Why is differential steering with 2 passive (non caster wheel) not used?


I'm curious why having a robot that uses differential drive and 2 passive non caster wheels not being used? The most common I see are differential steering with casters and skid steering.

What I'm asking about is something that looks like skid steering but only have one wheel on each side being driven with a motor.

Are you certain about "not being used"?

No I'm not. When I looked at articles discussing different drive systems, it wasn't included. I guess what I'm trying to say is it's not popular.

If it incurs skid friction to turn, that's a good reason not to use it. The more weight you put on it, the harder it will be to turn. Also it would be heavily dependent on a relatively smooth surface in order to work properly.

Wouldn't it end up behaving like the skid steering anyway? It's just that they rotate freely instead of being powered.

I'll explain my issue and let me know what you think.

Right now, I have the rover on differential FWD drive with 2 rear castor wheels. It's a big rover that's suppose to carry a heavy load (100kg) and this has been accounted for by the motor and the design. This has been achieved nicely indoors. The issues I'm facing are these:

  1. The front wheels (driven) are air pumped and offer some absorption/suspension, casters are rubber and don't offer that. So this rougher terrain it gets all bumpy and shaky.
  2. Since the casters are not the same size as the driven wheels, I have them at different levels, it's extremely hard to assembly them properly, so at a given time, one of the casters doesn't have good contact with the group and will spin randomly. This is really obvious with rough terrain, but also sometimes on a smooth flat surface (without the load onto it)
  3. Given that all wheels aren't connected on the same height on the chassis, I end up not using the available space of the rover for components and I'm stuck with only have the chassis, since the casters are quite big. I'll attach a picture of the current rover.

So by having two passive wheels, I'll have the suspension from the air pumped tires, they'll all be on the same level - increasing the compartment space and allowing better traction with the ground.

The reason I'm trying to avoid the skid steering mechanism is it needs more space for the chain/gears on each side. Heavier load, and more things that could mechanically fail.

Sorry, I thought you said, without casters... if not, my comments in reply #4 still hold, I believe. You have no comment on that?

Try pushing your existing driven wheels sideways and you might have second thoughts...

Yes, I'm asking to change the design to be without casters. It's currently with casters, but they're causing a lot of issues.

But wouldn't your comment still hold on skid steered rovers too? (ones where each side (2 wheels) is connected to a single motor? But these are very popular apparently.

You are avoiding the key question - how to deal with the extreme friction forces when any wheel is forced to move laterally. That's been an issue with wheels ever since Ogg.

I'm aware of your question. But I'm not sure how bad it would be. Like obviously if these aren't wheels and just fixed points it would be bad, but they can still rotate, even if it's in 1 dimension. How bad is it, I'm not sure. There's definitely going to be friction.

My question is, would this friction be the same assuming both wheels on each side are connected vs 2 wheels driven and 2 wheels driving? If yes, then the follow-up question is how come skid-steering is very popular. For example the jackal by clearpath.

Oh, so you want to have a 4 wheel drive, but with both wheels on each side driven together, for example with a chain? Like a tracked vehicle, but without tracks? I think this has a much better chance of success than 2 driven wheels and 2 idlers.

Skid steering is probably popular because it's easy. That doesn't make it good.

Yes I believe that is what is labelled as skid steering.

Would skid steering differ from having the same setup but with only 2 driving and 2 passive wheels (not caster)? Friction wise? Why would it be better?

You have to make some force diagrams. But consider a simple difference - the distance and direction the idler wheels have to travel for a given rotation, say 90 degrees.

That doesn't really answer my question, if one is better than the other. I guess I'm asking if you have experience with them or have seen them before?

I don't need to. I've been playing with mechanical things all my life. Okay, I'll answer your question. Skid steering would definitely be better than your proposal.

Can you explain how it would differ please? Like what makes the skid steering better, wouldn't they both behave the same since they spin in the same direction? I apologise if I'm asking too primitive questions.

I agree with @aarg that 4WD/skid steering will very likely outperform the pictured setup. However, no amount of forum blathering will give you good insight into how it will actually perform.

You will have to experiment with your robot, which looks heavy. And, the wheel base seems long, so the drive wheels will lose traction while turning, especially if the casters encounter obstacles.

Thanks for the reply. Although that was a helpful comment, that was not my question. My question is would skid steering (2 motors connected to 4 wheels) be better than 2 motors connected to two wheels and 2 idle (non caster) wheels?

I'm aware of the limitations of my current setup and that's why I'd like to change it. It's really good for turning but as you said, there's issue with the traction and vibrations.

To clarify.
Please check the picture attached. I currently have model A, I want to shift to Model C. However, I'm not seeing that this is a popular option. Model B is popular. How bad is model C compared to Model B.

so Model B vs Model C.

M denotes the motor

Model C is much worse than A, the pictured setup, so B.

That's what I'm trying to find out. Why is it much worse?