how to create an "unflippable" rover

[creator1629]

hello, i have tried many rover designs on the moon, but the problem is when achieving max speeds, hitting a small bump or irregularity will cause it to flip or bounce and crash. i was wondering if you guys had any design tips that would let me build a rover capable of maintaining high speeds. i have tried to apply general principles such as making a wide rover, keeping it relatively flat, making it have a low center of gravity, also i have tried adding girders sticking out the front/back/sides to prevent it rolling over. any ideas you guys may have, or even already built yourselves? basically id like a design where i can just hold down the forward button and not worry about it too much

[Klajan]

I have a very very stable design, even capable to flip itself on the weels if flippfed. the center of gravity is very low, and so is its profile. I can drive 30 m/s without having to worry about flippng, as long as i don´t make sharp turns (Timewarp at 4x works as well, but never steer while doing that, nothing can save you from flipping while doing that.). It is very sturdy, surviving 50 m dropps without brakeing the weels.

The rover itself is very lightweight, around 2-3 tons and has a SAS-Unit for enhanced stability (yes that works) It also has RCS thrusters to help with flipping it back.

[Klajan]

Here is my design, currently exploring Mun, Minmus and Eve:

[Tex_NL]

An "unflippable" rover does not exist.

Giving it a wide wheelbase and a low CoG is your best option. Next thing you might want to consider is re-mapping your controls. Regular rover controls are mapped to WASD but this also controls you your gyros. Re-mapping rover controls to IJKL enables you to controls steering both the rover and the gyros independently.

In case your rover does go turtle, RCS and/or landing legs like those in Klajan's design get you back on your wheels with ease.

[Tw1]

Low Cog is essential. Keep the mass down low, and you will have less problems. Heavier vehicles are harder to flip as long as the follow those rules.

You seem to have the basic idea, but one more rule you must keep in mind is positioning wheels around the CoG. Otherwise, your vehicle will be unbalanced, and tilt when accelerating or decelerating . I learnt this the hard way, building with the old cart mod.

I tend to start my vehicles based on this principle, positioning parts accordingly.

I also use the aircraft wheels to help the vehicle take impacts.

I tend to make large to medium sized vehicles.

ASAS can help, if you want it to keep pointing in the same direction. Also, if you haven't already, remap the rover wheel controls to IJKL, rather than WASD, and you won't get torque from the comand pod/ probe core interfering.

Edited May 16, 2013 by Tw1

[capi3101]

I started my own initial forays into rovery last night. So far I've got a simple rover - a mini-probe chassis stuck onto a rovemate, four type-I wheels, 4 PB-NUKs and a headlight. She'll do 20 on Kerbin, faster on the downhills and has enough juice to go for a while before needing to let off and recharge. Tried adding a stack separator to release and act as a navigational target to tell how far from KSC I am but the damn thing more often than not explodes as soon as it hits the ground.

Rolled the stupid thing over trying to drive it off the launchpad first thing. Tried adding a lander leg to right it again; that didn't work. I think tonight I might try an Oscar/24-77 combo; I remember that the first few times I successfully landed on the Mun and my lander tipped over, it was ample RCS thrust that allowed me to upright it again.

Anyway, I too am interested in this topic; might have to compare notes.

[Colonel_Panic]

That's it, I'm making a rover that can drive upside-down.

[creator1629]

thanks guys these tips have been very helpful. what im concerned about is the low gravity of the mun which means a small bump on kerbin wont knock you up very high but on the mun it might just be enough to flip you over. also another concern i have is that i am sending many unmanned rovers to multiple destinations, and so while a a manned rover can just have the kerbal come out and fix the tire if it gets damaged, an unmanned rover does not have that option, and multiple tire failure crash could effectively put it out of commission

[Klajan]

Yes, but if you want to keep it "realistic" you would also have to get the Kerbal back to Kerbin, while an unmaned Rover des not have such worries. A slong as oyu are on fairly low gravity worlds it will be hard to blow your tires. I have driven over 150 km in my rover on the Mun on 4x Physical Timewarp and have not blown a single tire.

[Khrissetti]

Just make sure it can run upside down like a 'Rollcage' car.

[nhnifong]

How could it be that hard? Just put wheels on the top

[capi3101]

Some updates on my forays with rovers...and a chance to see if I can get some images posted again.

First of all, the 24-77 idea worked, but you can't give it too much juice. Ultimately used a Round-8 instead of an Oscar; it was easier to sling under the chassis.

Having tested the design on Kerbin, the next step was to get some rovers to the Mun. I repurposed an early Duna lander for the mission: the Muad'Keeb 6 (sometimes known as the Phallus 7; never stick a chute on top of a PB-NUK). To do that, I slapped on a trio of BZ-52s, added some TR-2V stack seperators, and stuck the rovers onto them. Probably not a good idea to dangle a rover off the side of a rocket by a nuclear power generator, but it worked.

The deployment system leaves a little to be desired; first time I tried, I released all three rovers at once and they universally landed upside down. Not a lot of room under the lander for getting them upright either. Fortunately I had the foresight to quicksave before letting them go. I blew up all three driving around the surface. Next go I learned my lesson and released them one at a time, uprighting them as necessary.

The last one actually landed upright. Couldn't believe it: took six deployments before one landed on its wheels.

No quips on using that lander for a Mun landing; turned out to be a good thing I had that much thrust available to me. I overflew a Mun arch on my way to my intended landing site and decided to land there instead. Probably means the lander isn't coming back up ever.

Next thing I need to learn is how to drive on the Mun, apparently; damn things have a tendency to "drift" a bit. I know I can't be the first noob to ask about driving; anybody want to save me the searching time and just point me to the correct thread?

[MBobrik]

I've managed it. Creating rover so stable that it can drive on two wheels w/o flipping. Almost unflippable. I had actually hard time to flip it intentionally.

The secret ? Make it so tiny that probe SAS torque is stronger than the torque from the wheels.

now the same with a bigger can

[MBobrik]

Done. This one can make full turn at its top speed. 3 x SAS and wheels as close to COM as possible seem to do the trick. Now comes the Minmus test.

.

[DerekL1963]

Next thing you might want to consider is re-mapping your controls. Regular rover controls are mapped to WASD but this also controls you your gyros. Re-mapping rover controls to IJKL enables you to controls steering both the rover and the gyros independently.

You don't need to remap - just drive your rover from the docking screen rather than the staging screen. Swap back to the staging screen if you need gyro torque to right your vehicle.

Klajan, do you have a .craft file for that rover?

[Jokurr]

One thing you can also do is use RCS thrusters to push your rover down onto the ground if you think your rover is about to become airborne. This also is great on low-gravity worlds as it allows you to get more traction and accelerate much faster.

[Dunbaratu]

Landing legs on the roof, positioned upside down. can act as levers to try to flip the rover over. Test your design on Kerbin on the lauchpad first, before attaching the rest of the rocket stages to it. I can often deliberately cause a rover to fall on its roof by driving it off the launchpad very slowly so it falls off the edge of the platform and rotates as it does so. Then once on the roof you can test your self-righting mechanism.

[Ryu Gemini]

One way you could go for it would be to have three sets of wheels instead of two, each at a 120 degree angle from the other two. Or if you can afford a bit more weight, four sets of wheels at 90 degrees to each other.

It wouldn't be so much that it can drive upside down, as it would be that it has 3 or 4 upright positions.

.

[joppiesaus]

low center of mass, and this:

in crease the width of the rover will help