Project ICOVD Windmill Class Interplanetary Ramjet Development Thread

[Ursus]

Edit: I guess I could post a link to the original thread on the M6 Archive.

'Nuther Edit: Newest Dev Thread at http://orbiter-forum.com/showthread.php?p=529139

Since the official Orbiter forums seem to be dead, I figured it was time to start a development thread here for the ICOVD1 Interplanetary Ramjet. (Umm... ok... would there be any objections to me officially dubbing it the WindMill Class Interplanetary Ramjet, or just WindMill for short?)

[Edit: Decided not to capitalize second half of "Windmill" when I changed the thread title.]

Hmm... If I'm going to have a development thread here for a project on which I'm working... I wonder if I should ask for one of those cute little orange bars under my name...

(Most images I post here will be thumbnails, to save bandwidth for ImageShack and for readers with slow connections.)

Panel Colors

I have shifted the primary colors (RGB) in the graphs and readouts... well... away from primary colors, based on the results of viewing the panels through the color deficient vision filters in GIMP.

Hmmm... my latest screen-shots are all JPEG's, and JPEG compression distorts the colors. Here's a couple PNG's I took earlier.





Scoop Cross-Section

Right now (actually, I think I've changed this a bit since the last time I committed to the repository, and I certainly have since the above screen shots were taken), I've got the effective range of the magnetic field (when the vessel is 1AU from the sun) set at 96,100 km, or 1/4 the distance from the Earth to the moon. In collection mode, the scoop field forms an elliptical cone around the incoming solar wind vector, that is bound to within 45 degrees of the Z axis of the vessel, so if the vessel is pointed directly into the solar wind, the intake of the scoop has a diameter of about 136,000 km, and an area of about 1.451e+16 m², or 14.5 billion square kilometers.

(Anybody want to figure out how much energy it'd take to generate such a field? ...or maybe it's better if nobody does .)

When the vessel is further away from the sun, the field doesn't have to overcome as much interference from the sun's magnetic field, so it is effective for a longer range, so while the density of the particle stream is lower, the scoop can still collect the same amount of particles (hydrogen atoms) at any given speed relative to the particle stream.

Stabilization

I figure that the drag force from collection would be focused on the active scoop, which means any little deviation from being pointed straight into the wind would start the vessel spinning, so I figured the vessel would have to project a stabilizing field behind it. In one mode, the field will merely counteract any torque caused by collection drag. In the other mode, it will actually (very) gently steer the active scoop into the solar wind (or, perhaps eventually, to a certain offset from the solar wind vector).

This should supplement the collection navmode Imagineer is developing. If it doesn't, or if it actually interferes, I can rework it or tear it out.

I was actually able to go from Titania to the point shown on Orbit MFD (at 10,000x time compression, most of the time) without using RCS thrusters (after the initial line-up). Of course, I am a bit off course for Neptune, but... heck... there'll be plenty of time for course corrections during the braking phase.



Maximum Speed

One of the main objectives of my last run was to get some screen shots illustrating the maximum ram speed of the vessel. Between the adjustments in the field range and stopping to fill up the tanks (for an extra burst of speed once max ram was achieved), I wasn't able to get up to the maximum speed of a mere 1,010 km/s (that last bit of speed is hard to gain) relative to the particle stream (so, about 500 km/s inbound... and 1500 outbound... but you're going have to stop some time...)

I should've realized it earlier, without having to discover it through experimentation, but no matter how big or small the scoop is, or how powerful the engine is, the maximum ram speed is limited to the exhaust velocity.


Well, I've been yammering on, and it's getting late. I'll talk about the auto-throttles, sail mode and newbie C++ code (and whatever else I'm forgetting at the moment) some other time.

 

[trokkes]

online again

Hey,
I'm still around here :lol:.
I was thinking to open a thread about this, but you have been faster.

I'm working on texturing the docking ports. I'm playing with the size problems and trying to fit them.

And about the windmill. well :dry:
I like it. I'll make a shot with this name to see if it fits in its place.

 

[Imagineer]

(Umm... ok... would there be any objections to me officially dubbing it the WindMill Class Interplanetary Ramjet, or just WindMill for short?)

Clever. It uses the Solar Wind for propulsion and then theres the motion of the hab. ring...

I'm working on texturing the docking ports. I'm playing with the size problems and trying to fit them.

One thing that has frustrated me with most docking ports is there is no clear indication which way is "up" for visual approaches. Perhaps one of the cross hairs could have an arrow head?

 

[Ursus]

I've been knocked out of commission by a nasty flu bug for the last week but I'm hoping that I'll be able to pull myself together enough to get some work done over the next couple days.

After I get the RCS and Fuel sections of the Engineering panel done, I want to try and refine my auto-throttle algorithms a bit. More specifically, I want to refine the algorithm which determines the amount of fuel flow, as a fraction of the "unprotected reserve", that is available to the engine when it is approaching the protected level of the reserve.

The protected reserve is... to insure that we have enough fuel to maintain the main reactor for a couple months if the collection system fails or needs to be shut down for some reason? It seems like I had another reason that didn't involve system failure, but I can't think of one that makes sense, now. Oh, well... I'm pretty sure some amount of reserve is necessary... just how big of one is the question, and I plan on eventually making the protected reserve size adjustable.

... back to the auto-throttle... Right now, in order to keep the auto-throttle from jittering, especially at high time compression, I'm giving the unprotected reserve a high minimum "half-life" (well, sort of... the routine just aims to reach the protected reserve level at 900 seconds from any given time point), so the throttle tapers off slowly to a point where the reactor uses only as much fuel as is being supplied by the scoop. I might have a few ideas for lowering the drop-off time a bit, so the system doesn't seem so "mushy". Who knows, though... We might have to implement some sort of PID-type system to get it to really work right.

Still, even the cobbled code I've got now is a heck of a lot nicer than trying to manually adjust the throttle to maintain a "Pure Ramjet" profile.

 

[Linguofreak]

Ah, yes, good to see the project's still alive. Thanks for the link from my request thread Chupacabra/Ursus.

I would like to test fly it when it's ready...

 

[Chupacabra]

No problem

You can actually take a look at the source on Sourceforge if you wish, but it still has a little way more to go before the release of a beta.

On-Topic: Hey Ursus, is there something specific you might want me to model/texture? I find that I'm not very creative designing a feasible ship on my own.

 

[Woo482]

Does this have ummu ? It would be good to have it

 

[Ursus]

Does this have ummu ? It would be good to have it

Yup... In fact, I was just about to suggest that Chupacabra try his hand at an EVA station to attach to the truss. (There's a truss attachment template, "ICVAttProto", in the mesh directory of the repository.)

It'd need an airlock (with the outer door as a separate mesh group, so we can eventually animate it). It might also have a tool locker. (Mostly for looks, except for... hmm... is the turbopack included with UMMU, or is that only part of the DGIV package?)

Of course, I don't think that'd fill up an attachment block, but the extra space might later be filled with small parcels, and maybe eventually an optional work pod.

It might be a while before I get around to making it functional, but when I do get to that point, it'd save me the step of actually creating the mesh.

 

[Ursus]

Okey dokey...

Looks like I've got the fuel area of the engineering panel pretty much done.



It's amazing how I can get hung up over such a simple thing as the interface for changing the amount of fuel that gets set aside for maintaining the reactor (the "protected reserve"... "reserve reserve?").

Note that I haven't put the amount of time the protected reserve should maintain the reactor because there's not currently much in the way of reactor code, including the reactor maintenance cost. Eventually, that'll be shown in seconds. It will also be shown in days, hours, minutes and seconds (or as much of those that will reasonably fit in the alloted space).

I think that after I do a bit more work on the scoop system (mainly to cut down on the "scoopage", and the associated drag, when theres nowhere to put the collected hydrogen), I'll put together an alpha test package. I'll probably re-enable the Orbiter-native navmodes just for the test module, though I've actually managed to get along just fine without them during my own tests.

What do folks think of the exponential notation for the value displays? In a way, it makes more sense, due to the wide range of values involved. It's also a bit easier from a coding perspective, since the standard sprintf() function has a built-in option for such a thing. I'm just wondering if it might be a bit confusing for users who are aren't used to it. Then, back on the first hand again, it's an engineering panel; it should use engineering notation.

Regarding Source Code:

As I said when we started this project, I'm not really much of a programmer, so I'm doing a lot of learning as I go. I think I've found a system of assigning constants (at least for integral constants, which is good for the panel constants, such as area ID's, and positioning information) with which I can live (it involves enum's), after changing my mind a few times, so I'm going to go back and change a bunch of header code fairly soon, once I'm pretty darn sure I ain't gonna change my mind again.

There's also some other restructuring I'd really like to do to straighten out the spaghetti a bit. Some of it will probably happen gradually as I work around the parts that I think should be restructured. I'm thinking a major restructure might occur sometime around the first "public" release. In any case, the public methods of my classes shouldn't be affected.

 

[Urwumpe]

Is it possible to give the ship names and have the name printed on a texture?

When it gets released, I want to name a windmill class ship "Lady Devorgilla".

 

[Ursus]

Is it possible to give the ship names and have the name printed on a texture?

That's definitely somewhere on the "to do" list. Computerex and I were talking about it on the M6 dev thread. I was kind of hoping he was volunteering to implement it, but... looks like I'll probably wind up doing it myself.

Hmmm.... looking at Trokkes' tank texture...

If we extended the "name plate" through the front and rear bands (err... the thick gold bands, not the dark lines... ) around the tanks, and left the actual name off, we should have enough room for a fairly long name in that space. Oops, just a sec... :hmm:that's a separate name plate, rather than just a texture on top of the cylinder texture, and looking at the actual texture file, rather than the rendering... that might not be so easy. The texture might look all right if we stretched it, but that would stretch the name, too...

There seems to be a tiny bit of unused space to the left of the name plate texture; if Trokkes wanted to use that (which would involve a little bit of re-mapping)...

I looks like that'd just fit (yeah... I know... it's not quite aligned, but it was just a feasibility test), and if a person used the same correct spelling you did, it'd leave a little bit more room. (I actually Googled "Devorgilla", trying to figure out the significance of the name. Even after reading about Dervorguilla of Galloway on good ol' Wikipedia, the only significance I can find is that it's a long name.)

... or maybe it'd be better to use a separate texture for the name plate(s), anyway...

[Note: The following was written before I discovered that little complication.]

If we wanted to get really fancy, we might have an optional tag for a registry number or call sign in the scenario file, so the name could be printed "right side up", and the registry number could be printed "up side down" below it.

 

[Urwumpe]

(I actually Googled "Devorgilla", trying to figure out the significance of the name. Even after reading about Dervorguilla of Galloway on good ol' Wikipedia, the only significance I can find is that it's a long name.)

Actually, it is much simpler, if you don't follow the primary hits. :lol:

Lady Devorgilla is the name of a Scottish windmill which stands in downtown Gifhorn, Germany. It is related, but not part of, the international windmill museum in Gifhorn.

Why not name a windmill class ship by a windmill... OK, I must admit, it is only of local fame...

 

[trokkes]

Hi there,
Still working on the textures in the little spare time I have :beach:
For the docking ports, at the moment I'm using the "standard" docking port (only a little more blue)



the size of this one is a little big yet.


I have redone the mesh of the truss zone because the "rebels" normals was making me crazy . So these are fixed now and the texture has a little more detail.


I'm changing too the texture for the radiators. I didn't like the old one. Well, this one doesn't like, so I'm open for suggestions

Ursus said:
There seems to be a tiny bit of unused space to the left of the name plate texture; if Trokkes wanted to use that (which would involve a little bit of re-mapping)...

Yes, it's possible to do that. In fact, I'm changing the textures in some zones, because I'm learning new things. So I think I'll never finish it :rofl:

There are other places too to place the name of the ship. In the truss and in the central hab zone. So, "Lady dervorguilla" or "El ingenioso Hidalgo Don Quijote de la Mancha" could fit easily in the truss zone.
Perhaps we can use the tanks for the generic name and number, like ST ships and the truss and central hab for the different name of each ship.
I'll do some test.

Btw, I like the appearance of the engineering panel. It's :speakcool:

If you have more suggestions for the textures, please tell me.

 

[trokkes]

Ehm...

Yesterday at night my LCD monitor dead. I brought back my old CRT monitor to work again and it works fine but when I open orbiter, the textures of the windmill seems very (very, very) dark.
Even the images in the previous post are barely visible in my old monitor. They are too dark.

So guys, I have some questions for you:
Do you see the images of the previous post too dark? or it's my old monitor?
Do you use LCD or CRT monitors?
Do the textures need to be more clear?
What is the meaning of life?

 

[Ursus]

My condolences on your monitor.

Do you see the images of the previous post too dark? or it's my old monitor?
Do you use LCD or CRT monitors?
Do the textures need to be more clear?

The textures look nice and clear on my current (CRT) monitor. I have had monitors in the past on which they'd probably be indistinguishable from the blackness of space, even with the sun shining straight on them, let alone make out any of that nice detail.

Eventually, we'll have running lights, so people with at least be able to see where the ship is if they have brightness-disadvantaged monitors.

What is the meaning of life?

¿Quarenta y dos?

Er... I mean...

Forty-Two?

(I'd better not try posting in Spanish, or someone will try replying in Spanish, and the next thing I post will be "¿que?")


About places to put the name... The hab hub is a good place. Maybe the big conduit that runs down the center of the truss, too... but the name would get covered by truss attachments more often than not.

 

[Ursus]

I'm still working on it, or rather, trying to get back to working on it. Just trying to put together some documentation for the alpha, right now.

Also, I just added mesh and a config file for a 40ft self-propelled docking tunnel to the UPDock section the repository. It requires the UPDock Alpha 0.1 package to work, of course. Future releases of UPDock will include the SPDT.

 

[Woo482]

whats the spdt?

 

[Ursus]

whats the spdt?

It's a Self-Propelled Docking Tunnel (extension). It's a docking tunnel with its own RCS thrusters, so a person can deploy it from the cargo bay of an XR5 (or from its own carrier on a Windmill Class vessel, once such a carrier is built, or from just about any carrier that'll hold a forty-foot equivalent intermodal transport unit, or FEU), switch focus to the SPDT, and dock to one of the vessels the tunnel is preparing to connect.

A docking tunnel extension is neccesary when docking an XR5 to the Windmill, unless a person really wants to watch the bracing beams for the habitat wheel slice through the XR5's nose.

 

[Woo482]

ok

 

[Ursus]

I finally got around to putting together and uploading a distribution file for testers. I'm calling this a semi-private Alpha (maybe even Pre-Alpha) release. It's certainly nowhere near feature-complete, so I'm not publishing a link to it, but I won't (and actually, putting it on SourceForge, I couldn't if I wanted to) stop anyone who can find it from downloading it.

By "not feature-complete", I mean:

- No reactor simulation.

- No "navmodes" (attitude autopilots) whatsoever (except for KillRot, if you count that). I tried re-enabling the Orbiter native navmodes, and decided they where worse than nothing for this vessel.

- No attachment code, even though attachment points are in place on the truss.

...et cetera...

What we do have is what I think is a pretty nice (if I may say so myself) Bussard collection system, a magnetic sail and a fairly decent autothrottle modes to help with fuel management.

Next, I think I'm going to put a digital accelerometer, with vector and/or degree displays, on the... well, I was going to put it on the engineering panel, but it'd make more sense to have it on the helm panel. Yeah... I'll put it on the engineering panel. Eventually, I think I'm going to have to do a bit of re-arranging of the helm panel, and will add the digital directional accelerometer at that time, among other things. Right now, the engineering panel is turning into a nice place from which to navigate, even if it does have only one MFD.

The accelerometer shouldn't take long. Then...

Anyone know how a futuristic fusion reactor should behave?

Let's see... just off the cuff... the following probably isn't very accurate/complete.

First you add some reaction fuel to the reaction chamber. (How much? Would it need some sort of pre-treating?)

You compress the fuel down with a magnetic field, maybe adding some more as you go along.

You shoot a whole bunch of electricity through the fuel to excite it into fusing together.

In the hypothetical future reactor, enough electricity can be generated from the reaction to maintain the compression field (and the stimulus charge? or is that just necessary to jump start the reaction?) and other ship systems, such as those powerful magnetic field generators at either end of the vessel. I guess we could add a fission power plant if that seems too unrealistic.

To produce thrust, (lots) more fuel is added to the system, and the fused fuel is directed out through the big exhaust port.

When the reactor is not being used for propulsion, the spent fuel is vented through two or more symetrically placed exhaust ports, so that the thrust produced by said venting cancels itself out.

To stop the reaction, one simply stops replacing the spent fuel.

Of course, what I really need are key phases of operation, so I can put something more interesting on the reactor moitoring display than "Startup Sequence: Phase 1... Phase 2... Phase 3..." Though I guess that'd work I can't find anything better.