Orion aka "Ol' Boom-boom"

[IsaacHeron]

It would be a bit boring too. Flying in one direction all the time.

You're right. Would make more sense as an arcade game: repair/maintain the ship, destroy obstacles, launches the probes, etc.

[Canopus]

I love this Liberty ship. All our spaceflight problems would be solved. Of course it would be awesome to have every engine on you site as a mod

[nyrath]

arrgh and double arrgh.

I was just about to release a new version when a rather serious bug popped up. Weird too.

When you go to the Tracking Station, whatever Orion ship is first in the list has dead engines. That is, they will not respond to the Z key nor the throttle. All the others are fine.

If you delete that vessel, the new top vessel has dead engines.

There will be a delay as I try to track this down.

[nyrath]

New version for testing

see first post for download.

The bug scuttled away, I cannot reproduce it. Be on the lookout for Orion ships that when you move the focus to them apparently have dead engines.

New Features:

Users can now create their own magazine types and have them show up on the Orion Engine right-click menu.

Engine now shows user flag

Pusher cycle animation tweaked so it has larger extension.

Added 20MN magazine.

Removed 2MN magazine (nobody was using it).

Made the 3.5MN magazine hold 60 bombs instead of 92.

Lowered the mass of the 40MN magazines

Edited June 13, 2013 by nyrath

[Canopus]

The flags on the ship are cool. Nice Update

[nyrath]

The flags on the ship are cool. Nice Update

And after I uploaded that update, I found it unexpectedly easy to make parts attached to the pusher plate move with the plate animation. But I'm not going to make another update until a few people have tried this one. That phantom bug concerns me.

Other things I'd like to do include making the various atom bombs some sort of resource, so you can re-fill your tanks.

The next update will let you attach landing jacks to the pusher plate, and the jacks will move with the plate. This will avoid the need to put your jacks on ridiculous outriggers. But in any event, trying to land under atom bomb power is a challenge.

You can attach other things besides landing jacks, like extendable ladders so the Kerbals can make their way down the pusher plate as it glows blue in the dark.

Edited June 14, 2013 by nyrath

[Tommygun]

But in any event, trying to land under atom bomb power is a challenge.

Well you do get to create your own glass floored and self lighting parking lot wherever you land.

[Splode]

Well you do get to create your own glass floored and self lighting parking lot wherever you land.

"Houston, we're coming down in a boulder field... nevermind"

[Scoundrel]

Just an observation here, but considering the mass of the drive and the large amount of sudden force that the pusher plate is expected to absorb repeatedly (not to mention the giant shock absorbers), why would anyone need to put landing jacks on that thing except, maybe, to level it on an incline? Considering the Gs that ship has to endure again and again and again, I think it's safe to say that it could handle coming down at 20m/s on Kerbin.

On that note, a bug report: I'm getting phantom forces when I build an atomic rocket using just 6x 400MN canisters and then a CM+ASAS. It's kind of fun when you add a landercan + 2m ASAS module + cupola, then IVA the cupola and launch while everything is spinning like dinner plates. I imagine that's what it's like to be Scott Manley on acid. FYI: I have no addons or anything except for ol' boom boom (I build separate KSP folders for each of the addons I experiment with. Saves me problems down the road) so I know for certain that it's not an interaction with some other mod.

Oh, and before I forget, there's a strange interaction between the collision mesh and parts that partially clip the main body that cause everything to act like a sepatron just went off. Might have something to do with the phorces (it's a portmanteau of phantom and forces that I just created at this moment because I know you're busy and don't want to waste time reading "phantom forces" again, because I try to be thoughtful that way, so I hope you appreciate how much time I saved you from having to read that extra word - that's time you could better spend with your family or coding or re-reading a good Heinlein or Alan Steele novel), but it could also be the difference in mass between parts that's causing them to fly off, so it might just be KSP's physics engine.

And finally, I think it's time someone built an adjustable procedural ladder for use on our ATOMIC ROCKETS. And by someone I mean not me. And by adjustable procedural ladder I mean a sci-fi themed bar. And by ATOMIC ROCKETS I mean... wait, no, I actually meant that part... unless I'm allowed to mean BRAZILIAN SUPERMODELS, in which case that's what I totally meant.

[Canopus]

I have a question would a downsized orion powered by conventional explosives work as a probe engine?

[Sufficient Anonymity]

@Scoundrel I've been having the exact same issue - part clipping results in massive unplanned disassembly at launch (which can't even be avoided by turning off crash damage and switching unbreakable joints on), and with particularly large ships, crashes the game.

That said, I've still managed to design this - I'd have prefered to have been able use more panels to blend the Orion drive into the B9 parts, but anyway, presenting Prospero, a heavy interplanetary cruiser:

The cargo bays and solar panels are closed and retracted respectively when under acceleration. There's one current design flaw, namely massively insufficient amounts of monoprop. I have to be fairly conservative with RCS use when turning, and thus I can end up waiting a good 5-10 minutes for her to come about fully (but then she is around 700 tonnes).

Caliban and Sycorax to follow, and yes, I use Shakespearean characters to name my ships

Edited June 14, 2013 by SufficientAnonymity
Misc. stuff

[Tommygun]

I have a question would a downsized orion powered by conventional explosives work as a probe engine?

They tested the Orion concept on the small scale with dynamite.

Also the rumored Aurora aircraft may have use a pulse detonation engine with liquid or gas as the fuel source.

It gets more interesting about 7:12 into the video:

Edited June 14, 2013 by Tommygun

[Canopus]

They tested the Orion concept on the small scale with dynamite.

Thats why I'm asking. You would'n even need a pusher plate. The question is just if it would be efficient.

[magnemoe]

Thats why I'm asking. You would'n even need a pusher plate. The question is just if it would be efficient.

Realisticly it would not be an good idea, explosives give poor dV compared with other fuel and the orion itself is ineffective, however nukes contains so much energy it makes up for it.

[nyrath]

Just an observation here, but considering the mass of the drive and the large amount of sudden force that the pusher plate is expected to absorb repeatedly (not to mention the giant shock absorbers), why would anyone need to put landing jacks on that thing except, maybe, to level it on an incline? Considering the Gs that ship has to endure again and again and again, I think it's safe to say that it could handle coming down at 20m/s on Kerbin.

Ah, well, what you say is true, in reality. The question is whether I can get the Kerbal physics engine to cooperate. When I tried landing on the plate under atom bomb power about a month ago, the shock detonated the engine. I tried using landing jacks, which did not help at all. That was when I noticed that the jacks did not move with the pusher plate animation.

I probably have to play with the engine's part.cfg file, changing the crashTolerance or something. And try landing via weaker auxiliary rockets, like an LV-N Atomic Rocket Engine.

On that note, a bug report: I'm getting phantom forces when I build an atomic rocket using just 6x 400MN canisters and then a CM+ASAS. It's kind of fun when you add a landercan + 2m ASAS module + cupola, then IVA the cupola and launch while everything is spinning like dinner plates. I imagine that's what it's like to be Scott Manley on acid. FYI: I have no addons or anything except for ol' boom boom (I build separate KSP folders for each of the addons I experiment with. Saves me problems down the road) so I know for certain that it's not an interaction with some other mod.

Ah, by "phantom forces" do you mean the parts on top of the Orion engine chattering like Mexican jumping beans? Arrgh. I was having that problem with the magazines. With that, the problem was that the Kerbal physics engine did not like small dense objects on top of large uncompressed objects ("dense" meaning having high mass but low volume). I fixed that with a kludge, magically moving the mass from the magazines into the engine.

So theoretically, the engine should be a massive but uncompressed object underneath your stack of landercan and cupola; everything should be fine. But as per your report, it isn't. Perhaps you could send me a *.craft file so I could try and mercilessly kill this bug?

Oh, and before I forget, there's a strange interaction between the collision mesh and parts that partially clip the main body that cause everything to act like a sepatron just went off. Might have something to do with the phorces (it's a portmanteau of phantom and forces that I just created at this moment because I know you're busy and don't want to waste time reading "phantom forces" again, because I try to be thoughtful that way, so I hope you appreciate how much time I saved you from having to read that extra word - that's time you could better spend with your family or coding or re-reading a good Heinlein or Alan Steele novel), but it could also be the difference in mass between parts that's causing them to fly off, so it might just be KSP's physics engine.

@Scoundrel I've been having the exact same issue - part clipping results in massive unplanned disassembly at launch (which can't even be avoided by turning off crash damage and switching unbreakable joints on), and with particularly large ships, crashes the game.

I'm sorry the mod is giving you problems.

Ummm, can you please both go into more detail about part clipping? Was this with the Orion engine with the magazine rack on top or the one without? By "clipping" I assume you mean that some part stacked on top intersects with the engine.

There used to be a problem if anybody used the engine with the magazine rack, and tried stacking another part on top of a column of Orion magazines. This was caused due to the unfortunate fact that there was a single collider mesh for the entire engine, and colliders cannot have concave holes in them. The hollows in the magazine rack were actually covered with the collider. The magazines did not collide because their part.cfg turned off collisions.

I later used mulitiple collider meshes in order to allow concave holes, and the playtester reported that the problem went away.

In the first picture, the green wireframe is the collider for the top of the engine. Any other part that intersects the green has "collided" with the engine. The second picture just shows the collider for the central column and the cap. The third shows the colliders for two of the side racks.

That said, I've still managed to design this - I'd have prefered to have been able use more panels to blend the Orion drive into the B9 parts, but anyway, presenting Prospero, a heavy interplanetary cruiser

Ooooh, that's excellent!

I have a question would a downsized orion powered by conventional explosives work as a probe engine?

Alas, no. Conventional explosives are not as efficient as solid or liquid fueled combustion, at least when it comes to specific impulse.

The Orion works because nuclear explosives are more efficient.

[warmadmax]

don't know what the rover wheels use for suspension, but could a similar use be applied to the pusher plate to cushion a ground impact?

don't know what a ground impact would be like in relation to a nuke going off though, so might not even move the pusher plate!

[Sufficient Anonymity]

@nyrath Could have sworn I'd remembered to reply, but clearly I didn't, so here goes again:

The explosions I've been experiencing have been when structural panels intersect the main body of the drive (with integral magazine rack).

[RoboRay]

I wish they'd put the Delete Post button back.

Edited June 14, 2013 by RoboRay

[Scoundrel]

Ah, well, what you say is true, in reality. The question is whether I can get the Kerbal physics engine to cooperate. When I tried landing on the plate under atom bomb power about a month ago, the shock detonated the engine. I tried using landing jacks, which did not help at all. That was when I noticed that the jacks did not move with the pusher plate animation.

I probably have to play with the engine's part.cfg file, changing the crashTolerance or something. And try landing via weaker auxiliary rockets, like an LV-N Atomic Rocket Engine.

You really need to up those numbers. A lot. Currently it's less than 1/5th that of a modular girder segment. It should probably be about 3x times that, IMHO. Also, the LV-N sucks for landings! With their TWR you'd have to carry, like, 40 of them if you have any real mass worth of nuclear magazines. I would posit that the best option would be an aerospike since you're going to be using them in both atmosphere and vacuum. Decent TWR, TV is irrelevant considering they're basically for that last 100m, and Isp is good enough to use in both atmosphere and vacuum. I've had decent success with them in trying to land my orions - just have to figure out how to get that damned CoM lower so it doesn't keep rolling around on the baseplate like a quarter before toppling over... or maybe... just maybe...

Ah, by "phantom forces" do you mean the parts on top of the Orion engine chattering like Mexican jumping beans? <snip> So theoretically... <snip>... everything should be fine. But as per your report, it isn't. Perhaps you could send me a *.craft file so I could try and mercilessly kill this bug?... <snip>... Ummm, can you please both go into more detail about part clipping

Okay, I've discovered a few interesting things in trying to recreate it, and have set up 3 orions to display various clipping behaviours, as well as a fourth craft file to act as a control to show how clipping behaves with normal stock parts so you can see the differences in behaviour. Click here for the 4 craft files. Also, you will be relieved to know that the phorces do not appear to be actually present on the parts up top - what happened was a VAB glitch duped the ASAS for some reason and that caused the parts to separate. This doesn't appear to be linked to your mod, it's a relatively well known bug in KSP.

Also, no dead engine bug that I can see. I have three rockets in orbit (switch between them regularly) and I can report that all systems are nominal and glowing an appropriate colour of radioactive blue.

Still waiting for Jeb to grow that third arm though...

There used to be a problem if anybody used the engine with the magazine rack, and tried stacking another part on top of a column of Orion magazines. This was caused due to the unfortunate fact that there was a single collider mesh for the entire engine, and colliders cannot have concave holes in them. The hollows in the magazine rack were actually covered with the collider. The magazines did not collide because their part.cfg turned off collisions.

Interesting... very interesting. If I may ask, why remove the collider from the magazines and not remove the collision mesh on the racks? Is this why mags don't have a "full/empty" mass state themselves? Is full/empty mass calculated as part of the engine mass? If we jettison mags does that affect vessel mass?

Before I forget, I also thought you might be interested in this little post from a wobbly rocket thread:

Joints between rigid bodies are usually modelled as springs. Given a force or torque applied on the joint, the joint will bend to satisfy F=kx, where F is the force, x is the amount it bends, and k is the spring constant. There is also a threshold F that, if exceeded, breaks the joint.

The "wobbly" problem can be solved by increasing K: the same force would then produce a lower displacement, reducing the wobble.

The exploding rocket problem can be solved by increasing the threshold F at which the joint breaks. To test this, try turning on "unbreakable connections" in the debug menu. I betcha it just sets k and threshold F to ridiculously large values.

Perhaps you could include a specific plugin with your own rocket so the other parts use stock wobbliness, while the Orion hulls and mags have a much higher k&F than regular KSP parts - this might solve a lot of part compatibility issues and allow mags to have actual full/empty mass. Just a thought.

[nyrath]

The explosions I've been experiencing have been when structural panels intersect the main body of the drive (with integral magazine rack).

Structural panels, hmmmmmm. I'll check that out.

Okay, I've discovered a few interesting things in trying to recreate it, and have set up 3 orions to display various clipping behaviours, as well as a fourth craft file to act as a control to show how clipping behaves with normal stock parts so you can see the differences in behaviour. Click here for the 4 craft files. Also, you will be relieved to know that the phorces do not appear to be actually present on the parts up top - what happened was a VAB glitch duped the ASAS for some reason and that caused the parts to separate. This doesn't appear to be linked to your mod, it's a relatively well known bug in KSP.

Thanks! It is invaluable to have samples that exhibit the bug. So I can more or less assume I've stomped the bug if the samples start behaving. And I'm glad to find that the phorces was not my bug after all.

Also, no dead engine bug that I can see. I have three rockets in orbit (switch between them regularly) and I can report that all systems are nominal and glowing an appropriate colour of radioactive blue.

Yes, I had a feeling the bug was due to saved vessels constructed with a previous version of the parts suite.

Interesting... very interesting. If I may ask, why remove the collider from the magazines and not remove the collision mesh on the racks? Is this why mags don't have a "full/empty" mass state themselves? Is full/empty mass calculated as part of the engine mass? If we jettison mags does that affect vessel mass?

Ah, I did not make myself clear. The magazines have collider meshes, but in their part.cfg file their "attachRules" have a value of 1 for allowCollisions. This allowed them to interpenetrate the old magazine rack's collider mesh without unfortunate side effects.

In the current version, I carefully crafted the collider meshes on the rack to follow the contours of the rack. I kept the collider meshes on the rack frame to allow people to attach RCS and struts to the rack frame.

Magazines have their mass recalculated after every bomb expenditure. If you look at the magazine part.cfg file you will see it has values for an empty magazine, plus a mass per bomb. Multiply mass-per-bomb by number of bombs, add the empty mag mass, and you have the current mass.

In the latter kludge, this mass is removed from the magazine and added to the engine. If the magazine is jettisoned, the mass is removed from the engine and re-added to the magazine, before the magazine goes its merry way.

If you can read C# code, this is handled in the file OrionMagazine.cs in the source code supplied with the mod. upMagazineMass() keeps the mass updated in the engine instead of the magazine, and if you jettison the function onPartFixedUpdate() removes the mass from the engine.

Before I forget, I also thought you might be interested in this little post from a wobbly rocket thread:

Joints between rigid bodies are usually modelled as springs. Given a force or torque applied on the joint, the joint will bend to satisfy F=kx, where F is the force, x is the amount it bends, and k is the spring constant. There is also a threshold F that, if exceeded, breaks the joint.

The "wobbly" problem can be solved by increasing K: the same force would then produce a lower displacement, reducing the wobble.

The exploding rocket problem can be solved by increasing the threshold F at which the joint breaks. To test this, try turning on "unbreakable connections" in the debug menu. I betcha it just sets k and threshold F to ridiculously large values.

Perhaps you could include a specific plugin with your own rocket so the other parts use stock wobbliness, while the Orion hulls and mags have a much higher k&F than regular KSP parts - this might solve a lot of part compatibility issues and allow mags to have actual full/empty mass. Just a thought.

Umm, actually if you look in the OrionMagazine code you will see I apply what I call "atomic clamps." These are "ConfigurableJoint" which I use to attach each magazine to the engine, said joints are set so they have their xyz motion and xyz angular motion set to a value of "locked". In theory this should have stopped the wobble, in practice it only greatly reduced it.

NovaSililko said the Kerbal physics engine does not play well with dense objects (high mass + low volume) which perfectly describes the magazines. That is why I sucked the mass from the low volume magazines and injected it into the high volume engine. The vessel's overall mass does not change, so the performance should be unchanged.

The problem was not that the joints were breaking. The problem was that the magazines were jerking around like they were on invisible tethers.

But I will look to see if I can access the spring joints so I can increase the value of K.

[nyrath]

Thanks, Scoundrel! The test vehicles enthusiastically explode on the launch pad the instant the physics turn on, and the control vehicle just sits there; just like it says on the box. This will be a great help.

[Canopus]

Rareden fixed the picture

[nyrath]

excellent picture Canopus!

Somebody suggested that I talk about my Orion mod on my website

http://www.projectrho.com/public_html/rocket/sealofapproval.php#id--Computer_Simulation--Kerbal_Space_Program--Orion_nuclear_pulse

[Canopus]

Good that its on your website now

[Gringalf]

Squee! That Orion mod is super fun! Keep up the great work!

Here are some things I noticed:

1.) The pusher plate is really durable and suitable to use as landing gear. I managed to land at > 100 m/s and that thing didn't break. (pictures: http://imgur.com/a/11iR1) However, less durable and well-fixed ship components may break off and/or explode. Also, you have to touch down on the pusher plate. Landing sideways usually makes it explode.

2.) It would be nice if bomb feeding could be managed with action groups.

3.) There's a bug with deactivating/activating the bomb feeding via right-click: it always applies to all engines that were placed together by using the symmetry tool. I'd expect it to affect only the engine I right-clicked.

In case you're wondering why anyone would want more than one such engine in his space ship: I had the idea that if the ship is too heavy to be turned, I simply need one engine pointing in every direction. Like this: http://i.imgur.com/qub5642.jpg (It didn't work very well - it tends to start spinning slowly, and there is nothing you can do to stop that.)