Chief Delphi - How fast is fast to get on the hill?

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)

- General Forum (http://www.chiefdelphi.com/forums/forumdisplay.php?f=16)

- - How fast is fast to get on the hill? (http://www.chiefdelphi.com/forums/showthread.php?t=16689)

We were planning to o about 10.5 F/sec, but the main problem is what if you and your adjacent opponent both go up the hill in the first 15sec your white lines will intesect at the top of the hill. So it all depends on who gets to he top first, and still has enough power to push as many boxxes to there side as they can.

Thats the only problem I forsee, the only way to get around it is to program your bot to go up the center of the ramp instead of up the side. The middle holds alot of points that can go either way!

Re: A couple of limiting numbers

Quote:

Originally posted by Dodd
A robot of 130 lbm with a friction coefficient of 1.0 on the carpet (a popular assumption) can accelerate at 1 g (given enough motor power, proper gearing, etc) or 32 ft/sec^2. That means it can achieve a speed of 16 ft/sec in 0.5 sec, during which time it will have traveled 4 feet. The drive train power (delivered to the ground) to accomplish this initial acceleration is 1,400 w.

If this bot is to hang a 180 degree turn and storm up the ramp to clobber the wall, the frictional force on the carpet to make the turn is: F = m * V^2 / R. The frictional force is again 130 lbf (mu of 1, 130 lbm 'bot), and a turning radius of about 4' takes the bot from park to about the center of the ramp. This result says the fastest the bot can corner is at about 11.3 ft/sec, at which point it spins out. At this speed it takes about 1.1 second to make the 180 degree turn, and then another 1.0 second to make it to centered on the top, if the same speed is assumed up the ramp.

So, physical limits to drive to the top say about 2 seconds, assuming the bot can deliver sufficient torque to (almost) spin its wheels from rest all the way up to 11.3 ft/sec. I will bet no autonomous bot will be able to hit the wall of containers in less than 3 seconds.

Dodd

Good, now I don't have to work this out for myself :)

Isn't 1400 watts more then the motors can deliver? Maybe if someone uses stored energy in springs (though a team update kindof rules that out).

I forsee the biggest problem with speed being control. It would be very difficult for a human to make that turn that accurate every time, I see no way to do it without the line with the computer (and lets see you track a line at 10fps).

Greg

1400 watts/12 volts=116.7 amps.
Good luck.

Umm, programming?

Like was mentioned before, the challenge here is NOT how fast your robot can move.

Anyone who has worked with a feedback system (line tracking) will know that it will be very difficult to enable the robot to travel at full speed through the entire course of the turn and still hope to end up facing the boxes on top of the ramp.

Effectively, it would be VERY difficult for a robot to travel beyond 6-7 ft/sec and hope for the program to keep up with the curve.

Anyone that has worked in Lego-League will know that too basic a line tracking system makes your robot bounce left and round on the line, and a dead reckoning system with no feedback would be very difficult to get consistant with every condition.

Instead of the discussion being robot speed, perhaps it should shift to the real purpose of the challenge, sensors and programming :)

Location of Boxes

Did anyone stop to think that hitting the boxes at full speed is likely to knock them over your robot and into your opponents scoring zone?

If you do it right, it won't fall backwards, and yes, that was our teams first concern about building a short bot. :-D

fastest time, most bins to our side

We will be "there" faster than you can realize your robot is one, and a great deal of bins will be on our side very quickly...we have an arm specifically designed for this process (minimum, quick movement into position) that is very fast and strong, and just the right length. And we already have the code written to control it during autonomous (and for that matter human control) mode - its sweet! How many teams are using arms? How many of these are just long enough to get over and knock some bins? How many are uber-arms that go anywhere and do anything?

190's robot should be able to reach the top in 3 seconds this year. Control won't be a problem either. I can't say much more than that.

Ramp? What ramp?

:D

Maybe, if a group of programmers got ambitious, one team will make an inertial guidance system. :p All it would take are two of the yaw sensors, and good integration. However, it would take a huge amount of programming.

Quote:

Originally posted by FAKrogoth
Maybe, if a group of programmers got ambitious, one team will make an inertial guidance system. :p All it would take are two of the yaw sensors, and good integration. However, it would take a huge amount of programming.

it would take a few more sensors than just yaw rate sensors..... thats all i'll say

JUST HOW FAST???

rly, rly, RLY fast:D

We did some math here and figured that if your robot goes over the ramp at 16+ f/s your robot will "fly" over the ramp. Just something to keep in mind for some of you trying to achieve a very fast speed.

Quote:

Did anyone stop to think that hitting the boxes at full speed is likely to knock them over your robot and into your opponents scoring zone?

Thats why we designed our robot to knock over the blocks a certain way. I haven't the foggiest idea how fast our robot will be. I haven't had the time to calculate the speed yet. My one calculation made it around 5mph max.

bah..why bother with the bins... if so many people are going to knock the wall over at the very beginning, I say just wait until the mess on the ramp clears....... and just collect the bins you want!:p