# ProjectLarge Bussard Ramjet Project

#### Drake

Hi all,

I am getting back to Orbiter after a while away. I previously did a couple of little projects - the Stanford Torus and the Rolling Stone - just using Vinka's spacecraft.dll.

What I loved to do was to fly Vespucci (a very large ramjet) from LEO, where it was constructed, up to GEO where I had stashed the Torus. Fly 2 DGIII (from Earth) and 2 Shuttle A (from the Moon) to rendezvous with Vespucci, transfer "personnel" from Stanford over to Vespucci, and then begin a Grand Tour. At each planet, I would undock one of the DGs and explore a bit. If it was possible, I would locate a nice place to "build" a base with the DG and land a loaded Shuttle A with "base parts." If I was successful, I would "build" a base by adding it to the scenario code, load everything back up, rendezvous with Vespucci and head off to the next planet, moon, or asteroid.

Well, with each Orbiter update Vespucci got more and more buggy so I never got to finish my Grand Tour.

I recently came across the "Windmill" project, which is also a Bussard ramjet type. They are doing some really great work there on simulating the ramjet. I post over there with my thoughts about this part of it.

I thought I would build a model of my vision of a really big ramjet like Vespucci. I am an engineer (alas not in aeronautical) so I want to make it as realistic as I can.

So here is my development plan.

1. Develop initial model - 90% done
1. Model in gmax
2. Texture in gmax
2. Export to mesh format
1. Export model to mesh using max2msh
2. Create dds' for textures
3. Create flyable spacecraft using spacecraft3.dll - Release 1.0
4. Modify code from Windmill project to apply to this one - Release 2.0
So here is what I have so far:

View attachment 2141

The little pink eraser thing is the size of a Shuttle-A parked on the landing pad. The rings are 200 m in diameter and counter-rotate once every 30 seconds for 1g pseudo gravity. At the moment it is 1km long, though I might trim it.

I am still messing with concepts for the powerplant and the exhaust bell, and I need to add a sphere in front of the docking area as the refinery and some big honkin' radiators. I'll also add a collector around the exhaust for when we are slowing down. I also need to add some ship docks and cargo/people locks (I currently have placeholders). Little details like connection points need to be added

I expect that the initial model will be done pretty soon.

So, other than to let you all know that I am working on this, I am posting to see if anyone would like to contribute comments or work on this project.

What kind of help could I use?

• I am adequate in modeling - I don't plan on making a lot of complex doohickies that would look good but increase my work and decrease the framerate. However, I might have gmax questions...
• I am passable at texturing, but I am VERY open to help here in both generating the texture as well as doing the converstion to dds in a smart manner (e.g. square or not, pixel size, save options)
• I know how to program, but am not familiar with C++, so I will have questions about adapting the Windmill code for this project and would love to have help on this, or even have someone volunteer to do it when we get there!
• I need suggestions - what looks like it will work, what doesn't, what can look better, etc.
Current things I am thinking about:

• All the support struts are currently plain boxes. Should these be:
• Plain boxes
• textured to look like trusses
• texured to look like cables
• modeled to be trusses at the expense of lots more polygons
• Presumably the power plant is a tokamak - what should the power plant module look like in the model?
• How much will this mass?
• I'd like to be able to close or furl the collectors - how would that be best accomplished?
• Where can I get a model of a nice-looking lock and dock that I can bring into gmax? (I don't recall that max2msh goes the other way...)
• Should the axis support be a truss with tubes shown inside (for people, etc.) around a glowing column of plasma on its way to the reactor? Can the brightness of the glow be linked to the scoop rate?
• Where can I get a model of a communication dish into gmax, or shall I build one myself? Should this be animated?
• What shape would a fusion nozzle take? Not necessarily an expansion bell since it would be magnetically contaned I think
So let me know if you have input on any of that, or if you would like to lend your talent to the model. I encourage you to go over to the Windmill project page if you have input on modeling the Bussard ramjet code.

Thanks all!

#### Eagle

##### The Amazing Flying Tuna Can
• All the support struts are currently plain boxes. Should these be:
• Plain boxes
• textured to look like trusses
• texured to look like cables
• modeled to be trusses at the expense of lots more polygons
• Presumably the power plant is a tokamak - what should the power plant module look like in the model?
• I'd like to be able to close or furl the collectors - how would that be best accomplished?
• Where can I get a model of a communication dish into gmax, or shall I build one myself? Should this be animated?
I would suggest having a minimum of polygons on the truss and using textures for detail. Trusses for compression points, cables on tension. You could use trusses for tension as well, but cables do not provide any compression support.

Draw the power plant as you like. The tokamack is donut-shaped, but there is other necessary equipment. If you are lazy, just a sheet-metal shell over a volume will do.

You could use simple scaling for the collectors or divide it into parts and animate. Visualize how you want it to look and move first, then work out the animation of how to do it.

Satellite dishes and other aerials are extremely easy to model. Give a parabolic curve some thickness and put a receiver at its focal point. Add some support struts and an electronic box and wiring conduits and you're done.

#### Drake

Hi Eagle - thanks for the comments.

I am with you on minimizing polygon count, but I seem to remember from my first aborted attempt that if I texture a box with a truss it only shows one side, e.g. you look at the truss and only see the texture on the side closest to you, which would look weird. Do I remember correctly?

Also, if I build a truss with transparency and put something inside of it (like a glowing cylinder representing the plasma) will that show or not? I seem to remember it not showing - it is not "really" transparent IIRC, it is just not drawing anything where transparency is indicated.

I got an "A" in statics, so I know about cables in compression. (To paraphrase George Burns, like shooting pool with string.) What I was really asking and poorly phrasing was on tension members would cabling be used vs. a truss. I am attaching the ring supports (the only thing in tension) to a separate bearing, so the ability to move a little bit might preferred when starting up the rings, or not maybe the second bearing is undriven and just dragged around by the truss so maybe the stiffness of a truss is preferred. I am also not sure how cables as a member perform in vacuum - I know I wouldn't want to use steel cables. Aluminum would be OK but springy (low modulus). Maybe titanium...

Currently the power plant is in fact a big black box but I thought I'd put something interesting on it. Maybe parts of the tokamak stick out of the sides.

I think I'll have to animate, though it may be a combo animate and scale. That portion of the collector is material and the magnetic ring needs to be stowed somewhere. Hmm...maybe add a truss to the mag torus, deploy that via animation from its stowed position then scale the material part of the scoop like it is mylar or something being drawn up. That might be fun to watch...

I made a dish for the Stanford torus' heat plant but could never figure out how to get it to draw a parabola, so it is a section from a circle. Inelegant. If you know how to get gmax to draw a parabola, let me know!

Thanks very much for the comments!

#### Eagle

##### The Amazing Flying Tuna Can
You could put a slightly smaller truss with reversed normals inside the truss if you want it to be visible from both sides. If you modify the smooth angle of an object you can make even a triangular prism look like a cylinder (except that its diameter changes when you rotate it).

With transparencies you need to be wary of the layering. I'm not sure exactly what order it should be for the inside object to show up. I think the inside shape needs to be listed in the mesh file before the outside shape. In Anim8tor it ends up that the newest objects are put first in the mesh file. This tends to work out as I usually do the outside parts first.

If you want to fake a parabola, take a spherical section and squish the pointed end in more so it looks like one. You're not actually bouncing radio waves off it so it doesn't have to be perfect.

EDIT: I would suggest using meshwork for large and important trusses because if your resolution isn't good enough the truss will look blurry. But for small trusses, especially if they're repeated, you're golden with textures.

#### Drake

Oooo, great - I never knew about the order in the mesh. Geez, I wonder if gmax controls layering somehow...it has a hierarchy function...

Hmm, if anyone knows what order max2mesh gets the stuff in and how that is controled that would be great information.

I also really like the triangular truss texture idea - it would be more stable than a square truss anyway. On those small guys no one will notice if the other side disappears I bet. I think I'll make them trusses rather than cables. I don't know enough about cables in space and I know a truss would work.

Thanks again Eagle!

#### Drake

Current State

Here is the current state of the model -still more to do. I changed the center core to a polygon truss - I think it looks really nice - you can see the conduits and fusion fuel. Did a couple of other minor things - spent most of my time messing around with textures, which is frustrating since I know I don't remember the things to do and not to do in gmax so that the textures come out right.

#### Imagineer

##### New member
Adapting the Windmill code will not be too hard.

Techspecs.hpp contains nearly all the specifications. You will need to change a few numbers, well, more than a few... You will also need to update the names of the mesh and texture files. Finally, you will need to change the mesh group number in the animation routine and then clone it for the other centrifuge.

Have fun!

#### Drake

Thanks Imagineer! I downloaded the Microsoft tools the other day. Once I get this thing released to spacecraft.dll, I will delve into that! I can edit programs "real good!" A legacy of the old days of BASIC and FORTRAN and Pascal - I can program, but I don't know how to do anything in modern languages, so best I can hope for is to see how it is done and then modify/mimic. I am sure to have questions...

Getting close to having a usable model. Last night I exported the mesh to Orbiter for the first time - what a thrill! It actually worked first time!

Current questions:

gmax:

• The first time I exported the Stanford torus when I was dong that, I ran into the problem where the exported mesh is lots bigger than the scale I built in gmax, and the same thing has happened again. (It is so big is throws a shadow across the entire Moon!) Does anyone know how to fix that? As I recall, I used one of the utilities to just scale everything down, but I also seem to recall there is a setting somewhere in gmax that fixes this???
• Does anyone recall which direction in gmax is "forward" in Orbiter? I seem to remember that it is not what you would expect (e.g. forward). No big deal here - I'll just need to rotate it to where I need it.
• Is there a way in gmax to use one shape to "cut out" a piece of another shape, or am I stuck editing mesh nodes?
• I found out that max2mesh DOES import, so I can import in a dock from something - yay! I'll see if I can get permission to use one that comes in the core package so as not to add any more textures.

Textures

• I am having trouble getting transparency in some textures. One texture did just fine - the emissive material shows through the alpha channel holes and looks cool I think, but the truss I made is showing the alpha channel as black. Any idea what I am doing wrong?
• What is the best format to store the textures in? I am storing them as DX1 with alpha channel.
Orbiter:

• IIRC, we didn't have blinking lights in Orbiter, so people used zero-thrust tiny exhaust - is that still the case?
Thanks for any help!

#### Ursus

##### Rocket Tinker
ShipEdit in the Orbitersdk utils is handy for scaling, rotating, translating (meshes) and performing rough calculations of volume and moments of inertia, though Gmax ought to be able to handle the scaling and rotating. Another handy tool is MeshWizard.

The current version of the Orbiter SDK has light beacon management incorporated into it. (Look at API_Reference.pdf, starting on p. 118.) Actually... looking at the docs, it doesn't look like it's as much work as I though it would be... Hmm... You might just get some code to use in a couple weeks.

#### Drake

Thanks Ursus! Great info! I used to know some of this stuff...they say memory goes second...

Hmm... You might just get some code to use in a couple weeks.

Code:
BEGIN
GetCrackingOnModel();
HappyDance();
END

---------- Post added at 07:23 PM ---------- Previous post was Yesterday at 11:53 PM ----------

Notes to myself and any one else it might help...

Ah-ha - I figured out transparency. In gmax you can add a texture that makes it transparent, but max2msh and Orbiter don't think of it that way. The transparency needs to be in the dds file. The program I use to generate dds doesn't seem to use the alpha channel from the texture file png, so all I had to do was make a B&W version of the alpha channel and import that into the dds.

Now I have textured and transparent trusses!

I never would have figured that out but for a posting by Frying Tiger that was only available from the Wayback Machine...did we lose all the old orbitersim forums? I am still trying to figure out how transparency works when used within the diffuse texture map where I want the material to show up through the texture. Sometimes it does what I want, sometimes it does nothing, and other times it does the exact opposite.

Now to get the transparent stuff to show up correctly (the initial pass resulted in a truss that if you looked through it, you could see through all other objects like magic x-rays!), they need to be last in the mesh. I think max2mesh just goes alphabetically by object name - if so, all I have to do is add a "Z" on the beginning of names of those I want to be transparent.

I also figured out the scaling thing. In gmax if you don't set the units to be meters or whatever from the start, you are actually working in "units" even though it says meters. The version of max2mesh I have (not the beta I found out about later) takes these units and multiplies them by 39.370079 in an effort to convert units it thinks are inches into meters. Thus my 1km ship became a 39.370079 kilometer one! I want it big, but not THAT big! When I changed the units setting in gmax, it multiplied everything by 39.370079 so I had to do a uniform scale back to what it should be. Oddly, the old parts read in inches, the new ones in meters. Moral: set the units before doing any modeling!

I added some glowing safety lights on the instrument boom and need to add some more. I need to look in the SDK Ursus mentions to prepare for real lights though.

The mesh was on backwards (the exhaust came out of the front, which was amusing) but that is now fixed. I probably need to make it clear which direction is "up" for purposes of controlling the craft, even though it is almost radially symmetical. Probably a nameplate on the sides. What shall I name it?

I need to add some attachment points to hide trusses sticking into other support members. Also need to put in some pretty docking points for ships, a pretty human-door, and a pretty cargo door. All are just place-holding primitives now. Maybe add some cargo boxes to the landing area.

I added a throat to the front scoop - when I animate the scoop it will emerge from this. I need to add one in the back. It will be animated too, and will have to fit around the nozzle in an annulus I think.

Still need to make the engine room look good - just a box now. Maybe an ovoid with a double-decker tokamak. And I still need to figure out what the exhaust nozzle will look like. Since it will not be physically containing expanding gas, it won't need the special shape chemical rockets need in an expansion chamber. Looking at a VASIMIR nozzle for comparison, it is just conical and there to provide the magnetic containment. I think since it plasma, it will carry the magnetic lines out for a ways, so I think the exhaust will stay collimated for a long ways out. Or it will expand rapidly after the nozzle - I dunno. I guess it will depend on which looks better.

Need to add RCS nozzles and figure out weights, PMI, ISP, thrust, etc.

I'll post a pic soon.

#### Eagle

##### The Amazing Flying Tuna Can
Glad to hear its coming together.

I think you should name it yourself. After all Argus got to name the Argo. In light of self promotion Drake seems to work. But name it what you like; if you liked Philip P. Pauxtinaltygundermelk III, then go for it. We'll probably complain anyways.

#### Drake

Hmm, I just watched the episode of Cosmos with my girls where he talks about the ramjet.

Is Sagan already taken? If there is any justice in the universe, a Bussard ramjet would be named after him.

The Sagan-class interstellar ramjet testbed...

#### Drake

Current state of the model in Orbiter...

The grey box is the size of a Shuttle-A landed and docked.

Still can't figure how to put black "tiger stripes" on the glowing yellow safety area around the landing pad.

I am not convinced the engine pod (the black ovoid) is the way to go. If anyone has a better idea, let me know!

I added mysterious glowy magnetic tori.

A mistake in texturing led to a cool effect on the back of the scoop, so maybe I'll keep that.

Getting close to finishing the first model I think. Then a bit to get things animated via spacecraft.dll and ready for a beta release I think. So if you have a brilliant idea about the model, now is the time for it!

#### Eagle

##### The Amazing Flying Tuna Can
Your ship looks very nice. I have a small mesh suggestion. Your exhaust nozzle intersects those glowing tori in an unnatural way. Like its placement might block flow through the tori. Maybe curve a plasma pipe off or set it back a little further.

#### Drake

Your ship looks very nice. I have a small mesh suggestion. Your exhaust nozzle intersects those glowing tori in an unnatural way. Like its placement might block flow through the tori. Maybe curve a plasma pipe off or set it back a little further.

Moments ago I stuck a pipe there to set it off and moved the tori up...but your suggestion is cooler. How about if I pull a cylinder off the tori at a tangent then swoop them into the nozzle. I am pretty sure gmax can do that...

Thanks!

#### Drake

OK here it is with my interpretation of Eagle's idea.

It took forever to get those tubes to line up right buuut, I think the principle is flawed. As long as the plasma is going around the tokamaks, the net force is zero. (The new diversion tube is a mirror image on the other side going from bottom to middle to balance out that force.) But once I take the plasma out on the side like this, I need a magnetic force to keep it from hitting and burning through the tube, and that force is going to slow me down. We might get it back from the elbow right before the nozzle...not sure. I don't like how the nozzle is so far away from the engine room either. *sigh*

I think I need a tokamak at the centerline - I'll have to think about it more, or maybe two together like an infinity sign? Hmm...

Oh, and I christened it "UNSS Carl Sagan" (apparently it is a UN project).

#### Drake

Hmm, I just thought of this - maybe two nozzles - one coming out of the side of each tokamak? Maybe incline the tokamaks so that the tangent point is at the midline? Maybe....

#### Drake

Texture Size?

What is the best practice for textures? Do they need to be square power-of-two in size

Project Shuttle-A Mk.IX
Replies
72
Views
4K
Replies
180
Views
6K
Replies
8
Views
649
Replies
17
Views
1K
Replies
39
Views
2K