@harrison_warrick

V3 fabrication underway!

Right now, parts are being made for cloud diver v3, with hope for ground testing to begin by the end of the week at the earliest. I’ve come into possession of a p1s, which has dramatically improved print times, and as i write this its chewing through LW pla to make the other central elevatoraron (or whatever the name is). the central wing is done, and just needs to be glued together, and the last outstanding cad task (the wing retraction servo mount) is done, its been installed in the payload, which is somewhat contrary to the whole “the payload stands alone” bit, but im my mind theres no need for the wing servo if you dont plan on flying weather missions, so no point in lugging the thing about if you are just going to fix the wings extended.

V3 Cad Done

It took a long time, it made me suffer, but its done. New features include: an extra 200mm of wing at a 6412 profile for extra lift at low speed (it increases drag but shhhhhhh), a proper hotswappable payload (I’ll be doing flight tests unladen), and more control surfaces (and actual roll and yaw, isnt that crazy). Manufacturing is already underway, but im waiting on a roll of foaming pla to make the wings and whatnot out of. In the meantime im going to do some CFD sims to figure out how slow i can get this thing to take off.

chudbot lives (kinda)!

The motion base of the first prototype is mechanically complete and assembled, code is waiting on an order of raspberry pi picos that have enough PWM pins to drive all 12 servos onboard. the arm parts are printing right now. Ai integration is looking increasingly difficult and scary so this may end up being published sans hivemind part until I have the time to slam my head against the wall to deal with it.

HES THICC

wings are on the cad, and probably will be unchanged from the original design, but now it measures in at over a meter wide (of which 1000mm of it is wing… current take off speed is 7 m/s or bust).

The payload underslung bit has also been given a proper mount, so now it’s a dovetail with some holes for screws to hold it in…. much better for payload swapping, and i can just glue the nose and electronics bay together as one piece

speaking of which, next steps are a minor nose cone redesign, most likely to relocate the camera mount to one of the booms, the edf mount (and possibly the wing servo, im playing with adding in a pulley to let me mount it above the wing), and the tail being designed. tragically the 3-in-1 design is being killed off, now theres two type of control surfaces, pitch/yaw, which is basically the old rudderons at a new angle, and pitch/roll, which will be the center control surface split down the middle. I know that will have piss week authority but given the wings are liable to fold up in anything more than a 30 degree bank i dont think that will be much of a problem… getting it wired properly on the other hand…

also theres another 20 minutes i recorded but lapse bugged and wouldn’t show me my projects to add the time to. Also the first prototype of the quadruped hivemind (tentative name “chudbot”) is half assembled, and will be wired to a random arduino uno i have for walk cycle development and ai integration.

pain, suffering, and redesigns

So, cloud diver flight test 1 has been postponed. One of the rudderons broke and i couldn’t fix it so i need to redesign the part.

In other news, V3 work has begun and its quite a bit different. The older designs had a problem with being very large and heavy for what they needed to do, which meant that at full load (about 1.3kg total) cloud diver would need to get going about 30 miles and hour to even begin to fly. Needless to say this was a bit impractical, so I’ve scrapped most of the fuselage design to make something a bit better (hopefully). What was the fuselage is now the payload, which will be almost entirely self contained, only having a wire or two for signaling with the flight controller, which leads me to the second bit, which is that it will have a proper onboard flight controller (once i can get one). The biggest change is a new central wing section (based on a 6412 aerofoil for added lift) increasing total wing area by about 25%, and a new twin boom design which will have one boom for electronics (flight computer, ESC, etc) and the other for the battery, with some holes in the middle wing for wiring and whatnot and some 1/4 inch aluminium piping for the tail struts. New mount for the wing retraction servo isnt modeled yet. Between additional wing area, some fiddling with the EDF mount and the hopefully massive drop in dead weight (the tail struts will only weigh 28 grams, the old tail assembly was nearly 125), and some proper landing gear it will be less of a hazard to fly.

A new project approaches

in my drive to get Cloud Diver working, my desk has become somewhat of a mess. In an effort to remedy this, rather than spending an hour sorting the various pieces of electronics, 3d prints, and trash into appropriate receptacles, ive spent 2 and a half hours designing a little robot.

the idea is that these little bots will have a little claw on top, some sensors, and a system to hook them all up to an AI hivemind, at which point i will instruct them to complete tasks, like cleaning my desk, or suffering through some strange psychological torture experiment i have yet to devise (robo-sisyphus anyone?)

(cloud diver first flight pending the bloody smoke clearing… i feel like a victorian urchin choking to death in a coal mine)

Cloud diver at open sauce??? (I forgot what I did during the 45 minutes it’s been a long week here at outpost)

CAD Version 2

only 4 hours of cad later, i’ve fixed hopefully most of the problems with version 1, the parts fit together now with printed pins, and its much more compact. the electronics pay and the payload bay have been fused to make wiring and whatnot easier, and to keep the sensitive electronics upwind of the open section of the payload bay. also the tail servos have pilot holes now… learned that lesson the hard way.

But:
the actual physical implement wont be updated to v2 for a while. I plan to reprint the nose, tail, and payload bay and assemble those with the current parts to make a weird v1.5 hybrid monster for more flight testing. also it will have the outer wings… soon…

Two steps forward, one step back

I’ve been busy these last few days… First things first

INITAL FLIGHT HARDWARE ASSEMBLY
I’ve scrounged together a bunch of (kinda subpar) parts to use for flight testing while I wait for my grant, meaning I have just enough bits and bobs to bring this thing to life, especially now that (most) of the printed stuff is done (with the exception of the payload bay and outer wings) with “minimal” fit errors (the wings are a bit of a pain but c’est la vie)

GROUND TESTING
I installed flight critical hardware (no data logging yet, coming soon while i print off the next batch of parts… more on that later) and it works! the EDF is more powerful than i expected, and its got a hell of a moment around the COM, which…. yeah, more on that in a minute. The rudder/elevon/aileron things work great as well.

DISASTER
I got over confident and went “Oh yeah I can totally get this thing flying no problem even with half the wings missing and subpar electronics and random test parts in place of a payload bay” (spoiler alert: no) and tried to fly it. the combination of the moment i mentioned earlier, plus limited wing area, plus needing three hands to launch the bloody thing meant that it immediately nose-dived into the dirt… and broke the tail in half, bent the wing spars, bent the frame, and managed to disconnect the trasmitter from the esc… not great

WHAT NEXT
Well, I want to go to outpost, so I’m waiting on final approval there, but in the mean time i’ve done lots of little cad improvements, so now there will be actual connectors between segments, meaning less gaps and hopefully less bending. i’ve also changed the output duct for the EDF, so now it deflects upwards slightly, hopefully offsetting the moment and easing its nose down tendencies. I’ve got a cheeky idea for a wing locking mechanism for the outer edges to allow for a less sketchy take off, but thats not quite ready yet.

WE HAVE A GITHUB
(come take a look at github.com/harro09-07/cloud-diver)

Also yes I lied again this is still not the build devlog. I’m applying for funding now because it turns out ESCs are expensive and I don’t have any

Also I’ve had a rather devilish idea for an air launch mount for this thing so there might be another devlog today.

Cloud Diver Side show time!

this is designed to use much of the same parts as cloud diver, and act as a testbed for some future enhancements, the most notable being wings designed to deform to act as control surfaces, which should hopefully reduce drag on this airframe, and improve flight characteristics on cloud diver. it also has the goal of reaching high speeds, at the moment the goal is 100 miles per hour.

I know I said the next Devlog would be assembly, but I had some odds and ends to tie up, and I spent an embarrassing amount of time designing a part that wont actually work.

printing the first section of the wings, by my estimate it will take about 16 hours for that, plus another 16 for the tail. Payload bay won’t be printed until the second prototype, I want to make sure I’m not building an unflyable brick.

Cad V1.1 and the prototype!

Cad v1.1 is primarily focused on adding in the practical segments, mostly the payload bay (to be installed between the nose and wingbox) and battery bay (to be installed just aft of the wingbox.

The prototype, on the other hand, is rather less refined. at the moment all that has been printed is the rudderons, nose, and wingbox (though part of the tail is being printed as i write this). the problem is these parts are quite long, and my poor old printer has a z limit of 210mm, so that, along with constant print failures, has made assembly a long and painful process.

next devlog will be me putting the whole thing together (hopefully).

Cad V 1.0 is done! Last additions were a camera mount on the nose, internal geometry for mounting the big servo to pull the outer wing segments in, and a small fin on the tail to give it a bit of yaw stability (in theory). Printing has started, but ive only succeeded in making the rudderons and the nose before my old prusa MK3 decided to throw a fit and now nothing works. next devlog may well just be me doing a full teardown for maintenance on the old thing (it counts as project time right?) I’ve also started getting parts, currently just the structural rods and the EDF, but more should arrive soon(ish)

A bit of a side quest, a system to divert most of the outflow from the 50mm EDF to the rudderons, which should (in theory) give it a bunch of control authority. Meanwhile printing the test parts for the wings continues but slowly, given I’m having bed adhesion issues.

Engine Mount for a 50mm EDF added to wingbox, initial rudderon prototypes 3d printing

Initial CAD assembly and failing to set up CFD software