Orion aka "Ol' Boom-boom"

[Der Kosmos]

Which reminds me, I have to code it so when the magazine array blows up, it also destroys everything within about 220 meters.

Why? It's not a Zubrin drive. A smashed warhead shouldn't explode.

[nyrath]

Well this is quite awesome. But the size is a bit too large for me, so I changed scale and rescalefactor to 0.5, and the model scaled to a size which is better for me anyway, but the attachment nodes are messed up. I also need to tweak the power of the drive itself, but how would I fix the nodes?

I'm glad you managed to customize the Orion! Can you post some screen caps of your resized Orion ships?

You should be able to tweek the power of the drive by changing the "bombImpulse" parameter in the USAFOrionMagazine part.cfg file. I originally had it set to 1000 kiloNewtons, currently I have it at a historically accurate 3500 kiloNewtons. The latter allows one to boost payloads of 100 tons at about 1.5g

If you want further tweeks, I explain the effects of various part.cfg parameters in the ReadMe file.

Why? It's not a Zubrin drive. A smashed warhead shouldn't explode.

Very true. But since Kerbals are immune to the radiation from smashed plutonium shard, that's the only way to make it more of a disaster. And gamers sort of expect it.

What does everybody think? Should I err on the side of accuracy or cinematic excitement?

Edited May 24, 2013 by nyrath

[Canopus]

I'm all for accuracy:)

[nyrath]

That's two votes for accuracy.

Currently I'm trying to implement multiple tanks. The current "tank" is actually an array of 60 tanks ("magazines") in six stacks of ten magazines. This does not allow the user any fine tuning of the fuel load on the rocket.

I just figured out that one magazine has an internal volume of 4.62 cubic meters, curtesty of Blender's 3D printing tools. Each magazine holds 60 bombs. So if a 3,500 kN one kiloton bomb has a mass of 0.141 tons, the resource density is 1.83 tons per cubic meter. Unfortunately KSP's native resource system does not handle discrete bombs very well. I'm probably going to have to make my own magazine management system.

I'm also unsure how to individually jettison spent magazines. I'm not sure if there is an existing user interface that allows that in KSP. The only one I can find is the staging system. I might have to make it so the engine automatically ejects spent magazines. Otherwise I have to display to the user the load status of each magazine and give them a "jettison" button.

[nyrath]

Well this is quite awesome. But the size is a bit too large for me, so I changed scale and rescalefactor to 0.5, and the model scaled to a size which is better for me anyway

mushroomman, there might be a problem with the pusher plate animation. Watch the flat plate after each detonation. If the plate appears to vanish into the interior of the engine body, edit the USAFOrionEngine part.cfg.

Currently plateOffset and plateSpeed are 4.88. If you rescale to 0.5, change both of those values to 2.44.

If the plate animation looks OK, just ignore this post.

[mushroomman]

Here's my ship. Jeb took it on a shakedown cruse to the Mun and he crashed it. So I need to send a rescue mission some time. The thing gets very difficult to use in the atmosphere with the FAR mod. But it's pretty good. It has two atomic rockets (heh) on the side for fine maneuvering.

and yes, the pusher plate works just fine without changing it.

[nyrath]

Excellent! Glad to hear the plate works fine.

And that is too awesome for words that you are using atomic rockets for fine maneuvering!

[khyron42]

So, I had a curious event happen today. I had an orion interplanetary stage deliver a lander to the Mun, then detached a lander, landed, and returned.

I got to within about 1 km and almost matched orbits, but ran out of fuel and RCS on the lander. So, 1km away, I use ] to switch focus to the Orion drive vessel.

The orion drive and other orion parts disappeared, and all of the other parts (12 of the largest "grey" stock fuel tanks, a hitchhiker compartment+ASAS+RCS tank, some RCS thrusters) go flying in wildly different directions - not any explosion, just that the very wide central component they had been attached to suddenly vanished, and they were at different heights above the Mun so were on divergent orbits.

Any attempts to save the game failed, and using ] to scroll through the options, there was one completely invisible part; I'm guessing that was the drive and magazine, and somehow the graphics (and collision meshes) for them vanished. That's why everything else came apart like that, I think.

[Hremsfeld]

What does everybody think? Should I err on the side of accuracy or cinematic excitement?

Make that three votes for accuracy. Also, I've made an album of launching just a cockpit-and-ASAS and talking through it here; feel free to use it however you want, nyrath.

I'm finishing up my first few missions in .20, so I'll start testing this in a moment, but if you'll pardon me while I backseat mod design again:

The existing decoupling/staging systems work perfectly for ejecting spent magazines, if they're externally or stack mounted. If your heart was set on drum magazines, though, then yeah that's more difficult.

Implementing nukes as a resource...looking through OrionPusherPlate.cs, I see where in launchBomb() you decrement the number of nukes left; if I'm not mistaken, this works for doubles, floats, etc as well as integers. I know it does for C++, but I haven't used C# to know there. I'm going to go ahead and see if I can jerry-rig something, probably using the smaller RCS tank as a model. This'll be...Interesting, because I don't have a C# compiler and don't exactly know how to compile a .dll, but hey, hopefully it'll be easy. If not, I'll send you screenshots anyways of what I was trying to do to see if it gets your lightbulb going. I can't test the engine, because the only thing I could do without a compiler is make a very-high-thrust and high-specific-impulse rocket that operates identically to a liquid-fueled engine but uses nukes.

Okay, took some screenshots of what I'm trying to say, and put them in another album here. If the current engine becomes an actual engine, it can search for magazines above it like liquid engines do now. I think the best option would be implementing both methods, which requires "only" (says the guy not actually making anything) making a stack base above the engine, and making a magazines model. This will allow users to make their own ships look as realistic as they want, but won't prevent an actual realistic ship from being made.

As for the issue of the throttle not doing anything, the throttle only applies to actual liquid engines. Project Orion was obviously neither solid nor liquid fuel, but (oddly enough) seems to behave as a combination of the two. You control how many pulses there are, and you control when they happen, but you don't control how powerful they are. Theoretically in-game, you could scale the bombs' yields to the throttle, but Orion is about riding the shockwave of a nuclear blast, not about finely regulating your orbit.

In case I'm coming off as preachy (I'm not trying to, but it's early in the morning here), I'm not trying to say how you should or shouldn't make your mod, but you've got my brain going coming up with ideas on how this could work, so I want to share them.

Edited May 25, 2013 by Hremsfeld
Tested it a bit

[nyrath]

So, I had a curious event happen today. I had an orion interplanetary stage deliver a lander to the Mun, then detached a lander, landed, and returned.

I got to within about 1 km and almost matched orbits, but ran out of fuel and RCS on the lander. So, 1km away, I use ] to switch focus to the Orion drive vessel.

The orion drive and other orion parts disappeared, and all of the other parts (12 of the largest "grey" stock fuel tanks, a hitchhiker compartment+ASAS+RCS tank, some RCS thrusters) go flying in wildly different directions - not any explosion, just that the very wide central component they had been attached to suddenly vanished, and they were at different heights above the Mun so were on divergent orbits.

Any attempts to save the game failed, and using ] to scroll through the options, there was one completely invisible part; I'm guessing that was the drive and magazine, and somehow the graphics (and collision meshes) for them vanished. That's why everything else came apart like that, I think.

Yikes! That sounds like a serious problem. And it seems to be linked to having a lander.

Would it be possible to get a copy of your craft file, to see if I can reproduce the problem?

Canopus, have you tried using a lander with the Orion yet?

Edited May 25, 2013 by nyrath

[Kerbal01]

I can't get more than one pulse. What am I doing wrong?

[Canopus]

I can't get more than one pulse. What am I doing wrong?

You have to hold "Z"

[nyrath]

Make that three votes for accuracy. Also, I've made an album of launching just a cockpit-and-ASAS and talking through it here; feel free to use it however you want, nyrath.

...

In case I'm coming off as preachy (I'm not trying to, but it's early in the morning here), I'm not trying to say how you should or shouldn't make your mod, but you've got my brain going coming up with ideas on how this could work, so I want to share them.

Thanks for the images! And you are not being preachy, I need this sort of input. My trouble is that while I do know a lot about Orion drives, I am a novice at KSP. I was unsure about how the existing staging system works.

The existing decoupling/staging systems work perfectly for ejecting spent magazines, if they're externally or stack mounted. If your heart was set on drum magazines, though, then yeah that's more difficult.

Ummmm, I'm confused. How is the drum magazine different from an externally or stack mounted magazine? The internal details of the magazine do not matter. The only important thing is how many bombs are left inside. What I was fretting about is how the user can figure out which magazines are empty, and how to decouple just them. The baseline USAF Orion design has 60 magazines, in six stacks of ten high.

The main thing is to give the user the ability to adjust the number of magazines to match the mission. Those things are massive. And there are different sizes of nuclear bombs, the more massive, the more kiloNewtons they put out. I could easily make magazines for each bomb size. All you have to do is clone the part, rename it, and do a couple of edits in the part.cfg file.

You can do none of this if you are stuck with a huge monolithic magazine array, as I have currently.

Implementing nukes as a resource...looking through OrionPusherPlate.cs, I see where in launchBomb() you decrement the number of nukes left; if I'm not mistaken, this works for doubles, floats, etc as well as integers. I know it does for C++, but I haven't used C# to know there. I'm going to go ahead and see if I can jerry-rig something, probably using the smaller RCS tank as a model. This'll be...Interesting, because I don't have a C# compiler and don't exactly know how to compile a .dll, but hey, hopefully it'll be easy. If not, I'll send you screenshots anyways of what I was trying to do to see if it gets your lightbulb going. I can't test the engine, because the only thing I could do without a compiler is make a very-high-thrust and high-specific-impulse rocket that operates identically to a liquid-fueled engine but uses nukes.

When I was making the package, I downloaded the free version of Unity 3D to make the models. I later found out that it comes with the Monodevelop IDE for C#, altered so it works with Unity. Apparently it has the C# compiler as well, at least I did not have to download anything else in order to compile. I used the instructions here to set it up

http://wiki.kerbalspaceprogram.com/wiki/Plugins

Okay, took some screenshots of what I'm trying to say, and put them in another album here. If the current engine becomes an actual engine, it can search for magazines above it like liquid engines do now. I think the best option would be implementing both methods, which requires "only" (says the guy not actually making anything) making a stack base above the engine, and making a magazines model. This will allow users to make their own ships look as realistic as they want, but won't prevent an actual realistic ship from being made

That is excellent!

Currently, my magazine array includes a central core and some framework stiff in between the magazine stacks. That could be moved to the engine model, as long as the collider had holes allowing the stacking of magazines.

I've actually already got a magazine model, I just stretched it into the size of a 10 stack for the current model (to keep the polygon count low).

When you say "make a stack base", you mean placing attachment points on the top of the engine part, not creating an entirely new part, correct?

It would be nice if I could incorporate a decoupler into the magazine, to keep the part count down. And incorporate automatic fuel lines, also to keep the part count down. I was toying with the idea of making the engine magically find all the atomic magazines in the vessel, automatically suck bombs from the most empty magazine until it was dry, and automatically somehow ejecting the spent magazine. Probably a bad idea, but I wasn't sure if KSP could handle 60 decouplers with individual controls.

As for the issue of the throttle not doing anything, the throttle only applies to actual liquid engines. Project Orion was obviously neither solid nor liquid fuel, but (oddly enough) seems to behave as a combination of the two. You control how many pulses there are, and you control when they happen, but you don't control how powerful they are. Theoretically in-game, you could scale the bombs' yields to the throttle, but Orion is about riding the shockwave of a nuclear blast, not about finely regulating your orbit.

When the USAF Orion was designed, variable-yield nuclear devices were not common. Instead they would throttle the thrust by altering the bombs-per-second detonation rate. In the part.cfg file I have it set up to do 1 bomb per second, so the kiloNewtons per bomb is equal to the ship's thrust. If it is 2 bombs per second then thrust is x2 kN/bomb, if 1 bomb every 2 seconds it is x1/2 kN/bomb.

In practice though, people in this thread and other places tend to mount auxiliary propulsion systems for fine control. Like atomic engines.

[nyrath]

I can't get more than one pulse. What am I doing wrong?

When you hit the space-bar, it automatically sets off one nuke.

As Canopus stated, as long as you press and hold down the Z key, nukes will be detonated at one second intervals.

[mushroomman]

I think it would be more efficient to use secondary engines rather than variable yield nukes anyway. Even if you could have nukes that only change the velocity by 1m/s, it would weigh more than it's worth. So I'm content with only one power.

[Canopus]

Yikes! That sounds like a serious problem. And it seems to be linked to having a lander.

Would it be possible to get a copy of your craft file, to see if I can reproduce the problem?

Canopus, have you tried using a lander with the Orion yet?

I have not tried it yet but i'm going to now.

[nyrath]

I'm making a new engine model, with the frame on top, and individual bomb magazines.

[Canopus]

Looks interesting. Let me know when you think its ready:)

[Leonov]

When i first came across the idea of Nuclear Pulse Propulsion, I thought it would be perfect for Kerbal space program.

While using this for my latest endevor to the Kosmos i saw just how "Kerbal" this is. It opens up so many doors for posibilities on future missions.

I have nothing but praise for your endeavour into the making of this mod and wish you luck in its fufillment.

A quick question, Why is the fireing of the Projectile depended/bound on/to the Z Key?

Could there be a way to Adjust the timing at which the craft launches the projectile by using the throttle function?

[Hremsfeld]

Thanks for the images! And you are not being preachy, I need this sort of input. My trouble is that while I do know a lot about Orion drives, I am a novice at KSP. I was unsure about how the existing staging system works.

No problem! And okay, cool. I'm not an expert on the internal workings, either, but I know that parts can have multiple functions. There's a working Dragon capsule someone around here, made by CardBoardBoxProcessor, that has about 7 different things in one part; RCS fuel, engines, RCS thrusters, lights, ladders, landing gear, and of course a command capsule that kerbals can enter/exit. So, the thought I had was to add the decoupler module to the magazines, so when they're done they can just eject themselves, probably at an oblique angle so columns and radial attachment both would work.

Ummmm, I'm confused. How is the drum magazine different from an externally or stack mounted magazine? The internal details of the magazine do not matter. The only important thing is how many bombs are left inside. What I was fretting about is how the user can figure out which magazines are empty, and how to decouple just them. The baseline USAF Orion design has 60 magazines, in six stacks of ten high.

The main thing is to give the user the ability to adjust the number of magazines to match the mission. Those things are massive. And there are different sizes of nuclear bombs, the more massive, the more kiloNewtons they put out. I could easily make magazines for each bomb size. All you have to do is clone the part, rename it, and do a couple of edits in the part.cfg file.

You can do none of this if you are stuck with a huge monolithic magazine array, as I have currently.

Yeah, that was on me for completely missing the point of the magazines. I'd thought each magazine on the array was cylindrical, because I hadn't zoomed in/panned the camera to show the sides of the magazines facing the center. Having the nukes as a resource stored in magazines would mean you could just right-click a magazine to see how many nukes are left in it. If you don't want them to be able to be transferred like fuel/oxidizer etc. currently can, put transfer = NONE in the resource definition, rather than transfer = PUMP. Perhaps a resource for each yield? Or would the engine not know how to interpret that?

When I was making the package, I downloaded the free version of Unity 3D to make the models. I later found out that it comes with the Monodevelop IDE for C#, altered so it works with Unity. Apparently it has the C# compiler as well, at least I did not have to download anything else in order to compile. I used the instructions here to set it up

http://wiki.kerbalspaceprogram.com/wiki/Plugins

Thanks for the link, I'll look into that and see what I can come up with.

That is excellent!

Currently, my magazine array includes a central core and some framework stiff in between the magazine stacks. That could be moved to the engine model, as long as the collider had holes allowing the stacking of magazines.

I've actually already got a magazine model, I just stretched it into the size of a 10 stack for the current model (to keep the polygon count low).

When you say "make a stack base", you mean placing attachment points on the top of the engine part, not creating an entirely new part, correct?

It would be nice if I could incorporate a decoupler into the magazine, to keep the part count down. And incorporate automatic fuel lines, also to keep the part count down. I was toying with the idea of making the engine magically find all the atomic magazines in the vessel, automatically suck bombs from the most empty magazine until it was dry, and automatically somehow ejecting the spent magazine. Probably a bad idea, but I wasn't sure if KSP could handle 60 decouplers with individual controls.

As you've already got the model as part of a larger model, you could open up the magazine .blend file, do a save-as, and then delete stuff that isn't the central stack. It...doesn't actually bode well for structural stability, to have the spine be made of segments one magazine tall, but it allows for easily changing the ship's size if someone wanted a smaller or (God help us) larger amount of nukes. Would the word be payload, or is that reserved for what you're bringing along? Huzzah for words with multiple meanings :V

Sucking bombs from the most empty magazine...well, it has to figure out which magazine to draw the first blast from. Would something like the way liquid fuel/oxizider flows now be close to what you're looking for? Obviously there's need to be something else, because nukes are a discrete unit, unlike liters/whatever. So, it would have to draw one from one side, and then one from the opposite, etc whenever launchBomb() is called. People might want to hang onto their empty magazines so they don't drop spacejunk, though...but I guess that's not as big an issue when you're blasting nukes a few meters away and they're no longer part of the ship.

When the USAF Orion was designed, variable-yield nuclear devices were not common. Instead they would throttle the thrust by altering the bombs-per-second detonation rate. In the part.cfg file I have it set up to do 1 bomb per second, so the kiloNewtons per bomb is equal to the ship's thrust. If it is 2 bombs per second then thrust is x2 kN/bomb, if 1 bomb every 2 seconds it is x1/2 kN/bomb.

In practice though, people in this thread and other places tend to mount auxiliary propulsion systems for fine control. Like atomic engines.

Got ya. Yeah, I'll probably end up doing something similar. It'd makes landings a bit less dramatic, but hey.

Ooh! On that note, I attempted re-entry around Kerbin, and the entire rocket flipped forwards from a pretty high altitude. This is good for keeping it stable on ascent, but bad for landing on worlds with atmospheres, so the drag values might need to be looked at.

Edit:

I'm making a new engine model, with the frame on top, and individual bomb magazines.

Beautiful! Great work so far, I'm definitely looking forward to seeing how this turns out.

Edited May 25, 2013 by Hremsfeld

[Canopus]

When i first came across the idea of Nuclear Pulse Propulsion, I thought it would be perfect for Kerbal space program.

While using this for my latest endevor to the Kosmos i saw just how "Kerbal" this is. It opens up so many doors for posibilities on future missions.

I have nothing but praise for your endeavour into the making of this mod and wish you luck in its fufillment.

A quick question, Why is the fireing of the Projectile depended/bound on/to the Z Key?

Could there be a way to Adjust the timing at which the craft launches the projectile by using the throttle function?

nyrath is already working on using the throttle to control the bombs i think.

[nyrath]

When i first came across the idea of Nuclear Pulse Propulsion, I thought it would be perfect for Kerbal space program.

While using this for my latest endevor to the Kosmos i saw just how "Kerbal" this is. It opens up so many doors for posibilities on future missions.

I have nothing but praise for your endeavour into the making of this mod and wish you luck in its fufillment.

A quick question, Why is the fireing of the Projectile depended/bound on/to the Z Key?

Could there be a way to Adjust the timing at which the craft launches the projectile by using the throttle function?

Yes, that is on my list of things to do, but it is not top priority right now. It will have to be done because the Z key does not work with time warp. Though that is probably not much of a problem, blasted ship accelerates so fast.

[DasMike]

The new model is looking really good! I actually did a playtest with just the Orion and found it completely adequate as an SSTO platform, only problem, as several people have mentioned thus far, was that it wasn't very good at precise orbital maneuvers. So I slapped on a few Fatman NERVA engines, kind of like what mushroomman did, and it DID seem to do the trick. Of course, I'm eagerly awaiting a version that works with the throttle controls, but for now this seems like an effective workaround solution.

Also, I thought I'd include some artistic references for anyone with the skill (not me, sadly) and the interest to come up with different Orion-type designs.

http://www.projectrho.com/public_html/rocket/artgallery.php#WilliamBlack

http://www.projectrho.com/public_html/rocket/artgallery.php#id--Rhys_Taylor

[nyrath]

So, the thought I had was to add the decoupler module to the magazines, so when they're done they can just eject themselves, probably at an oblique angle so columns and radial attachment both would work.

*blink* *blink*

Enable both column and radial attachment? I did not think that was possible.

And what happens if you have a stack of magazines, then decide to eject the one on the bottom? I'd think that if they were using column attachment, the entire stack would be ejected. And if it was using radial attachment, the unsupported magazines would hang there until the acceleration stress tore them away. But I do not know what would actually happen.

The method used by the original USAF Orion was complicated and I am not going to try and duplicate it. The feed intakes on the engine ate bombs from the magazine at the bottom. When the bottom magazine was empty, it was ejected, and the entire stack would move down on rails until the new bottom-most magazine was seated in place.

I did have the bomb import ports and magazine ejection pistons modeled in the 3D model, you can see them at the bottom of the magazine array. I removed them for this new version because I was afraid they would be encased in the collider for the engine, and prevent the magazines from being attached.

I'm not sure I can visualize the oblique ejection angle that will accommodate both column and radial attachment. Can you make a quick sketch?

As you've already got the model as part of a larger model, you could open up the magazine .blend file, do a save-as, and then delete stuff that isn't the central stack. It...doesn't actually bode well for structural stability, to have the spine be made of segments one magazine tall, but it allows for easily changing the ship's size if someone wanted a smaller or (God help us) larger amount of nukes. Would the word be payload, or is that reserved for what you're bringing along? Huzzah for words with multiple meanings :V

Actually not increasing the payload, instead increasing the propellant fraction (i.e., the mass ratio).

I had thought of doing the spine in sections, but that both compromises the structural stability and increases the part count (which has a direct impact on the frames per second). The current spine allows for 60 magazine or 3,600 nukes. Canopus managed to go from Kerbal orbit to Duna with only 250 nukes. I'm not saying you'll never need more, but you have to make trade-offs.

Sucking bombs from the most empty magazine...well, it has to figure out which magazine to draw the first blast from. Would something like the way liquid fuel/oxizider flows now be close to what you're looking for? Obviously there's need to be something else, because nukes are a discrete unit, unlike liters/whatever. So, it would have to draw one from one side, and then one from the opposite, etc whenever launchBomb() is called.

ohboy, I didn't even think about that. Bad idea to keep sucking from one magazine until it is empty.

One magazine has a mass of 4.92 tons (4.74 tons for the 60 bombs, 0.18 for the empty magazine). Using up one magazine then ejecting it will move the ship's center of gravity by almost 5 tons. That's got to hurt the ship's stability.

Will have to try and absorb equally from the top 6 magazines.

Ooh! On that note, I attempted re-entry around Kerbin, and the entire rocket flipped forwards from a pretty high altitude. This is good for keeping it stable on ascent, but bad for landing on worlds with atmospheres, so the drag values might need to be looked at.

Both the engine and the magazine array have

dragModelType = default

maximum_drag = 0.2

minimum_drag = 0.2

angularDrag = 2

because those were the values in the part.cfg file I cribbed from, and I didn't know what reasonable values were. Suggestions?

[nyrath]

The new model is looking really good! I actually did a playtest with just the Orion and found it completely adequate as an SSTO platform, only problem, as several people have mentioned thus far, was that it wasn't very good at precise orbital maneuvers. So I slapped on a few Fatman NERVA engines, kind of like what mushroomman did, and it DID seem to do the trick. Of course, I'm eagerly awaiting a version that works with the throttle controls, but for now this seems like an effective workaround solution.

Also, I thought I'd include some artistic references for anyone with the skill (not me, sadly) and the interest to come up with different Orion-type designs.

http://www.projectrho.com/public_html/rocket/artgallery.php#WilliamBlack

http://www.projectrho.com/public_html/rocket/artgallery.php#id--Rhys_Taylor

I'm glad it works for you!

And of course I know about those pictures since that is my website.

I'm hoping that I can get the plugin to the point where other people can use it as is. All the customizable variables are exposed in the part.cfg file. The main thing will be for somebody to make an Orion engine with Blender or something, and note how far the pusher plate moves.

Which reminds me that I should make the bomb supply mechanism so it can be easily adapted to other Orion designs.