Hi! I just wanna know why Skid turn Car is more popular in Robotics than the one that can turn it's front wheels?(don't know what to call it). I thought the one that can turn front wheels is cheaper, can use only one motor, no need for H-Bridge,less programming??
Hi! I just wanna know why Skid turn Car is more popular in Robotics than the one that can turn it’s front wheels?(don’t know what to call it). I thought the one that can turn front wheels is cheaper, can use only one motor,
Mainly because they’re mechanically -drastically simpler and thus cheaper/easier. But you didn’t describe what you think you’d prefer? A conventional ‘car’ style? That still requires two motors, one to power the drive wheels, and one to direct the steering. The mechanical difficulty though is that a ‘car’ has a differential rear end to allow for the ‘outer wheel’ in a turn needing to travel further than the inner wheel.
no need for H-Bridge,
Required for reversing travel regardless of how many motors or drive/steer orientations. Otherwise once you’re stuck in a corner you’re stuck.
The skid with two drive wheels also has the advantage of turning in its own length.
The geometry of the skidder moving ‘forward’ -VS- moving in reverse reverse is DRASTICALLY drastically simpler than a conventional car format. Meaning the geometry of traveling ‘forward’ using front wheel steering -vs- the geometry of traveling backwards with rear wheels steering is completely different. So that requires two sets of programming depending on which way the vehicle of moving. The geometry of the skidder is little different whether the skid is either ‘in front’ or behind the drive wheels, but it’s still much simpler.
A front wheel drive - front wheel steer is geometrically simpler. But with that the ‘other’ wheels are simple dragging along anyway and not serving much purpose except to be a 'geometrical -burden- when traveling in reverse, so leave them off completely and just add the skid. Conventional cars have 4 wheels for stability and more payload capacity.
I recall a documentary about someone who built a rear wheel steering motorcycle, simply because it had never been done before. No one could drive it more than about 10’ before it tipped over. probably geometrical/balance/human reaction time issues. I’d bet a gyro-stabilised controller could do it, but that defeats the ‘cheaper/easier’ issues.
Tamulmol: Hi! I just wanna know why Skid turn Car is more popular in Robotics than the one that can turn it's front wheels?(don't know what to call it).
What you have termed "skid turn" is more properly called "differential steering", though you also see the more colloquial "skid steer" used as well (typically in relation to heavy equipment - ie, "skid steer loader"):
The other method is known as "Ackermann steering geometry":
Tamulmol: I thought the one that can turn front wheels is cheaper, can use only one motor, no need for H-Bridge,less programming??
I wouldn't say it is cheaper or easier, just different; you still need some kind of actuator to turn the front wheels to make the robot go in the desired direction. Invariably, this actuator is some kind of electric motor (although either pneumatics or hydraulics, among other methods, can be used). If using some kind of electric motor, you still need an h-bridge (even if using an RC servo - there is an h-bridge inside it). You also need some kind of feedback mechanism to know where the wheels are pointing (built into an RC servo, but other methods will need something custom if the actuator doesn't already have such a positional feedback mechanism). This may translate into more programming as well...
All in all - Ackermann steering can be more difficult to implement, both mechanically, as well as in software - for instance, to be able to get such a platform from one point to another, you have to know when and where along the path to angle the wheels, how much, and when to angle them back to keep advancing toward the goal location (and how to adjust if bumps, friction, or other problems cause the robot to deviate from the course). That's just one example of the extra software complexity involved.
Differential steering, on the other hand, is much easier to implement, and easier to program for; movement from point to point becomes a matter of vectors, which are fairly easy to visualize and code for. Of course, it makes for a more "jerky" movement - real differential steering is done so that the speed of the wheels is varied smoothly; this is especially important with real-world tracked vehicles (such as tanks and bulldozers, for instance) - because on certain terrain attempting high-speed central pivoting can actually cause the tracks to come off the wheels ("throwing a track"), because the rear and front of the track is being dragged sideways, pulling it off the idlers and such - which is why when such a maneuver is done, it is done fairly slowly, and only in soft ground to minimize this issue.
In the case of a multi-wheeled robot using differential steering, you don't have the issue of a track being thrown, but you do have the issue of the tire contact areas being "scrubbed" - literally the tread is ground off (particularly, again, on hard surfaces like asphalt and concrete), which can lead to other issues, least of all being the need to constantly put on new tires. There is also the issue (same as on tracked vehicles) of extra wear-and-tear on the wheel bearings, axles, drive system, etc - caused by the side-loading (this can be mitigated somewhat by using proper bearings in those areas).
Lastly - those two options aren't the only means of wheeled propulsion, but they are the two most popular, mainly because they can both be implemented using mostly or wholly off-the-shelf components. Other methods which you may run into (or you might want to investigate) for wheeled vehicles:
Both of these (and a few other methods) tend to be much more complex to implement, and don't typically lend themselves to off-the-shelf parts solutions (which tend to drive up the cost, sometimes greatly). But they should be kept in mind as potential alternatives, depending on what you are trying to accomplish.
Thanks for the replies! I understand now. I though the “car style” steering is cheaper because most R/Cs use it and super fast with just one motor compared to skid steering, still crawling with 4 motors but i havent seen a skid steering car with 4 brushless motors that rc use lol.
Although cost and complexity can play a role, I think that these are usually secondary concerns. Differential steering and Ackermann-like steering have intrinsic differences which usually make one or the other an obvious choice for a given application.
Differential steering is used primarily because it allows very tight turns. This is extremely important for an indoor robot, tanks, some kinds of construction vehicles, etc. Differential steering is also useful for tracked vehicles which can develop more traction longitudinally than laterally. Despite this, differential steering has severe disadvantages which make it a poor choice for almost everything else.
Think about cruising along in a car on a straight road. If you want to turn, this is done by rotating the front wheels. Unless you're going too fast for the corner, completing the turn without losing speed requires almost no energy. Ackermann steering is very efficient. It's also easy to change directions quickly and to travel in a straight line.
Turning at speed using differential steering is very different. It involves using the motor(s) to generate very large forces on the two sides of the car. It also skids significantly. This is wasteful, wears things out, and requires powerful motors. It can also be hard to drive in a straight line or to quickly avoid an obstacle. There are fundamental reasons that differential steering just isn't a good idea for a normal car (at least as the primary way to control direction. Modern stability control systems do use something like differential steering for small corrections at the performance limit. This is usually done by braking individual wheels, and is very complicated to do well.).