Jump to content

Mother Development Thread


Bonus Eventus

Recommended Posts

BGKpOXf.jpg

d02WIFs.jpg

gLstFeO.jpg

L7fMe4R.jpg

I think this version is the winner. I'm still not happy with the fusion reactor adapter (I think it needs more trusses), but I really like the engine shrouds. Reactor has a 4M diameter and a length of 12M.

 

Edited by Bonus Eventus
Link to comment
Share on other sites

36 minutes ago, Bonus Eventus said:

 

I think this version is the winner. I'm still not happy with the fusion reactor adapter (I think it needs more trusses), but I really like the engine shrouds. Reactor has a 4M diameter and a length of 12M.

 

"Moar TrussesTM" for the win.

Peace.

Link to comment
Share on other sites

Wait, you are planning to have large onboard fusion reactor and you not using it for direct thrust? Sorry but that doesn't make any sense. Usually when you have a big bulky fusion reactor on you ship, you attach a magnetic nozzle to it to generate High Isp thrust (which can be lowered by adding more propellant). The reason for this is that the most efficient way of converting fusion energy into useful thrust. Converting it first into electric power for your ion engines would incur huge efficiency losses which mean you need a lot of radiators to get rid of all the resulting wasteheat.

Instead I would simply call is a Propellant tank adapter which for a mother ship  would be refilled multiple times.

Edited by FreeThinker
Link to comment
Share on other sites

2 hours ago, FreeThinker said:

Wait, you are planning to have large onboard fusion reactor and you not using it for direct thrust? Sorry but that doesn't make any sense. Usually when you have a big bulky fusion reactor on you ship, you attach a magnetic nozzle to it to generate High Isp thrust (which can be lowered by adding more propellant). The reason for this is that the most efficient way of converting fusion energy into useful thrust. Converting it first into electric power for your ion engines would incur huge efficiency losses which mean you need a lot of radiators to get rid of all the resulting wasteheat.

Instead I would simply call is a Propellant tank adapter which for a mother ship  would be refilled multiple times.

I'm not trying to be flippant, but this just made me visualise a big bunch of ships out there with bulky fusion reactors. Have to say it would be a great sight though :D

Link to comment
Share on other sites

11 hours ago, FreeThinker said:

Wait, you are planning to have large onboard fusion reactor and you not using it for direct thrust? Sorry but that doesn't make any sense. Usually when you have a big bulky fusion reactor on you ship, you attach a magnetic nozzle to it to generate High Isp thrust (which can be lowered by adding more propellant). The reason for this is that the most efficient way of converting fusion energy into useful thrust. Converting it first into electric power for your ion engines would incur huge efficiency losses which mean you need a lot of radiators to get rid of all the resulting wasteheat.

Instead I would simply call is a Propellant tank adapter which for a mother ship  would be refilled multiple times.

 

-I've yet to see a fusion engine setup that didn't call for huge radiators.

-Most fusion engines I've seen produce low thrust at sub 100 KN. ISP in the 5000s though, but still kind of a bummer for moving around 100+ ton ships in KSP. Also, not very extensible. Can't cluster magnetic engine nozzles (like you pointed out). 

-If I drop the fusion reactor what do I replace it with—an equally bulky fission reactor or a million solar arrays? Those ion engines consume gigawatts of power.

-The fusion reactor I had in mind uses magnetic targeted fusion, by way of compressing plasma surrounded in a vortex of molten lead with a spherical array of pistons. Can't understand how this sort of reactor could be used for direct thrust. Maybe open gas core cycle? Use a heat exchanger to heat up propellant?

Edited by Bonus Eventus
Link to comment
Share on other sites

41 minutes ago, Bonus Eventus said:

-I've yet to see a fusion engine setup that didn't call for huge radiators.

You always need a decent amount of radiators for on board system and fusion production but the less energy you have to convert to electric power, the less conversion losses and therefore wasteheat and the need for huge radiators

41 minutes ago, Bonus Eventus said:

-Most fusion engines I've seen produce low thrust at sub 100 KN. ISP in the 5000s though, but still kind of a bummer for moving around 100+ ton ships in KSP. Also, not very extensible. Can't cluster magnetic engine nozzles (like you pointed out).

Yes but science always marches on, while tomorrow we might only be able to generate 100 KN, the day after it might be double. That's why I believe you should not have a fixed performance but it should scale with you tech progression.

41 minutes ago, Bonus Eventus said:

-If I drop the fusion reactor what do I replace it with—an equally bulky fission reactor or a million solar arrays? Those ion engines consume gigawatts of power.

Realistically I would put my money on beamed power by huge photovoltaic arrays, which can use the sun as a auxiliary power source. It would be realy cool if you could make Mothership sized solar array optimized for beamed power (meaning circular or squared dimensions)

Edited by FreeThinker
Link to comment
Share on other sites

47 minutes ago, Bonus Eventus said:

-The fusion reactor I had in mind uses magnetic targeted fusion, by way of compressing plasma surrounded in a vortex of molten lead with a spherical array of pistons. Can't understand how this sort of reactor could be used for direct thrust.

magnetic targeted fusion is probably the least efficient method for power production because it main fusion power will be thermal heat which is difficult to convert to electric power and required large amount of radiators. Instead I would recommend a Colliding Beam fusion reactor (a.k.a Tri Alpha) which main fusion product are charged particles which can be converted into electric power at much higher efficiency than thermal heat..

 

49 minutes ago, Bonus Eventus said:

Maybe open gas core cycle?

A open cycle gas core reactor would indeed be a good better for a Mothership, as the reactor is hot enough to fully ionize it exhaust which can be converted by a MHD power generator at relatively high efficiency to electric power.

53 minutes ago, Bonus Eventus said:

Use a heat exchanger to heat up propellant?

Thermalheat exchangers are limited by the maximum temperature materials can operate at which is about 3700 K , this mean the upper limit of their isp is only 1200s at best

Edited by FreeThinker
Link to comment
Share on other sites

  • 2 weeks later...
On 3/24/2018 at 9:34 AM, SkyHook said:

Have you give any thought to releasing a Version 1 of this pack? I love the parts, but I think working on this pack in phases would allow you to release something to us salivating users, while you work on different versions/iterations of the pack. For example, you could release the engines and trusses in version 1, but version 2 could incorporate the inflatable/rotating habitats, and version 3 could incorporate spaceport/shipyard parts. I only say this because I have been following development for 2 years, but I don't know that you have released anything. This makes me reluctant to support you on Patreon, etc simply because there is no plan to release the parts. Hope you take this suggestion into consideration.

 

Sincerely,  A long-time fan of the project. 

After mulling it over a bit, I think this is a good idea. Because of the scale of the mod, I've been prone to chip away at different aspects quite randomly. This will give the development some much needed structure. So, here's the plan:

The beta release will roll out in blocks.

Block 1 - Launch Hardware (Single resource tanks, Bipropellant fuel tanks, first/second stage M1 engines, decouplers, engine adapters, and size 4/6 fairings)

Block 2 -  Couplers (radial decouplers and docking ports)

Block 3 - Trusses

Block 4 - DA 42 Command Pod

Block 5 - Truss Adapters

Block 6 - Powerplant (solar panels and small fusion reactor)

Block 7 - SAS and RCS

Block 8 - Electrical (fuel cells, capacitors, and lights)

Block 9 - Radiators

Block 10 - Deep Space Engines (ion thrusters and fusion engine)

Block 11 - Storage (rocket part containers, food containers, cargo bays, KIS containers)

Block 12 - Manufacturing Module

Block 13 - Docking Bays

Block 14 - Habs

Block 15 - Centrifuges

Link to comment
Share on other sites

2 hours ago, Bonus Eventus said:

After mulling it over a bit, I think this is a good idea. Because of the scale of the mod, I've been prone to chip away at different aspects quite randomly. This will give the development some much needed structure. So, here's the plan:

 

Huzzah! I am glad you took it into consideration. I love the new plan too. Block 1 can't get here soon enough! 

Link to comment
Share on other sites

On 3/19/2018 at 12:49 PM, Bonus Eventus said:

9w4BcKN.jpg?1

Ean9uEQ.jpg

Finished modeling the docking bay adapter.

Time for me to build the venture star. BTW, can you talk to the Kerbalism and deepfreeze guys, so you could maybe make a giant radiation shield for kerbalism and giant deep freeze contianers, holding 100 kerbals etc, sumn so

Link to comment
Share on other sites

OD7Eued.jpg

 

bOslSUI.jpg

The new tanks will all have switchable textures. At least 3 options. Had to redo the textures unfortunately to fit them all into the same atlas. It came out good though!

EDIT: Thought I would explain a couple of things about the process to create the new version of the tanks. 

1) The initial problem I ran into was that I have many tanks in many different sizes, which won’t fit easily in a single texture sheet. Sizes are 1.25, 1.875, 2.5, 3.75, 5, and 7.5 meters. Heights for the tanks include, 1.25 ,3.75 ,7.5 ,and 15 meters. Diffuse textures weren’t too bad. The real problem was normal maps with 4 way symmetry. The normals for modular textures have to be made with flat geometry ( imagine high res geometry for a floor tile normal map). I was making quarter tile then mirroring the polygons vertically and horizontally for a larger tile. If the normal map had vertical grooves, smoothing would get very confused when the verts were merged. If the verts aren’t merged, then you will get light leaks. 

2) For the external feed lines, it’s nice to have some occlusion shadows. This means unique versions of some tiles/panels. So, even less UC texture space.

SOLUTION: The solution was counter intuitive. I made a quick experiment, modeling all of the panels like I would for a high res model for a unique normal map, just without the sub d bevels. I started with the largest tank. The tri count for the tank walls without end caps came to just under 6k. That’s about the polly budget for a typical engine in KSP. Surprisingly enough, not too bad. I still used normal maps for some extra details, but only in areas where merging mirrored polys together was unnecessary. 

There was one caveat though. Vertex smoothing became discontinuous. Panels looked too faceted and not quite round enough. I could have just upped the poly count, but instead I used a neat trick. I created a separate cylinder the same size as the tank and then transferred it’s vertex map to the problem polygons on the tank. It looked amazing.

Then I got ambitious. it would be nice if the rcs ports were real geometry too. Only problem there is it creates five way intersections when you cut holes into quads, which makes strange smoothing issues Here I transferred the vertex map too—result was perfect smoothing. This technique would work great for command pods that have to have their surfaces cut up for windows. I think that’s what Squad probably did for the rcs ports on the new capsule.

Sometimes the most straight foreword dumb method is the best method.

 

Edited by Bonus Eventus
Link to comment
Share on other sites

No date yet. Still have work to do...

  • one more tank cap to texture.
  • Only partially modeled the decouplers.
  • Need to model the S6 fairing.
  • Add smokescreen configs to the engines.
  • Finish the adapters.
  • Texture the S6 engine mount.
  • And a few other things. 

qJmU08N.jpg

pT26g8Y.jpg

 

3Im6hqY.jpg

Wzvb77t.jpg

vmocuXH.jpg

IRd6HEn.jpg

Edited by Bonus Eventus
Link to comment
Share on other sites

On 18-4-2018 at 6:32 AM, Bonus Eventus said:

qJmU08N.jpg

pT26g8Y.jpg

 

3Im6hqY.jpg

Wzvb77t.jpg

vmocuXH.jpg

IRd6HEn.jpg

 

 

Whoha, that a lot of engines. Are you aware this will cause a huge performance hit in KSP which have to simulate the physics of each individual engine pushing on your craft and that is something you want to avoid because big ships have unavoidable large part count already.

Edited by FreeThinker
Link to comment
Share on other sites

Yes I'm aware of all of that. I haven't finished the part yet. I just wanted to get it in the game early to test some things out.

Fortunately the outer two rings of engines will be built into the part. In reality it' will be eight engines. 7 with gimbal in the center.

EDIT: you know I’m wondering if I shouldn’t just combine them all. If I add a skirt, much like the Merlin’s on the falcon 9 octaweb, I could cut down on the polygons, and use bone animation instead of constraints for gimbaling. The part would act like the old behemoth engine.

EDIT: Bone animation successful. Now for the tricky bit of managing 42 radially placed fx transforms...sheesh

8cFpcTR.png?1

Edited by Bonus Eventus
Link to comment
Share on other sites

Started working on SmokeScreen configs today. I dug into Real Plume configs to learn the gist of how it works.

u3aweCy.jpg

I was able to get something working for the hypergolic RCS thrusters. Still a little weird. I feel like the plume is not long enough. I think that I don't understand enough 
about SmokeScreen to make the correct custom particle model in unity.

In other news I'm recruiting. If you'd like to contribute to Mother's Dev Team, check out this thread here for more info.

 

Link to comment
Share on other sites

  • 2 weeks later...

 

lAm9VQG.jpg

qNR4RbG.png

E7VVOub.jpg

Thanks to @Saltshaker @neistridlar @JadeOfMaar @cineboxandrew for helping me get this new nosecone drone core working. I've finally gotten the hang of smokescreen. 

The launch smoke and flare are almost where I want them. Block 1 is coming along nicely. Still need to finish another engine cluster (x9), finish the other nose cones and adapters, 

finish the air brakes, finish the decouplers (these are mostly in the bag), and finish the size 6 fairing. Also need to finish configs for modular fuels, IFS, and create some custom

bulkhead icons for size 6 parts.

Link to comment
Share on other sites

On 4/8/2018 at 5:42 PM, Bonus Eventus said:

If the normal map had vertical grooves, smoothing would get very confused when the verts were merged. If the verts aren’t merged, then you will get light leaks. 

This depends on how you import the model. the light leaks come from Unity extruding the vertices along their normals for the shadow pass. If the vertices are split but still have parallel normals they will extrude the same way and you wont get light leaks. In blender at least you can have a single "vertex" have multiple UV corrdinates, but whats really happening is its several vertices with parallel normals. Just make sure you select import normals in Unity

Link to comment
Share on other sites

  • 2 weeks later...
Guest
This topic is now closed to further replies.
×
×
  • Create New...