They See Me Rolling: A Rover Discussion

[Pezzhippo]

Greeting ladies and gents! I'm currently musing over the creation of some parts for the construction of medium-large rovers, but beforehand, I felt it necessary to do a little research and gather the insight of my fellow kerbonauts. Thus, I would like to posit a few questions regarding your personal criteria when designing your own aforementioned medium or large rovers, and get the ball rolling on what I hope to be an educational discussion on the subject.

Moving on, I would like to split this topic into two major parts; One concerning rovers for use on worlds with moderate gravity, eg. Duna, and another concerning rovers for use on low-gravity bodies, such as the Mun, or Minmus.

So, how would you folks go about designing rovers for these environments? For example, what kind of footprint do your rovers generally posses, how do you cope with various inclines or rough terrain, and how do you go about designing something for a given task, such as Kethane mining, or moving other objects over the surface of a given celestial body? Go nuts, be as detailed as you like, I most certainly look forward to your replies! c:

[CosmicScientist]

Whilst I never go overly big, usually either enough for science equipment and a sight-seeing tour through the cameras of a rover or just enough for a buggy that isn't going to weigh me down on the launchpad because I'm still just trying to explore and plant a flag on almost all celestial bodies in the system. I find that planning rovers for Gilly-like conditions is best in any situation so you'll have little as possible chance for it to tip, make sure its base is wide and possibly long, rotate the wheels to almost 45 degrees (so shift+eight twists to 40 degrees as it doesn't register wheels at 45 or beyond I find) to make the profile as low as you can without causing anything on the bottom to hit the ground.

When I'm on Minmus, I take advantage of its superbly flat and frozen ground and put on the airplane wheels that have no maximum speed and let ion engines take me to where I want to go if I need to go quickly (getting to orbit this way is certainly fun).

For rough terrain, I'm still puzzling how to make something to my aesthetic taste but if you take the structural girders and plonk wheels on them, you'll find they can give better suspension when on places like the Mün because they bend to the landscape from where they're joined to the body of your rover, they probably will do worse and worse when you use them on large enough rigs such as Kethane would likely bring. Then if you're looking to go up a hill, the Kerbal Attachment System mod will likely give something to get up with if you're struggling to get a Kethane miner up although I wouldn't rely on it, moreso something to lift you up and over the obstacle as you can always mine more rocket fuel with Kethane to replace the amount spent, right?

[Blue]

I've had trouble making large rovers on Minmus. It would either have to do with the large wheel motors accounting for low gravity with respect to momentum, or low surface friction; either way, large rovers move extremely slowly (<5m/s) on Minmus and require landing gear to make stationary, as their brakes are unreliable. Quicksaving while just sitting there at 0.1m/s will make the game detect the rover is "about to crash" and won't save.

Moho is a dream to drive, though. It has smoother features than the Mun, and slightly more gravity, ensuring much more Kerbin-like speeds and control behaviours. The downside is that travelling 12m/s (downhill) or faster means it is nearly impossible to stop without uncontrollably flipping over until you reach a flat area. I've yet to place rovers on Duna, Eve or Dres, so we'll see how those go.

[Levelord]

I made a dune bike for a Duna mission recently. It came with an embedded SAS to prevent flipping and can withstand bumps at speeds up to 50m/s. It's the fastest rover I've built so far, and it covers quite a lot of terrain if you travel in a straight line.

Basically I've added some decouplers to hold up the wheels as extra suspension, which makes the rover handle a lot better than if it were attached to a rigid surface. The extra bounce also allows it to withstand higher drops.

Edited August 24, 2013 by Levelord

[Pezzhippo]

So, from what I can gather here so far, is that the way to go with a rover is a good solid wheelbase of a low-profile nature, with some flexibility added to the mix to help compensate for terrain. This latter point, though, does highlight an issue I've been finding with the stock rover wheels, in that their suspension leaves a little to be desired. The newer Ruggedised model in particular, I found, runs in a near-constant bottomed-out state for example.

Blue's post regarding Minmus is extremely informative, and I've added a means of anchoring a rover to the surface either by landing gear, or more preferably, low-profile lander legs to my list of things to consider. I'd actually also like to add another point of consideration to this thread in asking: What would, in your opinions, make your lives easier when it comes to designing or operating rovers and why? And indeed, suspension or terrain handling is evidently a key point across the board, but is there anything else absolutely crucial to your general design processes?

[Levelord]

So, from what I can gather here so far, is that the way to go with a rover is a good solid wheelbase of a low-profile nature, with some flexibility added to the mix to help compensate for terrain. This latter point, though, does highlight an issue I've been finding with the stock rover wheels, in that their suspension leaves a little to be desired. The newer Ruggedised model in particular, I found, runs in a near-constant bottomed-out state for example.

Blue's post regarding Minmus is extremely informative, and I've added a means of anchoring a rover to the surface either by landing gear, or more preferably, low-profile lander legs to my list of things to consider. I'd actually also like to add another point of consideration to this thread in asking: What would, in your opinions, make your lives easier when it comes to designing or operating rovers and why? And indeed, suspension or terrain handling is evidently a key point across the board, but is there anything else absolutely crucial to your general design processes?

Well for me, absolutely all my rovers must come equipped with an SAS module. The massive torque their reaction wheels provide allows me to flip a rover over in case it lands upside down. The SAS module also prevents your rovers from flipping over at high speeds when activated (providing you are traveling in a straight line).

Edit: old-ish thread I made about the SAS system in rovers. http://forum.kerbalspaceprogram.com/showthread.php/42966-Strengths-of-the-reaction-wheel-system

Edited August 24, 2013 by Levelord

[Brotoro]

I've never built really large rovers, so my largest ones are probably in the "medium" class. But I've done a LOT of roving with them.

The rover below (or close variations of it) have been used on Eve, Duna, Laythe, Mün, and Minmus. I used one of these to drive completely around Duna. Low CoM, wide wheel base, and ASAS for stability when it "catches air," plus lots of wheels to distribute the stress so blowing wheels is minimized, are the features that make it successful. It works best in higher gravity, and used RCS on Minmus for those exciting "airborne" moments. (That version in back was a folding model that used Damned Robotics hinges, but it was less successful because the hinges made the frame weak.)

The BirdDog rover below is a combination airplane/rover that has driven extensively on Laythe (and Laythe has some very difficult terrain). It's a small plane so it can get from island to island on Laythe, and can land/take off in surprisingly small areas. When its nose gear is raised, the two front rover wheels are lowered to the ground to put the ship in rover mode. It can handle steep slopes approached at an angle (you have to learn to drive the contours of the landscape), and even handles well at 2x or 3x time warp on moderate terrain. The things that are most dangerous to it (and many rovers) are hitting v-bottom valley at speed, or trying to run along the top of a sharp ridge (where it can bottom out and rip off the jet engine). There have been a couple times where I've accidentally (or purposefully) taken it over a sharp drop off, and it was able to survive because it could glide down to a safe landing. The aircraft landing gear are very resistant to damage, and possibly the ship's mostly-triangular wheel layout makes it handle rough terrain well...but I haven't tried that configuration on airless bodies yet...mainly because the landing gear weigh a LOT compared to rover wheels (although it's not clear if KSP actually counts the mass of the landing gear in flight).

Edited August 24, 2013 by Brotoro

[Commander Zoom]

I made a dune bike for a Duna mission recently.

*drools* .craaaaaaaaft

[Brotoro]

This thread inspired me to play with some rovers... trying to make something suitable for Vall. For the first time ever, I tried out those ruggedized medium rover wheels... YEOW! Those things have a deadly grip! How do you guys use them? I tried driving around and the rover tries to flip violently on braking... and even at 2x time warp the thing wiggled out of control and flipped. Crazy wheels.

[Spartwo]

This thread inspired me to play with some rovers... trying to make something suitable for Vall. For the first time ever, I tried out those ruggedized medium rover wheels... YEOW! Those things have a deadly grip! How do you guys use them? I tried driving around and the rover tries to flip violently on braking... and even at 2x time warp the thing wiggled out of control and flipped. Crazy wheels.

Enough torque to balance on one wheel

[Pixel of Life]

Here's a tracked design I built yesterday. IIRC it weighs around 600 kgs. RCS is for flipping the rover back upright if something goes wrong - assuming the thing is still in one piece. The skycrane can easily be replaced with parachutes. Uses B9 Aerospace for the lights and Rubber Band Inc. Caterpillar Tracks for the cat tracks.

This works very well on planets and moons with stronger gravity. Due to the physics engine's limitations (the lack of a proper grip model, or the planets' grip values are set too low) some places like Pol and Gilly are out of question - despite the huge surface area of those tracks, the rover will slide all over the place if it's not on a perfectly level surface.

Edited August 25, 2013 by CaptainKorhonen

[Epthelyn]

Due to the physics engine's limitations (the lack of a proper grip model, or the planets' grip values are set too low) some places like Pol and Gilly are out of question - despite the huge surface area of those tracks, the rover will slide all over the place if it's not on a perfectly level surface.

Grip requires gravity to bring the two surfaces together, or some other downwards force. Unless you're using RCS or downwards facing engines to push yourself into the surface, only the heaviest of rovers will ever be able to drive on the lower gravity moons. It's not a engine limitation, it's physics.

[Levelord]

Grip requires gravity to bring the two surfaces together, or some other downwards force. Unless you're using RCS or downwards facing engines to push yourself into the surface, only the heaviest of rovers will ever be able to drive on the lower gravity moons. It's not a engine limitation, it's physics.

It's actually both. Because even on Kerbin and it's high gravity, the rover wheels lack enough traction to drive a corner at a reasonable rate. The rover wheels lack enough suspension and the game engine doesn't track the collison mesh of the ground accurately enough. Resulting in wheels unexpectedly buckling due to phantom forces.

[leax256]

modular wheels seem to have a grip of some sort, in fact if you put enough of their engines you can generate enough force to put the rover into a wheelie.

[Sufficient Anonymity]

On bodies without an atmosphere, I tend to send very heavy and squat rovers. However, I tend to concentrate on bodies with an atmosphere, and thus tend to send airships to avoid the issues of rough terrain, phantom forces and inadequate grip. For example, below is the colony ship that I'll be sending to Laythe shortly - sat atop the drive stage is a habitation block, then the airship re-entry heat shield, then the airship itself. Due to electric engines and RTGs, it has infinite range.

I'm still trying to get such a system to work well on Duna, however, as the atmosphere is very thin, and (as is probably obvious) I like sending heavy vessels.

[Max Grant]

I made a dune bike for a Duna mission recently. It came with an embedded SAS to prevent flipping and can withstand bumps at speeds up to 50m/s. It's the fastest rover I've built so far, and it covers quite a lot of terrain if you travel in a straight line.

Basically I've added some decouplers to hold up the wheels as extra suspension, which makes the rover handle a lot better than if it were attached to a rigid surface. The extra bounce also allows it to withstand higher drops.

Bruce Wayne Kerman Enterprises is interested in your prototype . . . .

[Brofessional]

My rovers tend to come out looking like dune buggies, with RCS to help with getting up and down slopes.

[NGTOne]

Personally, I subscribe to a philosophy of "Go big or go home" - my "rovers" are more along the lines of rolling bases. I've found the track mod to be invaluable for this, as the tracks are far more resilient than wheels, and can take a lot more weight. As an example:

This vehicle clocks in at about 650 tons. Landed on Eve without a hitch using airship parts (parachutes didn't work well because they would cut too early, when 90% of the vehicle's mass wasn't on the ground yet, due to its large size and the terrain not being ideally flat). I haven't tried sending it to other bodies yet, but its sheer mass would, I think, compensate for the grip issue. It actually has a problem on high-grav worlds, because the track pods have SO MUCH grip that trying to move it across anything other than flat ground (or, God forbid, turn it) has a tendency to tear it apart due to slight differences in grip between the pods (and, because of the vehicle's size and mass, the forces involved are too large to compensate for, even with super-struts from mods). I really should send one of these to Gilly at some point, to try it there.

The differential grip problem was acute enough that trying to roll that behemoth off the back of the runway would occasionally tear it apart. Its successor design (never completed, clocked in over 1000 tons last time I worked on it) tended to fare better.

[Kinglet]

They see me rollin' They hatin'

[Radam]

They see me rollin' They hatin'

WOW Amazing

[Kinglet]

WOW Amazing

I don't like it so much, I will built more soon.

[Wayfare]

This picture shows two different manned rovers from my MOdular Mission System (MOMS):

The big lug is the Munophant. It's fast, fairly stable and I have the rear wheel brakes bound to Action Group 1 so it won't flip ass-up (or even ass-over) when I need to stop in a hurry. Plus it has that lovely high-visibility Panopticon for a cabin, which means it's actually pretty fun to drive from IVA view.

In the foreground is a simple two-man buggy, originally designed for my Munshine V Mun rocket. I love those little things - they're really stable, they consist of just thirteen parts and they weigh under half a ton. You can bolt one onto the side of pretty much any lander, balance it out with either another buggy or an RCS tank and Bob's your uncle. They won't top 7m/s and they don't like climbing hills very much but hey

[Tiron]

It's actually both. Because even on Kerbin and it's high gravity, the rover wheels lack enough traction to drive a corner at a reasonable rate. The rover wheels lack enough suspension and the game engine doesn't track the collison mesh of the ground accurately enough. Resulting in wheels unexpectedly buckling due to phantom forces.

The Rovemax 1s lack enough traction to take a corner reasonably well. The TR-2Ls have too MUCH traction. I've personally never had problems with the suspension being inadequate. The suspension+ground mesh problem, however, has basically stopped my gameplay, because I got tired of it interfering with my rovers and I refuse to do anything else until I've found all the anomalies on Kerbin...which isn't going to happen unless the rovers get fixed.

[Brapness]

I have the rear wheel brakes bound to Action Group 1 so it won't flip ass-up (or even ass-over) when I need to stop in a hurry.

Why didn't I think of that? This would have saved a lot of my rovers.

[RickRastardly]

Personally, I subscribe to a philosophy of "Go big or go home" - my "rovers" are more along the lines of rolling bases. I've found the track mod to be invaluable for this, as the tracks are far more resilient than wheels, and can take a lot more weight. As an example:

This vehicle clocks in at about 650 tons. Landed on Eve without a hitch using airship parts (parachutes didn't work well because they would cut too early, when 90% of the vehicle's mass wasn't on the ground yet, due to its large size and the terrain not being ideally flat). I haven't tried sending it to other bodies yet, but its sheer mass would, I think, compensate for the grip issue. It actually has a problem on high-grav worlds, because the track pods have SO MUCH grip that trying to move it across anything other than flat ground (or, God forbid, turn it) has a tendency to tear it apart due to slight differences in grip between the pods (and, because of the vehicle's size and mass, the forces involved are too large to compensate for, even with super-struts from mods). I really should send one of these to Gilly at some point, to try it there.

The differential grip problem was acute enough that trying to roll that behemoth off the back of the runway would occasionally tear it apart. Its successor design (never completed, clocked in over 1000 tons last time I worked on it) tended to fare better.

How, in gods name did you get that into orbit around kerbin!