Jump to content

[WIP] THSS - Tri Hexagonal Structural Truss Reborn (v1.0 RC1 -7/19/14 )


noonespecial

Recommended Posts

Unfortunately, the docking adapter was the one thing I always hated about THSS. Especially considering the FusTek or KSO docking adapters are SO nice. But, we'll see.

To each his own! :) I'm just happy for more structural station parts.

Link to comment
Share on other sites

Wow, Greys, that helped a lot! That was exactly what I was looking for. I tried it out and it worked pretty good. I'd like it darker, but I'll play around and see what I can do.

DMagic, thanks for the link. I will most definitely checked it out.

Here is the result of the first texturing attempt:

LxC1IBh.png

Link to comment
Share on other sites

That's quite impressive. I'd smooth out the texture's patchiness a bit. What I's really suggest is cross members on the truss. IRL, the cross members prevent torquing, twisting, and ultimately failure of the girder/truss/whatever-is-supposed-to-be-structurally-sound.thats my only complaint. but the shape and detail is superb.

Link to comment
Share on other sites

Wow, Greys, that helped a lot! That was exactly what I was looking for. I tried it out and it worked pretty good. I'd like it darker, but I'll play around and see what I can do.

DMagic, thanks for the link. I will most definitely checked it out.

Here is the result of the first texturing attempt:

http://i.imgur.com/LxC1IBh.png

Awesome!

That's quite impressive. I'd smooth out the texture's patchiness a bit. What I's really suggest is cross members on the truss. IRL, the cross members prevent torquing, twisting, and ultimately failure of the girder/truss/whatever-is-supposed-to-be-structurally-sound.thats my only complaint. but the shape and detail is superb.

Agree

Link to comment
Share on other sites

That's quite impressive. I'd smooth out the texture's patchiness a bit. What I's really suggest is cross members on the truss. IRL, the cross members prevent torquing, twisting, and ultimately failure of the girder/truss/whatever-is-supposed-to-be-structurally-sound.thats my only complaint. but the shape and detail is superb.

I agree 100%. The model in the picture is incomplete. I read Greys's post and gave it a try with an unfinished model. I'm going to model and attach the radial modules before I place the cross members so I can find a suitable place for them that makes sense without clipping.

The modules look like this (again, this was just a half meter concept part).

3OhAl4X.png

Ip9MqTy.png

Link to comment
Share on other sites

That looks pretty good. I might start on a classic version that's nearly the same as the original. I can also answer any texturing questions or even make some textures for you guys if you like.

Link to comment
Share on other sites

That looks pretty good. I might start on a classic version that's nearly the same as the original. I can also answer any texturing questions or even make some textures for you guys if you like.

Hey SpeedyB, I was going to make a classic version when I finished these, but if you want to take the initiative, by all means! :D. Also, if you'd like to help with textures, I would love the help.

- - - Updated - - -

Updated models with Cross members that don't clip with fuel modules.

Javascript is disabled. View full album
Link to comment
Share on other sites

This is great! I was always a fan of the original THSS and I just love what you are doing here! I'm just wondering if it's not to complex for large structures? The curves are nice though, keep it up! ;)

Link to comment
Share on other sites

It would be interesting if you and sumghai got together and itegrated both mods together. The reason why is that picture from the original fusty parts mod. Would it also be appropriate to have longer truss sections I mean real long 10m plus parts so the builder can cut down on deployed parts. Could we also have sections that can link walkways togerther so they can be used with the connected living spaces mod?

Link to comment
Share on other sites

This is great! I was always a fan of the original THSS and I just love what you are doing here! I'm just wondering if it's not to complex for large structures? The curves are nice though, keep it up! ;)

There are a lot of places I could simplify the models. If performance becomes an issue, I will most certainly adjust.

However, I'll need feedback for this. My PC is a monster and it takes a lot to lag it.

My rig:

Intel i7 4770k

MSI MPower Max

GeForce 770

16 GB Ram

I run KSP (and all my games) off an SSD. Oh, did I mention that it's liquid cooled?

Yeah... I'll need other people to tell me if there is a performance hit. I probably won't notice it myself.

Link to comment
Share on other sites

6/29/2014 Update. I took a small break from modeling to work on textures.

Here's what I got so far. This is a concept piece of a 1.25m command pod with built in air intakes that I've knicknamed "Arrowhead". I don't know if it will actually make it into the release, I'm mainly using it to practice modeling and texturing.

Javascript is disabled. View full album
Link to comment
Share on other sites

Graphics aren't much of an issue for KSP, everything's very minimal and mesh complexity doesn't meaningfully contribute to memory load; plus KSPx64 is supposedly coming which will raise the memory limit by a lot. Texture resolution is currently fairly expensive but these parts shouldn't need more than a 1024 sheet, and would probably do fine with a 512 sheet; light weight users may desire 256 sheet but at that point they should be using the Active Texture Whateverthethirdwordiscurrently mod for compression.

Physics are expensive, so try to get your collision mesh as simple as possible. Mine are using 3 Cube Primative colliders, which are super cheap; it's how I got away with having mine be hollow. Because yours have those concave edges things are going to be complicated for you. Probably should have mentioned this earlier but mesh colliders on parts in KSP 'have' to be convex, technically they don't, but concave colliders are crazy esspensive. You have a few options, one of them is to just have a truncated triangular prism as your collider (could do with 6 cube colliders or 1 mesh collider) and ignore that your collision mesh is grossly unlike your visible mesh. Another is to have several convex mesh colliders overlapped that result in something similar to your shape. Say two triangle prisms rotated slightly so that each triangle is one set of corner points, then some other colliders filling in the space left over; I'll edit in a sketch of that.

Edit: This composite is composed of two mesh colliders and 9 cube primative colliders, more cube primatives can be used rather cheaply to enhance the accuracy of the representation, but at some point they'll have to reach into the tip of the triangles and then you'll need to use mesh colliders so that you can make it thin enough at one end but not leave any air pockets inside, as these colliders will compose the ends of the struss too. Looking at it I just realized that I have airpockets, oh well, just a sketch, you'd want to make the short side cubes thicker to fill in that space, and maybe add a third one to cover that little slice

j2cDHvT.png

Approximating the visual mesh is key, accuracy isn't; and you can make your modules so that their mesh doesn't contain the entire module; and you can let things clip into eachother for building purposes.

Edited by Greys
Link to comment
Share on other sites

Greys, this is the second time you've posted here and given me more information than I've found in a dozen tutorials. :D. I think without these little instructionals I'd be sitting here with barely functional naked models of unreinforced tri-struts. Haha. Thanks a million mate. The colliders I was using did leave a fairly large air pocket on all three sides.

Link to comment
Share on other sites

Yes, performance is one thing, putting parts together is another. The collision box (with triangles and cubes as pictured) is as simple as possible, but I do believe the short ends should be flat, not concave. It will be much easier to use these parts that way: imagine z booster attached to that edge (I like it that way) - it will probably by only attached to one of those collider edges, so if it moves, it will not be in-line with the vehicle, hence inducing a rotary movement on the whole. I could draw a picture, but I'm too lazy right now (just came back from work), maybe in an hour or so, if it's unclear?

Link to comment
Share on other sites

Good point Frederoo, but that can be fixed by making sure that all the sides have a 'parallel to flat' collider face at the trough point of the curve,, which basically just means always have an odd number of colliders make up that surface.

However, I had a worry for a moment yesterday and asked some more professionally skilled modders and the way I've done it with all that overlapping may cause some severe physics generally-being-a-doof, the only way to tell for sure is to try. The problem is that there are a lot of coplanar faces that overlap, if this were a rendered mesh there would be Z fighting everywhere:

sQPI4iE.png

Now, this isn't a rendered mesh, it's a physics mesh, and collision is supposed to generally be a yes or no thing, but there's not much faith that Unity's physics mixed with KSP's physics will result in things working correctly in this situation. The suggestion is that instead of using primatives just make a well fitting mesh and cut it up. This way you get the same effect, but the mesh never overlap, they meet at a lot of places, but that doesn't matter. This will unfortunately make the part a lot more expensive. You may still be able to make use of a lot of primatives and then fill in the space with mesh as long as there's no overlap.

To reiterate, this is theorycrafting, it may work fine so test it; but it also might explode all over your face. What I recommend testing is placing things against the ends in the editor and with KAS because it lets you do that in flight; and ramming various things into the ends. For example, walk a kerbal up to it and push them against it walking back and forth; drive a buggy into it at reasonable velocity. Place it upright on the runway and then drive a buggy 2.5km away and back to see what it does when reloaded; also build some basic ships and do the same; they should be fine because of clipping rules but you don't want to learn that they aren't when you go to space. Place one upright and have a kerbal walk on the top side. Timewarp while things are touching it. But the most valuable thing will be when you release some parts and users can start bulk testing; assuming it survives your tests.

Link to comment
Share on other sites

Thanks for all the information!

What I'm doing is rather simple. On the models that would have complex meshes, I do a two step process. First, I export the completed model to dae, then I build a simplified shell over the model (pretty much following the example you gave above). I then export that into a separate dae. Then I finish them both in Blender (remove doubles, texture (on "hero mesh"), scale, align, etc.) and then move to Unity. I import both models, then substitute the collider mesh with the outer shell mesh I made.

It's probably doing more work than needed, but as I'm still learning Unity and Blender, it's the best method I've found for me so far and it gives me total control on exactly how the collider will wrap around the model.

Link to comment
Share on other sites

I've finished a small medium and large truss.

I have them textured and looking perfectly stock. Unfortunately my main internet is out so I'll upload the pictures once its fixed tomorrow.

Link to comment
Share on other sites

I've finished a small medium and large truss.

I have them textured and looking perfectly stock. Unfortunately my main internet is out so I'll upload the pictures once its fixed tomorrow.

That's GREAT Speedy! I can't wait to see them :D

Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...