Mechanics and Construction > Mechanics and Construction

<< < (2/2)

Duane Degn:

--- Quote from: jwatte on April 22, 2013, 03:20:28 PM ---Specifying "exactly on the border" numerics is how cheap consumer junk stays cheap, and also keeps disappointing with its inability to actually achieve stated specifications. You need the margins for anything robust.

--- End quote ---

Agreed but you need to know the initial requirements in order to figure out what kind of margin you want. The previous calculations regarding torque were wrong.

480oz*in of torque is what is needed to move the robot up a vertical wall (if the tires could maintain friction). A robot travelling on a horizontal surface (or a 45 degree incline) doesn't require the same amount of torque.

It's good to have more power than the bare minimum but it helps to know what the bare minimum is. 480oz*in is much more than the bare minimum in this case.

I've seen this type of error made many times in hobby robotics. There's a tendency to use the weight of the robot and diameter of the wheel to calculated the torque required to move the robot. If a robot were on a flat surface with little friction on its wheels a very small about of torque could propel the robot forward. The torque doesn't tell you how heavy the robot can be, it tells you what the acceleration of the robot will be. If you want the robot to drive up a slope (without needing a running start) you want to make sure the acceleration is at least as much as the acceleration of gravity pushing against the robot.

Once this bare minimum is calculated, then all the other factors such as friction, traction, etc can be addressed.

jwatte:

--- Quote ---480oz*in of torque is what is needed to move the robot up a vertical wall (if the tires could maintain friction).
--- End quote ---

You are right; that wasn't made clear. Another way to put that: With 480 oz*in of torque, you could accelerate your robot at 1g on a lossless, perfectly flat horizontal surface. That's a decent amount of acceleration.
The problem with "what is the absolute minimum" is that, in a lossless system on a flat surface, the absolute minimum is epsilon above zero. This is what trains use to be able to move very heavy loads with comparatively small engines. (The first steam locomotives just had a few horse power of power!)
So, using the 1g number for the estimated torque needed is a decent rule of thumb that lets you ignore all the corner cases. Once you have 1g of torque, you're going to be able to climb any incline you can get your tires to stick on, and you're going to have reasonable stopping power. For heavier robots (over 5 pounds, say,) stopping power is actually more important than starting power!

Duane Degn:

--- Quote from: jwatte on April 23, 2013, 11:46:24 AM ---You are right; that wasn't made clear.
--- End quote ---

Yes, but I think your last post clarified very nicely. Thanks.