Stock Tank Track (belt drive) development

[Majorjim!]

Hi all,

 

 I am very excited to be able to share with you a KSP stock first, tank tracks! Or at least that is what I am trying to make! The idea came to me when playing with my little 'therm-o' hinges (trade mark) lol not really. I began to realise that it may be possible. So I tried to imagine how something like that would look like. It would be a collection of 'links' looped back with the first and last connecting to each other.

I knew if in future it would be used to drive a rover it would need parts that don't disappear when decoupled and reloaded, so that means probe cores. So I took the smallest probe core and made a therm-o hinge on one end and two 'spider' motors on the other.

I then placed one on to a stack decoupler and offset (download the no offset limit mod folks) it away. Then take another link and attach it to the decoupler and offset it in line with the first. This can be seen in the second video below.

 So each link is attached to the same decoupler. So when it releases all the individual links detach but remain together as therm-o hinges. This can also be clearly seen in the videos below.

I made a very simple rig to power the track using the rotating antennas. There was a lot of torque from the track and they could not be made 100% reliable in this configuration. I am currently testing a new rig that is completely stable and can run unaided using reaction wheels. The issue I have now is designing a reliable drive wheel..

It can become quite laggy as each link is 21 parts. And every single one of them is needed! This will most likely be a 1.1 rover drive.

 

Anyway talk, or text rather is cheap! I made a kind of Progress report video from recordings I made during the experimentation stage. Failures included!

I took a few pictures as well. Of the new WIP test rig and the links themselves. I have also made a video of me putting together a basic chain of the links. There is also a Sub-assembly download for the links.

 

 

Here is the development video

 

 

And the track construction video

 

 

What you see is just the test rig to develop all the parts that need to work. The track itself worked first time. The drive and the wheels not so much.. The end result will look very different from what you have seen here.

 Thank you all for looking and have a go with the link sub-assembly. Maybe someone will have another idea for possible applications.

MJ

 

Track link Sub-assembly: KerbalX DL

[RixKillian]

AHH! This is brilliant - You've no idea how many times I've tried to make tracks work. I'll be following this very closely! (And possibly trying my own variant if you don't mind )

[kmMango]

You're a wizard Jim. This is amazing! 

I bet zekes would love this, if he were still around. 

[katateochi]

errr, wow! That is very cunning! 

How are you able to connect all the links to the same node on the decoupler? 

After you've placed the first track, could you copy the group of two tracks and then connect them as a pair, then copy the group of 4 etc.....oh wait, they're not actually connected to each other are they? 

I'm going to have to have a play with this (& try n' fathom what ya done here).
I wonder if you could sort of coil the track up (or rather fold it like a stack), put a command chair on the last one and (somehow) anchor the other end to the top of a building and go bungee jumping!  I imagine there would be some flex in it.  
 

Edited January 15, 2016 by katateochi
typo

[Majorjim!]

Just now, RixKillian said:

AHH! This is brilliant - You've no idea how many times I've tried to make tracks work. I'll be following this very closely! (And possibly trying my own variant if you don't mind )

Please do!  I will most likely need help at some point! There are quite a few issues to overcome. I don't mind at all! Let me know if get stuck with anything. You have no idea how long it took just to get it to run a few cycles without self destructing..

 

Just now, kmMango said:

You're a wizard Jim. This is amazing! 

Cheers mate, I appreciate that.

Just now, katateochi said:

errr, wow! That's is very cunning! 

How are you able to connect all the links to the same node on the decoupler? 

After you've placed the first track, could you copy the group of two tracks and then connect them as a pair, then copy the group of 4 etc.....oh wait, they're not actually connected to each other are they? 

I'm going to have to have a play with this (& try n' fathom what ya done here).
I wonder if you could sort of coil the track up (or rather fold it like a stack), put a command chair on the last one and (somehow) anchor the other end to the top of a building and go bungee jumping!  I imagine there would be some flex in it.  
 

You need to have the Ignore restrictions box ticked in the f12 menu. Then you can attach as many things as you like to it. There is always an available node on it. No each link is a separate craft. That means it should run a lot better in 1.1. Hopefully.

Yeah fiddle around with the subassembly and see what you can come up with! No offset mod needed. And thanks for the kind words man.

[pTrevTrevs]

I fear you're tampering with a power beyond our comprehension. Should the Kraken arise to stop your experiments, I'm afraid the best we can do is pray for your soul. 

 

If you get this working though, you will have made a breakthrough that can change the game forever. Imagine it, proper tanks! I guess that with 1.1, the high part count won't matter too much. Maybe I can finally finish that WWI tanks project I started but never finished!

 

[Mad Rocket Scientist]

Incredible!  This makes my old chain design look like a flint knife compared to this.

[Dman979]

Hey, @Majorjim, you must have a KSP IQ of over 133, because this is genius!

I had tried with the radial decouplers, but it didn't work.

I think the wheels might be a problem in 1.1, though. I've read that the colliders in Unity 5 are different.

[kmMango]

I would try to build a tank, but I have a feeling that fiddling with this will anger the Kraken. Also, part count.

[sgt_flyer]

@Majorjim

have you tested it with only 1 powered wheel on one extremity, and a free spinning wheel on the other ? 

[selfish_meme]

Of course, if you don't care about reloading, you would not need the probe core, you have done it again Major

Edited January 15, 2016 by selfish_meme

[Majorjim!]

42 minutes ago, pTrevTrevs said:

I fear you're tampering with a power beyond our comprehension. Should the Kraken arise to stop your experiments, I'm afraid the best we can do is pray for your soul. 

 

Believe me the blighter is doing his utmost to foil me!

 

Just now, Mad Rocket Scientist said:

Incredible!  This makes my old chain design look like a flint knife compared to this.

Many hours where spent.. Not all of them enjoyable!

 

Just now, Dman979 said:

Hey, @Majorjim, you must have a KSP IQ of over 133, because this is genius!

I had tried with the radial decouplers, but it didn't work.

I think the wheels might be a problem in 1.1, though. I've read that the colliders in Unity 5 are different.

Yeah I am a bit worried about changes killing it. I think its more likely that the update will help though.

 

Just now, kmMango said:

I would try to build a tank, but I have a feeling that fiddling with this will anger the Kraken. Also, part count.

Yeah my test driving rig with a set of tracks and mechanism is over 700 parts. If 1.1 does no good to the performance, instead of letting it die I would weld the links. Each one is 21 parts.

 

Just now, selfish_meme said:

Of course, if you don't care about reloading, you would not need the probe core, you have done it again Major

Yup! There are quite a few ways you could make these. Some may indeed be better than others. Thanks man. I am hoping someone will have a go too.

Edited January 15, 2016 by Majorjim

[Bubbadevlin]

Nice! Cant wait to see an actual tank using these! (although good luck with the part count)

This is the first time I have seen someone using a NBSD on a craft too, and yea, they are really tricky and NOT for precision parts

 

Edited January 15, 2016 by Bubbadevlin

[Majorjim!]

56 minutes ago, sgt_flyer said:

@Majorjim

have you tested it with only 1 powered wheel on one extremity, and a free spinning wheel on the other ? 

Yup, my newest test rig, in the pics, is powered by the front wheel only. It has a reaction wheel inside that can be trimmed to spin unaided.

 

Just now, Bubbadevlin said:

Nice! Cant wait to see an actual tank using these! (although good luck with the part count)

This is the first time I have seen someone using a NBSD on a craft too, and yea, they are really tricky and NOT for precision parts

 

Me too! It may take some time, and I don't even know if they will hold up.. It's fun trying though!

[Azimech]

Not the first but surely the first stock, very impressive indeed! I gave up when the chain started to fly through the colliders of the guide wheels, fell on the ground and continued to move like a psychotic snake eating itself - probably need to blame Infernal Robotics (sorry guys :-P )

So I'm glad the concept still has life! Have you posted it on Reddit yet?

[Azimech]

7 hours ago, pTrevTrevs said:

I fear you're tampering with a power beyond our comprehension. Should the Kraken arise to stop your experiments, I'm afraid the best we can do is pray for your soul. 

 

If you get this working though, you will have made a breakthrough that can change the game forever. Imagine it, proper tanks! I guess that with 1.1, the high part count won't matter too much. Maybe I can finally finish that WWI tanks project I started but never finished!

 

I'd use it as a timing belt for camshafts. There's a new engineering revolution coming some day :-)

[Majorjim!]

6 hours ago, Azimech said:

Not the first but surely the first stock, very impressive indeed! I gave up when the chain started to fly through the colliders of the guide wheels, fell on the ground and continued to move like a psychotic snake eating itself - probably need to blame Infernal Robotics (sorry guys :-P )

So I'm glad the concept still has life! Have you posted it on Reddit yet?

I will never post anything to that place again. And thanks man. I have seen tank track mods but none of them are made from actual separate links.

5 hours ago, Azimech said:

I'd use it as a timing belt for camshafts. There's a new engineering revolution coming some day :-)

I had that exact thought! If 1.1 does not deliver I will weld the links. They are 21 parts each. That will make them a lot easier to use in game.

[Castille7]

Congratulations on the progress so far MJ it is definitely impressive work. Also I have never ventured off to other forums such as Reddit because it's enough to post and reply on the KSP Forums already, I do this for enjoyment and don't want it to become work. KSP is my relaxing playground to hang out and have fun while learning from others. I'm guessing there is some good feed back coming from Reddit but I am ok with staying just on here. Again really nice tracks!

[Majorjim!]

10 hours ago, castille7 said:

Congratulations on the progress so far MJ it is definitely impressive work. Also I have never ventured off to other forums such as Reddit because it's enough to post and reply on the KSP Forums already, I do this for enjoyment and don't want it to become work. KSP is my relaxing playground to hang out and have fun while learning from others. I'm guessing there is some good feed back coming from Reddit but I am ok with staying just on here. Again really nice tracks!

Thank you Castille7!

 Here is a pic of a set of tracks ready for the drive mechanism. 721 parts...

 

[Rune]

Just... no. This will never work. I don't know what kind of black magic you guys are expecting out of 1.1, but in my best days, I am afraid to hope for more than a 50% increase in the number of onscreen part before lag kicks in, and then only if they are more than one vessel. This will never run smoothly, because KSP is not a game about complex mechanical simulations, and therefore terribly ill-suited to simulated something like this.

And even granting you could make it run somehow? It will work horribly, making the same work as any other wheel, only worse and in a gazillion more parts. If I want something to look like this, I would install a track mod and solve it with a single extra part, a proper animated tank track.

Now, that said: I do appreciate the work involved, and it is certainly impressive to see. But I feel obligated to tell you that you are running smack into a dead end ÍMHO...

 

Rune. No one can deny you can make the KSP build editor sing arias, though, MJ.

[Dafni]

I agree with Rune, but it is very very impressive nonetheless.

I feel like just the video alone is making my computer suffer

[Redshift OTF]

Wow, Major Jim, this is impressive stuff! I love it when people push the boundaries of what this game can do.

I hate to be that guy again but...I do have my own idea on how to make tank tracks that might work. I had thought about it since I first came across the infinite offset mod and a quick test seems to yield results. It doesn't look as cool as yours though.

[RixKillian]

7 hours ago, Rune said:

And even granting you could make it run somehow? It will work horribly, making the same work as any other wheel, only worse and in a gazillion more parts. If I want something to look like this, I would install a track mod and solve it with a single extra part, a proper animated tank track.

I believe the point of this is simply to make it work, regardless of how well it will function. Of course there are better ways to move around on the ground, just as there are more efficient ways of air travel than Azimech's turboprops, but for those of us who are tinkerers, nothing beats taking an extraordinary problem and solving it - especially with what we have on hand.

[Rune]

1 hour ago, RixKillian said:

I believe the point of this is simply to make it work, regardless of how well it will function. Of course there are better ways to move around on the ground, just as there are more efficient ways of air travel than Azimech's turboprops, but for those of us who are tinkerers, nothing beats taking an extraordinary problem and solving it - especially with what we have on hand.

By all means, tinker away to your heart's content. I'm just warning you that the magical day when KSP can run 3k part ship without lag is probably never going to come, certainly not in 1.1. "Substantial improvements" doesn't mean 3000% increase in performance, it most likely means something in the range of ~50%, which would turn ~300 part builds once again playable (to some extent).

 

Rune. Right now the lag spike is at around 100 for me...

[Majorjim!]

8 hours ago, Rune said:

Just... no. This will never work. I don't know what kind of black magic you guys are expecting out of 1.1, but in my best days, I am afraid to hope for more than a 50% increase in the number of onscreen part before lag kicks in, and then only if they are more than one vessel. This will never run smoothly, because KSP is not a game about complex mechanical simulations, and therefore terribly ill-suited to simulated something like this.

And even granting you could make it run somehow? It will work horribly, making the same work as any other wheel, only worse and in a gazillion more parts. If I want something to look like this, I would install a track mod and solve it with a single extra part, a proper animated tank track.

Now, that said: I do appreciate the work involved, and it is certainly impressive to see. But I feel obligated to tell you that you are running smack into a dead end ÍMHO...

 

Rune. No one can deny you can make the KSP build editor sing arias, though, MJ.

Well, certainly not with that attitude!

 I cannot begin to describe the feeling of seeing this work for the first time. It was comparable to my first Mun landing. The fact that nobody has ever done this before in KPS just added to the excitement.

 To be honest I was amazed that it worked at all. And it does work. Exactly as I intended it to.

The final working vehicle will be under 1000 parts. If 1.1 performance is like 0.90 1000 parts will be playable. For me. And as I have stated many times, I make craft for myself first. If I know it's cool I share it with others.

The whole point to this was, 'can it be done'. It was a test of my KSP building knowledge and skill. I like to test myself and come up with new ideas and new solutions. This machine is as Kerbal as it gets!

6 hours ago, Redshift OTF said:

Wow, Major Jim, this is impressive stuff! I love it when people push the boundaries of what this game can do.

I hate to be that guy again but...I do have my own idea on how to make tank tracks that might work. I had thought about it since I first came across the infinite offset mod and a quick test seems to yield results. It doesn't look as cool as yours though.

Thanks man, I would love to see what you came up with! Care to share a pic or two?

1 hour ago, RixKillian said:

I believe the point of this is simply to make it work, regardless of how well it will function. Of course there are better ways to move around on the ground, just as there are more efficient ways of air travel than Azimech's turboprops, but for those of us who are tinkerers, nothing beats taking an extraordinary problem and solving it - especially with what we have on hand.

This man gets it. Cheers mate. 

Edited January 16, 2016 by Majorjim