Brikoleur's Simple Plane Race [closed]

[peridoot]

On 4/23/2021 at 4:10 PM, Klapaucius said:

I've been thinking of  creating a challenge of building planes without gear.  I love how @peridootinspired this one which is now leading to other ideas.

BTW: With a few small changes, your original plane qualifies for the Asymmetrical Aircraft Challenge.

you called?

[QF9E]

7 hours ago, swjr-swis said:

The good news: I finally managed to clock one run that while losing a few seconds from my previous entry, still retains first place in performance category with a good margin. Even at mach 1-ish max speed, runs of under 5 mins are possible. I'm just tired of trying. So this one will have to do.

The better news (and reason I'm submitting this entry at all at this point): it takes first place in the complexity/part count category, being the first entry to use only 7 parts.

That's a very efficient design, I love it! And I am amazed that you got such a fast time without a capsule, and all that without a vertical tail.

And I feel your frustration. I also got really frustrated with trying to get a fast time. It took me the longest time to get my 3'36" run, and it even came to the point that my neighbour complained as I was cursing so much... 

[QF9E]

7 hours ago, Klapaucius said:

I am still at 11 parts, but have cut costs to 1665.  Control surfaces are expensive, but hinges and cheap wing parts are not.  The hinge is mapped to IK and the servo to JL.  If anyone is keen to try and fly this death trap, here is the craft file:  https://kerbalx.com/Klapaucius/Woland

That's a wonderful bit of out of the box thinking! Sorry to hear that the plane is so hard to fly, it's such a natural idea - given that birds control their flight in this way, how hard can it be? Thanks for sharing it, I might give it a spin if I find the time today.

I think it would be fun, once this challenge is over, to share some of the prototypes we designed that may still have some promise but that we were unable to complete the course with. I've been working on dual rotor helicopters to get around the "coaxial" innovation rule and I managed to build and fly one with two intermeshing rotors, like the real-world Kaman Huskie. Problem is that it is almost impossible to control in the yaw axis (as I understand it, this is a problem with tandem rotor designs in KSP in general) - overall it handles ok, and it is easy to land, but it is almost impossible to make it fly in a specific direction.

 

[Guest]

Congratulations @swjr-swis for taking the #1 spot in the complexity category, and defending it in performance, for an overall third place rank, just one point from @QF9E. That nacelle + engine combo is what I was fishing for, earlier.

I'm not going to give it a patent, however, because it's just a matter of combining two parts in the way the KSP designers intended. So if someone else wants to build another craft around a nacelle + engine, go for it. Just don't make it an exact copy.

– That's a really cool out of the box design @Klapaucius. I really like your lateral thinking. Now all you need to do is make it work!

@camacju @Klapaucius you have until 2021-04-27T06:00:00.000Z to update your entries.

[QF9E]

3 minutes ago, Brikoleur said:

That nacelle + engine combo is what I was fishing for, earlier.

I got that bit, but I was stumped because I thought you'd need a vertical stabilizer or lose yaw stability of your craft. In fact my 2'58" run from yesterday uses the pointy nosed cockpit, the inline F-16 style air intake and a Panther. I haven't tried how well it flies without a vertical stabilizer, but is does fly at Mach 3.

Edited April 26, 2021 by QF9E

[swjr-swis]

2 minutes ago, QF9E said:

given that birds control their flight in this way, how hard can it be?

Ha! You'd think by now we would've learned to flap our aircraft wings for thrust... or to run them on mitochondrial energy... like pretty much every other flying creature we have ever encountered.

Or to autoshape our lifting/control surfaces as required at any point of the flight envelope; use them for seamless transition between V-H-V flight; translate in hover with 6DOF with 'static' wings; make our wings of materials and shapes that take optimal advantage of vortex formation/transition/release for physics-'defying' lift to drag ratios; tail sections offering perfect control without vertical stabilizers; etc etc etc.

Yes yes, we can go insanely fast, and make really heavy/bulky stuff fly, and we're good at making things go boom while flying. But we're like... toddlers, playing at flying, while clearly still having a very limited understanding of the whole concept. Humans are the real life kerbals of the flying world.

 

[Guest]

Just now, QF9E said:

I got that bit, but I was stumped because I thought you'd need a vertical stabilizer or lose yaw stability of your craft.

I haven't flown @swjr-swis's design but I think there might be one change you could make to improve its yaw stability with the same set of parts. The ones I've played with (with the same part count) are fairly stable, they do want to crab a bit at low speeds but not so much it's a real problem.

[swjr-swis]

1 minute ago, Brikoleur said:

I haven't flown @swjr-swis's design but I think there might be one change you could make to improve its yaw stability with the same set of parts. The ones I've played with (with the same part count) are fairly stable, they do want to crab a bit at low speeds but not so much it's a real problem.

The craft file is there to try out. In flight it's very stable. I think much of it is passive stability due to the insane drag of the kerbal behind the CoM. Which is much more forward than you'd expect, because of the jet engine's CoM being artificially pushed far ahead. The wings being slightly cambered add a bit too, but that's mostly roll stability.

Honestly, the only points where I felt it severely lacked control was on the ground. The skid allows no steering whatsoever, hardly does any braking, and has zero  margin for bumping or bouncing. Hence the frustrating part.

[QF9E]

1 minute ago, swjr-swis said:

The craft file is there to try out. In flight it's very stable. I think much of it is passive stability due to the insane drag of the kerbal behind the CoM. Which is much more forward than you'd expect, because of the jet engine's CoM being artificially pushed far ahead. The wings being slightly cambered add a bit too, but that's mostly roll stability.

Honestly, the only points where I felt it severely lacked control was on the ground. The skid allows no steering whatsoever, hardly does any braking, and has zero  margin for bumping or bouncing. Hence the frustrating part.

Thanks, I overlooked the KerbalX link in your earlier post when I first read it. And I agree, all improvised landing gear that I've tried (basically everything with a high impact speed rating) had very bad ground handling, and some of those (the large plate that I added to my Goliath design, for instance) also affects flight quite negatively, especially at high speed.

 

[swjr-swis]

34 minutes ago, Brikoleur said:

because it's just a matter of combining two parts in the way the KSP designers intended

Somehow... I highly doubt this... *any* of what we're doing here... is in the way they intended.

https://i.imgur.com/1400OgM.mp4

(how I picture KSP devs more than once watching us use new parts/features they just added to their game)

Edited April 26, 2021 by swjr-swis

[camacju]

How about an application that both uses parts in the way they were intended and uses the same parts in a way they weren't intended?

Retracted landing gear makes for great landing skids, and you only need two for stability!

Edited April 26, 2021 by camacju

[QF9E]

35 minutes ago, camacju said:

How about an application that both uses parts in the way they were intended and uses the same parts in a way they weren't intended?

In the same category: a lithobraking maneuver that deliberately destroys an engine on landing a rocket. It changes the speed of the craft like an engine is supposed to do, but in a rather unconventional way.

[swjr-swis]

59 minutes ago, camacju said:

Retracted landing gear makes for great landing skids, and you only need two for stability!

Yes, I've tried this one for good effect (my submission numbering may give a clue as to a number of unsubmitted entries). I ended up not using it due to it still being one part more than I needed. They're also rather expensive, although I wasn't going to lose much on that account anyway. But it works!

[QF9E]

With @swjr-swis' tailless craft in mind, during my lunch break I briefly tried my paramotor design without a vertical tail, and it turns out to work fine, albeit with even worse flight characteristics than before: roll stability is almost absent, and it has a tendency to get into an irrecoverable spin at low speeds. However, I did manage to land in one piece, and with some practice I'm confident I could finish the Airfield run with it.

Edited April 26, 2021 by QF9E

[Guest]

6 hours ago, swjr-swis said:

The craft file is there to try out. In flight it's very stable. I think much of it is passive stability due to the insane drag of the kerbal behind the CoM. Which is much more forward than you'd expect, because of the jet engine's CoM being artificially pushed far ahead. The wings being slightly cambered add a bit too, but that's mostly roll stability.

That would explain it! The ones I built used a command pod and were pretty slippery; I did have to find a solution to the longitudinal stability problem without a vertical stabiliser. But I eventually did. I'm looking forward to a debrief once this is wrapped up – maybe we should move the end of contest back a bit so we can do it sooner? I could always run a second round of this with rules somewhat changed to shake things up.

[QF9E]

19 hours ago, swjr-swis said:

So, presenting the Brik-SPR-4:

@swjr-swisWhat engine setting did you use on your runs? I've tried your plane and when I run it full throttle all the way it runs out of fuel. When I run it at 70% there's enough fuel but you lose some 20 m/s top speed. I found that your plane has a tendency to crab below 70 m/s or so (probably due to the missing tail fin), which makes landing quite tricky. I found that lining up with the runway at 100 m/s and then gently glide down while incresing pitch trim to keep descent rate acceptable worked quite well. Your plane generates a lot of drag, which is an advantage when trying to slow down for landing. I did not need any fancy maneuvers - just cut the throttle when near the runway, use a high-g turn to align with the runway and you're at the correct speed for final approach.

After a handful of attempts I got a time of 5:10 at 70% throttle setting. I slightly increased the aileron authority limt as it rolls a bit slowly for my tastes and I moved the landing skid forward a bit - it had a tendency to tumble forwards on landing as the touchdown point at the back is quite a bit behind the CoM. It also flies with a pronounced nose-down attitude in flight, is this deliberate? Perhaps you angled the wings upwards to make takeoff easier?

Edited April 26, 2021 by QF9E

[swjr-swis]

31 minutes ago, QF9E said:

What engine setting did you use on your runs? I've tried your plane and when I run it full throttle all the way it runs out of fuel. When I run it at 70% there's enough fuel but you lose some 20 m/s top speed.

I ran it at full throttle, and it only flamed out moments before touch down  on the KSC runway. I did however climb pretty steeply right from take off and leveled off at  around 3km - at that altitude it uses less fuel and suffers less drag. Then on approach to the island, I first lined up, then pitched into a steep dive so I could throttle way back and flare at the last moment.

 

31 minutes ago, QF9E said:

Your plane generates a lot of drag, which is an advantage when trying to slow down for landing.

Remember there's the thrust reverser as well. The most efficient landing requires engaging it (brakes) and then shortly burst throttle to shed speed. But the landing requires an almost perfectly level touch down... it's very stressy to get all that done at the same time.

 

31 minutes ago, QF9E said:

I moved the landing skid forward a bit - it had a tendency to tumble forwards on landing as the touchdown point at the back is quite a bit behind the CoM.

Like I said... perfectly level end of approach. I placed it there to keep the CoM balance, but you can't afford to bump or bounce in any way.

 

31 minutes ago, QF9E said:

It also flies with a pronounced nose-down attitude in flight, is this deliberate? Perhaps you angled the wings upwards to make takeoff easier?

Hmm, were you adding too much trim perhaps? If I remember correctly, at neutral trim, it wants to climb, no need to 'make' it rotate. Yes, both to help with the take off (to combat the frequent 'stumbling' over completely invisible bumps in the KSC runway), and to regain a natural tendency to flare up on that final approach, when I sometimes needed to Alt-X to completely reset all trim to have a sane last-minute setting and get it perfectly level a meter over the runway.

 

 

Edited April 26, 2021 by swjr-swis

[QF9E]

3 hours ago, swjr-swis said:

Hmm, were you adding too much trim perhaps?

I flew it at about 400 m above sea level like I do with my own designs, and I needed quite a bit of downward trim for level flight at that altitude. I am having a bit of trouble with the concept of "too much trim" - in my view you trim as needed for level flight.

I tried some more, and I am afraid I don't like the 3000 meter altitude approach. I think you lose a bit of time climbing to that altitude, and top speed is a few m/s lower as well. I also tried the thrust reverser but could not get that to run reliably - I had several craft tumble forward and explode on the runway after I enabled the reverser. I then tried a run with 5 units of extra fuel at lower altitude, and I managed a time of 5 min 00:

Note: I forgot to enable world time, but the video runs at normal speed, has length 05:00 (to be exact: 05:00 and 9 frames (at 30 fps), which is why YT shows 5:01 in the preview and 5:00 while the vid is running) and starts and ends with the craft with speed 0 on the KSC runway. Also note that my Island landing left the craft in an awkward position from which I had some difficulty of getting on the correct course back to the KSC.

 

22 hours ago, Klapaucius said:

If anyone is keen to try and fly this death trap, here is the craft file:  https://kerbalx.com/Klapaucius/Woland

I gave your craft a go, but I'm afraid I can't get it to fly stably - it invariably enters a roll to port after takeoff that I can't seem to stop. I find it really awkward that the hinges do not snap back to the center position like the roll, yaw and pitch axes normally do. I tried to bind the hinges to these axes, but could not get pitch trim to work - apparently pitch trim is not sent to the hinge as it remained in the center position no matter the trim position.

Edited April 26, 2021 by QF9E
Fixed embarrassing typo

[swjr-swis]

12 minutes ago, QF9E said:

I flew it at about 400 m above see level like I do with my own designs, and I needed quite a bit of downward trim for level flight at that altitude.

I am not surprised, that's exactly what I would expect if you do that. And it explains...

 

13 minutes ago, QF9E said:

I am having a bit of trouble with the concept of "too much trim" - in my view you trim as needed for level flight.

..this too: you're trying to fly it at a much lower altitude than I build it for, so yes, you end up having to 'trim too much' (or as you worded it, 'pronounced nose down') to force it to stay at that lower altitude. You're literally fighting the plane (with roll control too, I get the impression from the video).

Hats off for that run. It also shows it's indeed capable of under 5 min runs when it all goes well.

 

[Klapaucius]

7 hours ago, QF9E said:

I gave your craft a go, but I'm afraid I can't get it to fly stably - it invariably enters a roll to port after takeoff that I can't seem to stop. I find it really awkward that the hinges do not snap back to the center position like the roll, yaw and pitch axes normally do. I tried to bind the hinges to these axes, but could not get pitch trim to work - apparently pitch trim is not sent to the hinge as it remained in the center position no matter the trim position.

It is very counter intuitive, and how you position the hinge depends on whether you are bent forward or backwards.  A little goes a long way, but the two need to work together.  Outside of takeoff, you need to do everything slowly.

At first, I found it easier to open the editor and use the mouse slider for the pitch and then key the roll.  Or you could do vice versa. It gives a better visual sense of where you are.

[QF9E]

7 hours ago, swjr-swis said:

You're literally fighting the plane (with roll control too, I get the impression from the video).

Looking back at my run, I noticed that I use quite a bit of upwards pitch trim during both takeoffs, and find myself having to force the plane into level flight. The solution seems obvious: don't use that much pitch trim during takeoffs.

I don't quite see it with the roll control, perhaps you can point to a specific moment in the video? I did increase the roll rate, perhaps I overdid it a bit.

And I hope I don't come across as overly critical of your craft. I really like your design, and I enjoy flying it a lot.

Edited April 27, 2021 by QF9E

[Guest]

I've changed the rule about ending the contest. I will now close it if there are no new entries during seven days (or the original date, in case entries keep coming in). The closing date is on the first page; currently it's midnight, May 3. 

[swjr-swis]

5 hours ago, QF9E said:

I hope I don't come across as overly critical of your craft. I really like your design, and I enjoy flying it a lot.

On the contrary: it's feedback from a -clearly- better pilot than I am.

I would not expect this kind of design to win any handling prizes. It's a flying engine with a steel H-profile for a skid and an open seat. Pretty sure we wouldn't find anyone to insure this craft, let alone certify it.

[QF9E]

Turns out a 6 part craft is possible (2 motors, 2 blades, solar panel, lander can):

The reason this works is that with a tail rotor you don't need a vertical stabilizer for yaw stability, as the tail rotor provides yaw control. And the lander can has a built-in battery. And yes, I did manage to make this thing fly without SAS ... for a couple of short hops. I am a newbie when it comes to helicopter design and flying but I flew a precursor to this one (with dual rotor blades on both rotors and proper landing gear) to the Island runway in about 20 minutes, so a total trip time of about 40 minutes should be doable.

Edited April 27, 2021 by QF9E

[OHara]

22 hours ago, QF9E said:

don't need a vertical stabilizer for yaw stability, as the tail rotor provides yaw control

. . . and the disk of the rotor blade gives passive yaw stability as well.

This works with the chair as well, flying very carefully using slow changes in trim, and letting it fly at whatever odd angle works best. 

I wonder if I can make it to the island and back on the charge in that battery.

(Helicopters with fore and aft rotors have enough degrees of freedom to control the craft, if we vary the rotor rpm, but there are awkward interactions between controls and I found them very tricky and unpleasant to fly.)

It looks no current entrant has been overtaken within the past 24 hours, so we are hoping for new entrants to the challenge.