Small cargo ship

[chuck_starchaser]

Thanks.. Uh.. How about a trade? A model for a song...

[zaydana]

hmm, i havn't actually got FL Studio installed at the moment because i'm using the family computer. I'm still waiting on the parts I ordered for my new PC... stupid online store. Anyhow, i should have them in the next couple weeks, and I'll see what I can do about some music then.

[chuck_starchaser]

No rush; I'll definitely want Klauss' arrangements on the bass line, and he's too busy for now... If you have any models in mind, let me know; I might pay in advance.
After I finish this model, though; I'm just about 50% of the way.

[spiritplumber]

2001tastic! What's the 3 coils for?

[chuck_starchaser]

Well, I fancied that perhaps there's a way of making a controllable, nuclear fuel cell, like an active device; so, instead of using positive and negative materials for the plates, using some kind of same-material 2 layers, plus two inner grids, or something of the sort, and depending on the polarity you apply to the controlling grids, the fuel cell produces a similar polarity on its power outputs. That way, we could divide each nuclear fuel cell internally into 3 pie wedges, and apply a 3-phase, 50 KHz (for efficiency) sine-wave signal to the 3 sets of grids, and get direct 3-phase power output. (Am I dreaming in technicolor?) But I'm assuming it would be ultra-low voltage, ultra-high current, so the 3 toroids are a step-up transformer. They are pretty massive, actually: The cores are 4 meters in diameter. That thick, U-shaped "pipe" in each of them is the 1/2-turn primary; --solid copper... er... make that aluminium, 2 feet diameter.. What's the gauge number for 2 feet dia.?

[chuck_starchaser]

Update.

When you need to fix something on the engine back there on EVA, you take this ladder down from the lock. We're looking "up" towards the front of the ship; --remember that "down" is towards the back as per the 0.3 G (max) gravity created by the forward acceleration. Hope you're not afraid of heights...



Half-way down the ladder there are like scaffolding walkways that lead to the maintenance check-points around the ship's mid-section.



If you take the left walkway to the end, turn around and look back, this would be the view, approximately...



Here we see the mid section in more detail in this rendering. Looks more complex than it is due to shadows. The maintenance scaffoldings lead to the controllers and lube-points on the 2 centrifugal and 1 wankel pumps, and on the liquid/gas separator:



The front of the ship is up in the pic; back end is down; to make it more intuitive to look at, gravity-wise.

Haven't done any of the electrical stuff I said I'd do...

[Ryder P. Moses]

Jesus, dude, that's awesome.

[chuck_starchaser]

Thanks. I've no idea how one would go about unwrapping so much stuff; but I'm thinking maybe the pipes and scaffolding could be on separate materials that are flat-colored, as opposed to textured; --maybe made specular so as to not stand-out for their lack of shadow sensitivity...

[klauss]

As said, you want the spikes with a bit more space between them.

The idea follows:

Spikes are very little massive. So... no matter the material's specific heat, it becomes very hot very quick. Since radiation is proportional to T^4, that increases radiation a great many times, increasing thus cooling efficiency. The trick, I imagine, is avoiding the spikes themselves to melt under heavy stress - I imagine modern systems would find a way to overcome that issue, either by active compensation (pumping less heat onto them, and routing somewhere else), or by simply using very resistent materials. In any case, they could be pushed to incandescence (which would make an interesting sight).

They would also suffer little from inter-spike heating, because they also present a very thin profile: if properly spaced, from one spike's POV, 90% of the "visible environment" would be open space, with only the remaining 10% being other spikes (thus, inter-spike heating is cut down to 10%). Add highly reflective materials as per Chuck's idea, and you may cut it further down.

And I never saw them anywhere... it just ocurred to me. Funny.

PS: I won't even try to guess the polycount - I guess, though, many of the stairs and thin structures could be approximated by 1-bit alpha channels. Even the spikes. But that can be done afterwards, I guess.

[chuck_starchaser]

As said, you want the spikes with a bit more space between them.

The idea follows:

Spikes are very little massive. So... no matter the material's specific heat, it becomes very hot very quick. Since radiation is proportional to T^4, that increases radiation a great many times, increasing thus cooling efficiency. The trick, I imagine, is avoiding the spikes themselves to melt under heavy stress - I imagine modern systems would find a way to overcome that issue, either by active compensation (pumping less heat onto them, and routing somewhere else), or by simply using very resistent materials. In any case, they could be pushed to incandescence (which would make an interesting sight).

Right, I'm not used to taking per-weight efficiency in specifying heatsinks, but I guess in space applications this is a priority. And for a spaceship capable of acceleration, the very flat and long heatsinks Ryder was mentioning, and that you see on the ISS would present a structural support problem.

And I never saw them anywhere... it just ocurred to me. Funny.

Good thinking it is. I'll call it the "Klauss heatsink type" in-game, in the ship's owner's manual.

PS: I won't even try to guess the polycount - I guess, though, many of the stairs and thin structures could be approximated by 1-bit alpha channels. Even the spikes. But that can be done afterwards, I guess.

No need to guess: Facet count is at 37506. Vertex count at 43262.
Like I said from the start, it will probably be 100 k tris when done.
I'd say forget about the alpha channel idea. Why? Because the texture for this ship is already going to be pretty big *without texturing the greebles*. This is what I'm thinking: I'm thinking of making all the scaffoldings, pipes, ladders, hex nuts and electrical ducts into a separate object, with its own material, and use per-vertex color instead of textures for them. The only problem is they can't receive shadows or have radiosity baking, that way; BUT, if we make them all look "chromed", specular, then we remove the justification for shadows of any kind on them; --although they would *cast* shadows, of course.

EDIT:
Thinking about it again: That the spikes get hotter is not really an argument for efficiency, as, if the heatsink IS hotter that's a measure of its failure to dissipate... Unless you have a pelletier (probably mispelled) device under each spike...

EDIT2:
I'm thinking of calling this ship the "Tadpole".

EDIT3:
Klauss, what do you think of this idea?:
Having an alternative set of shaders that use a four-section glow map. The glowmap doesn't need as much rez as the other components, so we could halve the rez and give it four quadrants. Two of them could be alternative radiosity bakings that the shader can interpolate between; and the other two could be, again, two alternative glows (lights) that the shader can interpolate between. The reason for separating radiosity from lights being that the lights don't change with the intensity and color of ambient light, wheras radiosity does. The reason for interpolating radiosity being that sections on my ship, like the rads, get hotter in proportion to engine power setting; and indoors, you might want to have two radiosities for two sets of light sources that you can mix as per their being on, off or dimmed.

[klauss]

EDIT:
Thinking about it again: That the spikes get hotter is not really an argument for efficiency, as, if the heatsink IS hotter that's a measure of its failure to dissipate... Unless you have a pelletier (probably mispelled) device under each spike...

Not really.
CPU heat sinks are designed to stabilize at around 70ºC.
That's not that they're failing at their job - CPUs needs higher temperatures... if you cool down a processor too much you also take it out of its operating temperature.

So... think of these ones as... say... designed to stabilize around 300/400ºC.

Those would dissipate heat 1066 times faster!

Of course... it's not magical - they already took little heat to get that hot, so they would dissipate small amounts of heat faster. I'd consider them a mild boost over traditional heatsinks. I don't know... simulations would be in order to verify this - and it's not impossible... remember that objview program I sent you along with the radar code? That did thermal simulations of a sort. I would only have to change the formula from atmospheric convection/conduction-based cooling to radiation-based cooling... not a big deal.

About glowmap stuff: not impossible... there will be ways to attach physical states of the unit to shader parameters, as velocity, acceleration and - in your case - reactor output (that one I have to add... perhaps add a generalized interface... hm...).

[chuck_starchaser]

Uhmm... No; Klauss, you're wrong, for a change; CPU's work better the colder they are. There are CPU cooling kits that you can buy (if you can afford them) that push the CPU down to -40 Celsius. You can then just about double the clock speed on them, at those temperatures, if your motherboard and bios have the multipliers

True, the heat dissipation of a heatsink rises dramatically with temperature; but if heatflow per heatsink size was the parameter we were trying to optimize, we'd then make them the size of pinheads and let them go to millions of degrees. Unfortunately, we more often need to get down to a given temperature, so letting the heatsink get hotter to increase heat dissipation per surface area is not the right solution. But pushing heat towards the heatsink by peltier or other methods is.

R.e.: Reactor output, the interface must be there already, as the exhausts get longer or shorter, I've noticed. Not a "shader" interface, tho, of course...

[Ryder P. Moses]

That's not because microchips are faster or more efficient when they're colder, it's because they give off a lot more heat when they're overclocked; more than standard heatsink kits can handle, enough that even with coolant systems they'll usually degrade more rapidly than those left at settings they were rated for. Computers do work less well at extreme cold temperatures; this is supposedly a problem at polar research stations and such. However, it's not a relevant problem unless you live in Siberia or are using a liquid-nitrogen coolant system without a clue what you're doing.

[hurleybird]

Yeah, chips usually can run faster the better they are cooled, but klauss is right that the processor has an operating range... it's just a lot larger than he thinks. Cool a chip to much, and it may not run. Lots of A64's have a low temperature barrier... somewhere way down there... around -50 I think...

[hurleybird]

That's not because microchips are faster or more efficient when they're colder, it's because they give off a lot more heat when they're overclocked; more than standard heatsink kits can handle, enough that even with coolant systems they'll usually degrade more rapidly than those left at settings they were rated for. Computers do work less well at extreme cold temperatures; this is supposedly a problem at polar research stations and such. However, it's not a relevant problem unless you live in Siberia or are using a liquid-nitrogen coolant system without a clue what you're doing.

Eh... when you say that computers work less well in extreme cold conditions it's really the hard drive, not the chips, that suffer. In this case we are talking about cooling ONLY the CPU.

And I hate to say it, but you're wrong here. If you look at extreme overclocks, the CPU is almost always at a (very) negative temperature. If you were only to cool the CPU down to its original temperature, you would not be able to get nearly as high of an overclock.

[chuck_starchaser]

I'm still thinking about this spiky heatsink and I have a feeling there's something to it, tho I can't yet put my finger on it (ouch!).
In the meantime, I got bored with the mid section, so I'm working on the aft section ...



Another, smaller, liquid/gas separator. Gassified H2 will be sent to the 3 annular tubes, from which it will come out into ionizers and get injected on the sides of the exhaust bell. The liquid part will be sent to cool the last superconducting coil, at the plasma choke point. The 3 larger but thinner coils are for shaping and conditioning the plasma exhaust.

[Halleck]

I must say that watching an engineer model a ship from the ground up is quite different from watching an artist do the same- with functioning parts in the place of decorative "greebles". It's no eye candy to be sure, but it has a great sense of being completely utilitarian, undoubtedly built by someone with a "function-before-form" mindset.
And on a ship schematic, the individual parts could be labeled without resorting to technobabble.

It will be interesting to see this mod come together.

[chuck_starchaser]

Exactly; I frankly don't care about prettiness; I'm much more concerned with someone asking "why does this pipe, or wire, connect here?" and be able to answer the question. There will be larger ships you can buy, but more high tech and less hands-on. I think the player will develop an emotional attachment to this ship like to no other ship, after learning exactly how to maintain it, how often you need to go on EVA to service the pumps, and modding it with third party add-on's.

Bottom side there will be bay you can use as fighter bay. A fighter will probably be defined as a small gunship, chemically propelled, very limited fuel (2 or 3 minutes worth of accel, use it in bursts). No lasers or missiles. The former wouldn't write on armor, let alone damage it; the latter are too heavy to carry for a chemically propelled gunship. Instead, your cargo ship would have like an MLRS, and in your fighter you have the illumination electronics and remote launch button. The actual weapons on the fighter will probably be slug type munitions --i.e. a machine gun. Put a few slugs up the tail pipe of a ship and it will have to deploy a solar sail to get home. And possibly a rail gun with a real heavy slug; --for a single, killer shot.
But yeah, fighters' forms will also follow function.

[Halleck]

Interesting... sounds like you're planning a combat model quite different from the 'turn, shoot, turn, launch missile, turn, shoot' style of dogfighting in VS.

I'm curious, have you given any thought to the breadth of this mod and the universe it contains? It would be frikkin' awesome to have real gravity and orbit plotting...

Also, what about stuff like ringworlds, or gas rings like in The Integral Trees? I've been contemplating about making a ringworld for VS... would that not be "hard" enough sci fi for your universe? (no pun intended- ringworld material must be quite hard indeed, though in a different sense.)

[chuck_starchaser]

Funny you mention Integral Trees; I bought the book a week ago or so at a used bookstore; but haven't started reading it yet. Ringworlds are not as problematic as Dyson Spheres, but do have their share of problems. Someone pointed out to Niven, at a lecture, that a ring would not be stable. Essentially, a planet can orbit on a circle, but if disrupted it can follow an ellipse; but each part of a ring doesn't have that freedom. That's why when he wrote the sequel, Ringworld Engineers, he added jets for active adjustment of orbital eccentricity. Ok in basic principle; out to lunch in the numbers. Basically ring worlds are about a billion times less absurd than Dyson spheres, but still pretty absurd, IMO.

Real gravity is a must. Orbital plotting will be the standard interface. No joystick on this ship. You'll have like graphs and tables to help you optimize a trip for fuel/time/cargo. Lots of work optimizing a trip; but that's what you'll be doing most of the time during a trip, anyways: planning the next one; --plus advanced trading, ship maintenance, and daily dogfight practise on the simulator. And you'd better practise, because pirates will be rare to find, but when you do, you'll find they know their job well. And this mod will be hardcore: If you die, the fact is stored to the disk and you cannot reload from a save; you'll have to start from the beginning with a new name.

[klauss]

Yeah, chips usually can run faster the better they are cooled, but klauss is right that the processor has an operating range... it's just a lot larger than he thinks. Cool a chip to much, and it may not run. Lots of A64's have a low temperature barrier... somewhere way down there... around -50 I think...

Hm... probably I misread the specs. It's quite easy to miss the sign

Anyway... chuck... the idea is to actively push heat towards the spikes, not to let temperature differential do the work. Pelltier devices are one way... AC-like pumps are another. Mostly because I was actually meaning this kind of thing for the big cone-ish heatsinks (being big... you can mount a small AC pump into each of the... say... twelve spikes on each heatsink plate).

I'll see if I can draw it...

Ok... try this: spiked-heatsink-schemmatic.gif

[chuck_starchaser]

Sure, actively pushing the heat is definitely the way to go. What I'm still not sure about is the spikes themselves. Ryder's argument put succinctly what I always thought about heatsinks in space: Maximize area as seen from any perspective. However, I found, to my surprise, that the nuclear fuel cells on the Cassini probe had fins partially facing one another. This is not really a counter-argument, as those fins do increase the area seen from any perspective. But then I saw diagrams of plasma thrusters with parallel fins.
But there are other non-linearities to consider. "Maximize profile area" is a simplification based on the assumption that the heatsink itself is superconductive of heat; which is a fairly good approximation when you're working in electronics, since most heatsinks are made of thick material. But in space you're trying to optimize profile area per amount of weight rather than per-size. This would seem to be an argument for spikes, but only if they are arranged in random or pseudorandom fashion, such that from any perspective I see spikes. In other words, you'd probably want spikes that block the sight through from any direction. The spikes I have there fail in this regard.
But then that denies the argument of low reabsorption, as each spike would probably see spikes all around.
The argument for heat reflectivity of the spikes probably doesn't hold water; since radiation efficiency and absorption go hand in hand. Best radiant color is black; but black is also the most heat absorbing color.

So, I'm still torn...

Perhaps we should be looking at nature. Plants try to maximize exposure to light per amount of weight, as weight is both a problem to hold, structurally, and consumes the plant's limited resources of cellulose, which it must make from glucose, which it must produce by photosynthesis. Their problem does not include heat conduction considerations, but then again, there are so many shapes of plants we can find species that for some reason or another are fleshy at the core and have gradual thinning out characteristics. I'm thinking of like yucca plants, the iron plant, and succulents. They don't have computers, but they are computers, in an evolutionary sense. perhaps our ideal heatsink would look like a spider plant or something...



Or like a yucca...



But then there's another problem: Plants don't try to be accomodating or neighborly to other plants; but we need heatsinks that coexist side by side; so, in a way, we do need to optimize for size, and not just weight. So maybe they should look like flowers and have black top-sides and white under-sides...

I wonder if there might not be a role for pyrolytic graphite, here. Fascinating material: It's flexible like rubber; but it's a chrystal; and it has anisotropic electrical, magnetic and thermal characteristics. On the xy plane it conducts heat better than copper; but perpendicular to the xy plane, in the z direction, it is a terrible thermal conductor. Same thing electrically. Magnetically it is the most antimagnetic material known in the z axis. Yeah, the chrystal structure is like hexagons of carbon in the xy plane, IIRC...

[Halleck]

...Ringworlds are not as problematic as Dyson Spheres, but do have their share of problems. Someone pointed out to Niven, at a lecture, that a ring would not be stable. Essentially, a planet can orbit on a circle, but if disrupted it can follow an ellipse; but each part of a ring doesn't have that freedom. That's why when he wrote the sequel, Ringworld Engineers, he added jets for active adjustment of orbital eccentricity. Ok in basic principle; out to lunch in the numbers. Basically ring worlds are about a billion times less absurd than Dyson spheres, but still pretty absurd, IMO.

On top of this, ringworlds require the existance of what I would think to be an impossibly strong material, "scrith". That's what I meant by being 'hard'.

Nontheless, I'm still interested in the possibility of creating a ringworld for vega strike. The easiest way to do it would be to make a "halo"-esque structure and punch it in as a space station. The real challenge would be to make one up to niven specifications, centered around the sun (and with "shade panels" to boot.)

Real gravity is a must. Orbital plotting will be the standard interface. No joystick on this ship. You'll have like graphs and tables to help you optimize a trip for fuel/time/cargo. Lots of work optimizing a trip; but that's what you'll be doing most of the time during a trip, anyways: planning the next one...

It will be cool to see this in action. Something to do while in midflight will be good.

Are you planning on having frontier-style time accel? Even with all the planning to be done, players are still liable to get bored in mid-flight.

--plus advanced trading, ship maintenance, and daily dogfight practise on the simulator.

Dogfighting still? What exactly is your combat model?
I've always imagined real space fighting to be at standoff range, using turrets and explosive ordinance (torps and missiles.)

And this mod will be hardcore: If you die, the fact is stored to the disk and you cannot reload from a save; you'll have to start from the beginning with a new name.

Unless somebody has the audacity to hack that info.
I've seen games (pakoon) that store such info in plain text, but also store a special hash/checksum in the file so that the game can tell if it's been modified.
If you really want to prevent players from restoring a game, you could implement something like that... and everyone will probably hate you for it.

[chuck_starchaser]

Exactly; no material could ever take the outward "weight" force of the world in it. The only solution I see to that is to have nanotube ropes every few yards, and compensatory weights that hang "up" (sunwards). Still, the main problem is stability. Someone loads one one watermelon too many on his truck and that section starts to buckle; centrifugal force increases with distance at constant angular velocity, so the whole area that buckled now weights more... But in the Vegastrike universe, everything is possible

Not sure if we will have time accel or a bed and alarm clock. I'm for the latter. CoffeeBot wants to have a drug you take that slows you down. I don't think it's gonna work; because you have a lot of things that need taking care of, to be on drugs. And if pirates show up...

There isn't a single combat model; depends. Pirates for example, don't want to destroy you; they want your ship, and everything in it; and if possible, working, and flyable. They just want *you* out in the vacuum. They'll try boarding you, but if you have a fighter you can fly out and shoot their helmets off, careful not to damage your bigger ship. First you have to get their ship, tho, or it can shoot your fighter.

But yeah, some combat would be long range missiles, and anti-missile missiles. You'll rarely see combat, tho; but you want to be prepared for when it happens. Specially if you know you'll be going into combat, you'd practise a lot on the simulator in advance.
Yeah, I'll do the checksums thing, and not even notify the player right away, but like a day after, so that hackers will have a hard time verifying whether a hack worked or not.

Here's a succulent heat radiator.



Each "leaf" is a slab of pyrolytic graphite with the top side rounded ellipsoidally. The point is that heat flows very easily horizontally, and each plane comes to a peripheral edge that grow in surface area linearly from the top down to the edge. The top side is then dusted with regular graphite to maximize radiation. The bottoms painted white.

[Average Earthman]

What's the old saying on Dyson Spheres? "if you can build a Dyson Sphere you don't need it"...