[1.2] SpaceX Interplanetary Transportation System by Thrimm Aerospace [0.91]

[reekchaa]

11 hours ago, Draconomial said:

I saw the video on the SpaceX subreddit. How's development going @Thrimm?

Very comprehensive discussion still going on at subreddit.  Thanks for that.  Full link here:  https://www.reddit.com/r/spacex/comments/5itma8/spacex_mars_colonization_simulation_by_mark/
Eager to know how the Split-Flaps development is going as well for aero control surfaces.  

Edited December 18, 2016 by reekchaa
visual

[Thrimm]

Hello guys! A quick update on what's been going on for the past months with this mod.

As we have no real news from SpaceX about the design for the split body flaps, I tested multiple different options myself. I tried adding them as built in parts to the landing legs, to the body of the spacecraft  and unfortunately none of those solutions seemed to work. What I was getting was some level of control over the ship, but nowhere near what is needed for realistic reentry let alone flipping the vessel for landing on Earth (without reaction wheels that is). Problem with this ship design is that is looks like a bullet and really wants to fly straight, so adding small control surfaces doesn't really counter that. Also, for some reason FAR enforces somewhat strange orientation of the vessel even in level flight (rotated by about 60 degrees, on fin facing down) and I'm not exactly sure how realistic this is - and it is important because if want to stick to original SpaceX concept of having control surfaces for pitch and yaw only, then the roll problem has to be tackled separately.

I've been toying around the idea of making the ship modular, so command pod, fuel tank, engine mount, cargo bay, engine mount etc. so assembling a crew or tanker version of this ship would be possible, as well as any other type of monster you could come up with. It would require some changes to the overall shape of the mesh but it might be possible - I would like to hear your opinions on this.

Last, with all of the aerodynamic problems listed above, I decided to acquire the Aerospace Toolbox for Matlab and I will use to calculate real aerodynamics of the vessel and then transfer the approximated values to the game. At least we might hope it will be realistic and maybe we will find if there are some real problems in the design, and not just the game related ones.

On top of that there are still optimizations and patches for RO. I have mostly finished working on RealFuels patch, but I have little experience with other parts of RO, so if anyone would like to contribute it would be greatly appreciated.

Another thing I tested is adding the heatshield as a separate part that is in fact a heatshield, not just a decoration. It is possible and I think it would add to the realism, but we're limited by game aerodynamics (docking port already causes issues sometimes, adding extra drag to one side).

 

[reekchaa]

Beautiful, Mark... would love to Beta-test anything you have made changes to so far.  
Making the ship Modular sounds like a great idea.  Having the ability to customize and configure a monstrous beast would give everyone much more replay option, and balance out to the realism of development.  If it works, this would greatly enhance our inspiration to tinker and build.
A Heatshield (ablative or not) sounds great too, but it's true getting aerodynamics to work as advertised by Elon should stay a top priority in my book.  I'm still unsure how easily this 'bullet' will ultimately turn, testing and developing control surfaces of my own, investigating how the other early ITS designs out there are working (and not working).  
I'd be happy to offer back more direct feedback, share my own design mods, whatever you need to keep your development on this going, inspiring more peeps to the real world engineering table. 

[Probus]

@Thrimm,

  Are you still developing this awesome mod?  I wonder if @ferram4 could help you with the aero calculations.

[Thrimm]

22 hours ago, Probus said:

@Thrimm,

  Are you still developing this awesome mod?  I wonder if @ferram4 could help you with the aero calculations.

Yes I am. The actual problem that I am facing right now are not the aerodynamics calculations itself as this is rather straightforward, but rather implementing realistic aero behaviour in somewhat simplified model we have in KSP. You see the thing is that you can have a relatively accurate representation of a spacecraft in a numerical environment using complex numerical models and derive some sort of behaviour patterns from it and then you want to implement those patterns in much less sophisticated model that is represented in game, without the loss of generality. This is problematic, because game physics and aerodynamics has to be much less complex to be calculated in real time.

It can be relatively easily achieved in the case of planes or simple rockets but here in ITS there are a number of somewhat interesting and non-canon solution that were implemented by SpaceX. Take grid fins for example - they work as a deployable control surface that generates very little actual lift when deployed and close to 0 when stowed. This is impossible in KSP. When a lift value is attributed to a game object it is always there, you can animate it and change its orientation but you cannot suppress it without developing a special plugin for it. Another thing is the vessel itself - it clearly is designed to rely on body lift with legs/fins being stabilizers. This creates another problem, because body lift in KSP follows much different lift-to-drag curve than regular lift does and is also omnidirectional. Regular lift is directional and is governed by an attributed game object. The problem arises when, as is currently the case, you attribute a regular lift to the stabilizers (in order to have them create a force in a specific direction) and rely on body lift for the vessel. This will cause the center of lift of the vessel to be moved very far backwards that will also follow a very different lift-to-drag curve as a function of velocity, resulting in the vessel being extremely nose heavy upon reentry. You could argue that one can add a lift value to the vessel itself to counter that, but then you end up with directional lift only which is clearly not the case in reality, as ITS has a point symmetry - I'm currently exploring that idea.

Another thing that is a great simplification in KSP is the absolute lack of influence of airfoil on the lift parameters. That causes enough issues if you want to build a regular plane (especially if you want to copy an existing, real design) but here with all the new solutions such as gridfins, split body flaps and general use of different types of airfoils for the booster and the ship itself simply becomes very complicated.

So for example imagine that you have the ITS vessel sorted out by flat out replacing body lift with regular lift and you've managed to make it omnidirectional to represent point symmetry of the vessel. Assuming that your vessel is now maneuverable upon reentry, total lift value you have in the game is now so high that it is unstable when attached to the booster, as its CoL has moved very far in front of CoM. This is an exiting issue in KSP that can be easily visualized if you try to build a DynaSoar replica or in fact put any spaceplane on top of a rocket. Obviously I could just simply increase the lift values for booster fins but that will just move it further and further away from being realistic.

So the bottom line is that the main problem lies in finding those behaviour patterns for the spacecraft and the booster in a complex and accurate numerical environment that is provided by Aeorospace Toolbox (which can be very accurate if applied correctly) and the simplifying those patterns without the loss of generality in such a way that they are represented in the aerodynamics model we have in KSP, even with FAR. And since SpaceX decided to use a lot of non-standard solutions, it's currently causing an issue.

Thanks,

Mark

 

P.S. I'm sure you're aware of this, but to further illustrate the lift problem we have in KSP, check out how the same design will behave in Simple Planes, where you can actually change the airfoil type on the wings and control surfaces.

Edited May 5, 2017 by Thrimm

[Probus]

Thanks for the feed back Mark.

I am not 100% sure of this, but I think I recall @NathanKell working on lifting body mechanics a year or two ago.  He may have some suggestions too.  At any rate, good luck at finding a solution.  This is a great mod.

[The-Doctor]

Can't wait to play with this

[The-Doctor]

Hope you take my considerations regarding making it like the ships in NatGeos MARS series

[paneledberry]

I read through the issues post above, and it seems like there's lot to work out in this mod. I also get that people have real lives Even so, is there an estimate on when it will be updated for 1.3? I came late to the party and would love to try it out but have no idea how to revert back to 1.2, and have quite a few mods installed on 1.3. Thanks

[RandyRawgust]

Sorry if this is a tad late, if you haven't figured it out yet, this is what I did.

Check the properties tab for KSP. And check the participation box on the Settings page to opt-in to the latest beta. You can opt-out of the beta at any time by removing the checkmark. Once you've checked the box you can select the "check for updates" button to begin downloading the latest Steam client beta if available, like KSP 1.2.2. (Thanks Google :P)

Oh and if you got the space for it, I actually copied the game folder directly to 'My Programs', that way I can have a clean Steam version that stays current and a client specifically running 1.2.2. Best of luck!

On 8/24/2017 at 7:45 PM, paneledberry said:

I read through the issues post above, and it seems like there's lot to work out in this mod. I also get that people have real lives Even so, is there an estimate on when it will be updated for 1.3? I came late to the party and would love to try it out but have no idea how to revert back to 1.2, and have quite a few mods installed on 1.3. Thanks