The Great Piston Engine Thread

[BlueCanary]

MK2 cargo bays as cylinders wouldn't work because the door colliders are not fully functional.

But I use them for simple bearings. I'll upload one soon. 13 parts.

Awesome, I want to try and figure out the lightest, lowest part car I can (driveshaft, wheels, clutch, maybe very simple gearbox) for whenever an ICE arrives to power it.

[Redshift OTF]

Yeah I tried Mk 2 cargo bays as cylinders and they were pretty bad. The collision skins on them seem a bit buggy and anything the size of a mk 2 fuel tank moving inside them gets stuck very easily.

This is where I'm at with an ICE. It's only decorative at the moment as the axle bearing is producing the power not the cylinders but it's a start. It only works 1/6th throttle right now and gets stuck fairly regularly, mainly because there is too much play in the rings I made from octagonal struts but there's a lot of ways to tweak this. Eventually I will mount the drive shaft under the power unit and maybe half its size.

Edited September 16, 2015 by Redshift OTF

[BlueCanary]

I was doing some testing for wheels that could be used with ICEs, the best thing I found was using landing gear on the outside of the wheel with the brakes set, they seem to grip quite a bit more than using other parts. Also, to overcome the low power of current engines, you can disengage the brakes and let the wheels spin up without having to move the vehicle, then engage the brakes inertia starter style to jolt the vehicle into movement.

[Redshift OTF]

EDIT: Also do you have a subassembly for a small, low part bearing somewhere?

I'm working on this. When I have it perfected I will post it.

- - - Updated - - -

OK, I'll upload this in the interim. It is a double-ended axle with engines with a small double ended bearing on the back. You can remove the small bearing and use it how you wish.

Test engine and bearing

The whole setup is only 140 parts ignoring the launch clamps and probe core. I think the small bearing is only 21 parts!

- - - Updated - - -

I was doing some testing for wheels that could be used with ICEs, the best thing I found was using landing gear on the outside of the wheel with the brakes set, they seem to grip quite a bit more than using other parts. Also, to overcome the low power of current engines, you can disengage the brakes and let the wheels spin up without having to move the vehicle, then engage the brakes inertia starter style to jolt the vehicle into movement.

Good idea! I never thought of using retractable landing gears as a clutch mechanism as Azimech and you have suggested.

[Azimech]

What you see here is a turboshaft, driving a clutch, driving a gear, which is driving an axle. This method of using gears is new, the big plus being colliders not biting each other, creating massive amounts of drag and/or wrecking everything.

The downside is that it's heavier, you need to brake the wheels before you start and there can be some slip if there's a lot of slack in your bearings.

This was a first try, I'm very pleased with the result!

Javascript is disabled. View full album

And I can add another invention to the list.

The golden glow is due to Squad's overheating bug, so I just switched off max temp.

Edited September 22, 2015 by Azimech

[Azimech]

Almost ...

Javascript is disabled. View full album

[Bubbadevlin]

OMG wow..

[Redshift OTF]

Oh this is looking nice. Have you managed to get some kind of combustion engine going yet? Mine hasn't worked reliably enough to do much with it yet which is a shame because I think I've come up with the perfect "ignition system".

[Azimech]

No, I've refined part count and size a lot, and I have an untested stock mechanical solution but an electronic solution is still far away.

Edited September 23, 2015 by Azimech

[erasmusguy]

Two speed gear box. It works by lowering landing gear into the cogs above. It still needs a lot of love but it works more than it doesn't. Its major problems so far are that you have to switch to the gearbox to change gears and the landing gear sometimes want to do its own thing. Im also working on a version where landing gear slides the selector into place like in real gearboxes but its taking longer to figure out than id hoped.

edit: turns out those landing legs break far to easily.

Edited September 25, 2015 by erasmusguy

[Azimech]

That looks great! I don't really understand it's operation though.

[Azimech]

It's time to bring this topic back to life.

[Azimech]

Klond made this fantastic flat-four with drivetrain and steering!

 

[klond]

 This whole thread is oddly satisfying.

 Might as well keep everything in one place.  I'm gonna pimp the ol' mechanical valves while I'm here.

video

mechanical valve demo

 

 If we had 1 input sensor (crank position maybe) that controlled one digital output (flip fuel flow on/off or spin a SAS module) we could make a lot of things.  Essentially a transistor -something where an input causes an output of our choosing.

[Azimech]

1 minute ago, klond said:

 This whole thread is oddly satisfying.

 Might as well keep everything in one place.  I'm gonna pimp the ol' mechanical valves while I'm here.

video

mechanical valve demo

 

 If we had 1 input sensor (crank position maybe) that controlled one digital output (flip fuel flow on/off or spin a SAS module) we could make a lot of things.  Essentially a transistor -something where an input causes an output of our choosing.

I invented this. Check over here:

 

[klond]

 Instead of temperature, is it possible to do through proximity of 2 parts, or by breaking an invisible beam?  It would avoid the analog to digital conversion.  I don't know KSP mods/modules.

[Azimech]

That's exactly what my system uses: invisible beams :-)

[life_on_venus]

On 05/09/2016 at 8:39 PM, klond said:

 Instead of temperature, is it possible to do through proximity of 2 parts, or by breaking an invisible beam?  It would avoid the analog to digital conversion.  I don't know KSP mods/modules.

Klond, you can use a solar panel as your sensor; a separate craft maybe spins a rover wheel or something and has no batteries, only a solar panel. As the crankshaft turns it shades the panel and causes the wheel to stop moving. The wheel controls a camshaft or something

Edit: It seems you would need some kind of amplifier to make this useful, as one rover wheel moving may not have the torque to be useful. I have no ideas right now but I'm sure you can think of some!

Edited September 8, 2016 by life_on_venus
Added ideas

[klond]

 I like this line of thinking.  IDK if I'm doing something wrong, but I can't get a rover wheel to turn unless it's touching the ground.  Is it just me?

 I wonder if a headlight will power a solar panel, like a perpetual motion machine.  Makes me wonder if any of DaVinci's machines would work in KSP.

 Maybe 1.2 will have some new toys.

 

[life_on_venus]

12 minutes ago, klond said:

 I like this line of thinking.  IDK if I'm doing something wrong, but I can't get a rover wheel to turn unless it's touching the ground.  Is it just me?

 I wonder if a headlight will power a solar panel, like a perpetual motion machine.  Makes me wonder if any of DaVinci's machines would work in KSP.

 Maybe 1.2 will have some new toys.

 

Yeah I don't think perpetual motion is possible even here (unless you use a kraken drive!).

With regards to the wheels not turning without being on the ground, maybe it's the new "friction control" feature? or drive control.

This is the closest I think you can get to "digital" signal processing in KSP, the only other way being mechanical.

So far the rover wheel method would only allow for rotary motion however if you made some kind of a wall climbing device robot thing it could be used as a solenoid: 

Just a random vid I found  That kind of thing with a stick protruding from the top would make a good solenoid for opening valves methinks but would be quite slow

[Azimech]

I like the train of thought you guys have, some nice out-of-the-box thinking!

Unfortunately, I've tried it in the last 3 years. Solar panels don't work the way they should. Using heat is the only way to go forward ... but when you learn how to use it, it can be used for almost anything!

[Wabbit]

Man, this is really interesting to watch. Can't wait to see more compact, more efficient, and more stable designs coming out. Maybe as SQUAD introduce new parts we'll see the oppurtunity for more tech like this?

[Azimech]

My straight six having a thermal analysis.

 

My radial (WIP)

12 minutes ago, Wabbit said:

Man, this is really interesting to watch. Can't wait to see more compact, more efficient, and more stable designs coming out. Maybe as SQUAD introduce new parts we'll see the oppurtunity for more tech like this?

But we'll always be stuck with the stock material setting, having a friction value which is way too high to make efficient, fast running engines possible. The only way to really enjoy piston engines is to use mods or they'll always be slow crawling animals and certainly never capable of powered flight. I know @Majorjim! won't agree :-P

[Majorjim!]

18 minutes ago, Azimech said:

My straight six having a thermal analysis.

 

My radial (WIP)

But we'll always be stuck with the stock material setting, having a friction value which is way too high to make efficient, fast running engines possible. The only way to really enjoy piston engines is to use mods or they'll always be slow crawling animals and certainly never capable of powered flight. I know @Majorjim! won't agree :-P

Not at all man. I never have. My comment about pistons before was that they were possible stock. I said nothing about their performance. ;-)

[life_on_venus]

On 05/09/2016 at 8:39 PM, klond said:

 

Azimech, how does the whole system of using heat work? I admit I haven't done much searching for old posts on it but would love if you could explain