). We actualy just finished up with a roll controll test to see if we would have the controll authority we needed. It seems that (on the rig atleast) with the help of a GY240 we have a fairly solid platform with only minor wing rock. As for how that translates to free flight we will see. Glad to hear the swivel method for rudder works though! thats a little comforting .

VTOLicious that video you linked was actualy the reasion we went with two futaba heading hold gyros on the pitch/roll axis. Its a great video, it is amaizing how rock solid that craft hovers in the air. We have our fingers crossed that ours is half as stable.

RCvertt, we did test our lift fan which has the same motor and battery as the swivel duct just different fan and has a straight duct. From that we measured thrust to be a healthy 3 lb 4 oz. So thats about a ~11 oz loss from the 90 deg swivel duct. Pretty sizeable, but not nearly as bad as was expected.

So now for a little update.

our main focus continues to be on testing the power sources, but with a healthy fiberglass fuse just pulled from the mold the focus has slowly begun to shift to assembly.

After the last post we started running some tests on our roll controll. For that we took a dummy fuse and mounted the edf 40's in their expected locations and then rigged the up fuse in such a way that it pivotted along the roll axis. To ensure that the plane was not recieving any stability from the rig, the axis was intentionaly placed below the cg to make the plane inherently unstable; That is it continually tries to roll belly up. ofcourse we took video of this. Bellow shows first the inherently unstable rig and then the subsequent test. From what we could tell the edf40's held the plane rock solid with only minor wing wobble. This may require some further tuning of the gyro gain to iron out. It may also be due in part to the constantly destabalizing rig or maybe its the relatively inboard location of the fans as Cray has pointed out. Only an unteathered flight can tell. enough talk.. video..

Roll controll (0 min 17 sec)

You can hear the gyros really working the throttle inputs on the fans. We tried a run without gyros and it is simply impossible to keep the plane steady otherwise. It is the quickness of the gyros that really make the plane stable. The fuse in that video looks pretty we know . That was attempt 1 for our fuselage. It was a sandwiched coremat construction with 2 oz glass on the exterior and 3/4 oz on the interior. despite using sand bags to apply pressure to the skin while curing, the plane came out with many spots having a poor bond with the coar material. It seems vacuume bagging would be the best option for that type of construction. On the bright side it was very light (10 oz) the problem was it had the stiffness about that of warm jello.. you can see it jiggle a bit there at the beggining.. So... vacuume supplies are on the way and for now we are going with a more conventional layup method for the hovering test plane.

[G-force237]

So with roll controll looking acceptable we moved on to testing the entire power system as a whole. starting with short 2-3 second runs to make sure nothing was being stressed tooo much we slowly worked up to a continuous 15 second full throttle burst. This way we could see what kind of sagging we could expect in flight. It was pretty amaizing to have all the fans screaming at once, precautions for ear and eye protection were definately taken!. We do not have a watt meter that is large enough to know for sure but based off of online test data we should be pulling close to 2600 watts. Ofcourse we didnt pass on the oportunity to take video of this either.

15 second full fan run up (0 min 21 sec)

From the video you can see with all fans drawing power the output of the main fan droped about 3 oz off the standalone test, then an additional 2 through the duration of the video. I assume the thrust from the other fans follow suit. more tests are ahead to see exactly how much thrust we can expect to have, this was also not including the intake ducts. Once all is said and done, we will have to see if we are left with enough thrust to hover..

To wrap it up, attached are some pictures of the airframe progress. The fuse has had all the hatches and doors cut out, has been panted, and is now ready for components to be installed! Once we are satisfied with the power system we have, things will start to go in!