Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Technical Discussion (http://www.chiefdelphi.com/forums/forumdisplay.php?f=22)
-   -   Minibot climb rate (http://www.chiefdelphi.com/forums/showthread.php?t=88680)

Ether 02-04-2011 20:04

Re: Minibot climb rate
 

pi*5.5" wheel diameter = 17.3" wheel circumference.

90"/17.3" = 5.2 wheel revs to climb pole

(5.2 revs)/(2.2 seconds)*(60 seconds/minute) = 142 rpm

At that speed, the motors are generating virtually no torque.

Somebody please check my math.




Kevin Sevcik 02-04-2011 20:37

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1048475)

pi*5.5" wheel diameter = 17.3" wheel circumference.

90"/17.3" = 5.2 wheel revs to climb pole

(5.2 revs)/(2.2 seconds)*(60 seconds/minute) = 142 rpm

At that speed, the motors are generating virtually no torque.

Somebody please check my math.

He noted in passing that he modded the gearboxes to 10:1. Which probably puts him on the wrong side of peak power, but makes a little more sense.

Ether 02-04-2011 22:24

Re: Minibot climb rate
 
Quote:

Originally Posted by Kevin Sevcik (Post 1048492)
He noted in passing that he modded the gearboxes to 10:1. Which probably puts him on the wrong side of peak power, but makes a little more sense.

OK, is that what he meant? I thought that was an aside. The way I read it, he modded some gearboxes but then decided to use stock (as-is) ones.

Your interpretation makes more sense I guess.




Mr V 03-04-2011 01:25

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1048548)
OK, is that what he meant? I thought that was an aside. The way I read it, he modded some gearboxes but then decided to use stock (as-is) ones.

Your interpretation makes more sense I guess.




I also took it to mean that the gear boxes on this bot were the original 52 to 1 ratio.

The 10.4 to 1 boxes, based on our experiences and pile of smoked motors, won't work with a 5.5" diameter wheel. We ended up with ~3" wheels with our 10.4 to 1 gear boxes.

The math doesn't add up to a 2.2 sec time with 52 to 1 boxes either, unless that bot's weight is less than the total of the battery and motors, let alone the wheels and frame, again based on our experiences.

WileyB-J 03-04-2011 09:13

Re: Minibot climb rate
 
We used stock (52:1) gearboxes on the minibot. The average time was 2.2-2.3 seconds from bottom of the pole to the top. It weighs around 4.5-5 pounds.

If this helps:

AlDee 12-04-2011 13:04

Re: Minibot climb rate
 
Quote:

Originally Posted by jskene (Post 1045662)
We were consistently getting ~1.2 seconds. We got our minibot measured by team 1279 at the DC regional and got a higher accuracy measurement:

Our deployment arm, when it was working properly :(, was taking ~0.2 seconds to get the bot on the pole, so total time was ~1.27 seconds.

Below is a photo of the bot, which uses a cnc machined chassis. Our goal was to use only the minimum of materials necessary to hold the components in place. Weight came in at 2.5 lb.


Very Impressive. What type of switch are you using for the trip down?

Garret 12-04-2011 16:50

Re: Minibot climb rate
 
1 Attachment(s)
Our minibot had a 3.5 second climb. Unlike a lot of robots that I saw that used magnets to hold on to the pole. Our magnets were located in our wheels and then covered in electrical tape for traction.

I have attached a picture of one of the early forms of the minibot. the magnets are underneath the thin layer of black electrical tape.

We had some problems with deployment early on, mainly due to the fact that the robot was driving down the pole, but other than that it worked reasonably well.

Ether 12-04-2011 17:01

Re: Minibot climb rate
 
Quote:

Originally Posted by Garret (Post 1052047)
Our magnets were located in our wheels and then covered in electrical tape for traction.

I have attached a picture of one of the early forms of the minibot. the magnets are underneath the thin layer of black electrical tape.

Is there any gap between the magnets, or are they butted right up against each other?

How did you secure them to the wheel?

Did you alternate them N/S ?



wireties 12-04-2011 17:05

Re: Minibot climb rate
 
Late in the process we got a 1.7s minibot working with 1/4" tubes covered with surgical tubing. I never understood why everyone thinks magnets are necessary. If you center the minibot's mass left to right and balance the top to bottom mass difference by angling the minibot a bit, magnets are not necessary. Sure the minibot is not trying to climb straight up but it is also not trying to pull the magnets up with it - and its simpler.

Garret 12-04-2011 17:14

Re: Minibot climb rate
 
Quote:

Is there any gap between the magnets, or are they butted right up against each other?

How did you secure them to the wheel?

Did you alternate them N/S ?
The magnets are butted right up against eachother and are alternated. To stick them to the wheel we took electrical tape and made a loop to make it act like double-sided tape. Then we used a layer of electrical tape on top of the magnets to hold them to the wheel. Overall the minibot worked well, the main issues were silly mistakes made when hooking the battery up or when setting it up before a match (forgetting to set it to turn on). but overall when we didn't mess up before the match the minibot worked well.

Ether 12-04-2011 17:15

Re: Minibot climb rate
 
Quote:

Originally Posted by wireties (Post 1052061)
I never understood why everyone thinks magnets are necessary.

What method do you use keep the bot from falling off the pole?



wireties 12-04-2011 18:17

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1052065)
What method do you use keep the bot from falling off the pole?

if the mass if distributed left and right and the angle of the bot as it climbs the pole adjusted to account for top to bottom weight distribution (hang the minibot from a string and check the angle it wants to hang and almost duplicate that) it will just climb straight up. Ours did that though we did add a small cross-section of a pvc pipe to make it fall straight.

HTH

Tom Line 12-04-2011 18:27

Re: Minibot climb rate
 
Quote:

Originally Posted by wireties (Post 1052094)
if the mass if distributed left and right and the angle of the bot as it climbs the pole adjusted to account for top to bottom weight distribution (hang the minibot from a string and check the angle it wants to hang and almost duplicate that) it will just climb straight up. Ours did that though we did add a small cross-section of a pvc pipe to make it fall straight.

HTH

Am I right in assuming that you are using a wheel on each side of the pole, with the battery on the same side as one wheel, so that you're using torsion to lock the minibot 'on'?

Ether 12-04-2011 18:47

Re: Minibot climb rate
 
Quote:

Originally Posted by wireties (Post 1052094)
if the mass if distributed left and right and the angle of the bot as it climbs the pole adjusted to account for top to bottom weight distribution (hang the minibot from a string and check the angle it wants to hang and almost duplicate that) it will just climb straight up. Ours did that though we did add a small cross-section of a pvc pipe to make it fall straight.

HTH

I've never seen your bot so I cannot quite picture what you are describing.

I guess what I was asking was do you have some sort of snap-action device that springs shut and locks the bot to the pole, or do you "press-fit" the bot to the pole, or some other method?




wireties 12-04-2011 18:49

Re: Minibot climb rate
 
Quote:

Originally Posted by Tom Line (Post 1052100)
Am I right in assuming that you are using a wheel on each side of the pole, with the battery on the same side as one wheel, so that you're using torsion to lock the minibot 'on'?


Close - we did have a wheel on both sides but the weight was even left and right but the bottom (below the wheels) was heavier than the top (above the wheels). We launched it with the wheels at a 45 angle, it pulled itself onto the pole (righting itself in the process) and then fights (a little, maybe 2 or 3 degrees) against the cantilever on the way up. We depended mostly on flex in the motor mounts and sticky wheels to create the rolling friction against the pole.

HTH

wireties 12-04-2011 19:04

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1052114)
I've never seen your bot so I cannot quite picture what you are describing.

I guess what I was asking was do you have some sort of snap-action device that springs shut and locks the bot to the pole, or do you "press-fit" the bot to the pole, or some other method?

Press fit is the best description, flex in the motor mounts and sticky wheels created most of the rolling friction. We used a small cross section of PVC but that was needed to make it fall straight down

http://flic.kr/s/aHsjuentrP

Ether 12-04-2011 19:27

Re: Minibot climb rate
 
1 Attachment(s)
Quote:

Originally Posted by wireties (Post 1052123)
Press fit is the best description, flex in the motor mounts and sticky wheels created most of the rolling friction. We used a small cross section of PVC but that was needed to make it fall straight down

http://flic.kr/s/aHsjuentrP

See attached JPG.

1) The rollers look angled. Is that intentional?

2) It seems like the weight of the battery pack (2) would cause the bot to fall away from the pole in the direction indicated by the "->" arrow (assuming that the picture is right-side-up) ?

3) Is this part of the minibot?

Have you burned up any Tetrix motors with this configuration?





wireties 12-04-2011 22:39

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1052134)
See attached JPG.

1) The rollers look angled. Is that intentional?

2) It seems like the weight of the battery pack (2) would cause the bot to fall away from the pole in the direction indicated by the "->" arrow (assuming that the picture is right-side-up) ?

3) Is this part of the minibot?

Have you burned up any Tetrix motors with this configuration?

1) They do look a little angled. The frame must have bent some last time it came down, it is quite thin. They are supposed to be normal to the plate everything is mounted on.

2) Indeed it would if you did not tilt it fwd a wee bit, which means we are sacrificing some power I know but it was still quite fast

3) Yes, we had a slower minibot (very reliable) which launched from a 45 degree ramp and we wanted to use both minibots interchangeably. That is the only purpose for that piece of the assembly.

We burned up one motor after running it up many dozens of times during testing but that motor smoked some while testing an interim design so it was a little suspect.

We tried button magnets but this configuration seemed to work better. Perhaps a nicer shaped magnet would have worked better, maybe we'll try it before the off-season events.

HTH

billbo911 12-04-2011 23:54

Re: Minibot climb rate
 
I never would have guessed when I started this thread that it would have grown to 220+ posts. I also didn't consider one thing, this is FIRST. When I did the very first rough calculations, I had to make a lot of assumptions. My first guesstimate of climb time was in the ball park of 7 seconds. I should have known that the envelope would definitely be pushed, heck, almost to the breaking point.

The first minibot we built was very reliable and won a lot of races. It is not fast, just reliable, 4.5 seconds consistently. We rode that little bot all the way to the finals in Sacramento. It will be with us in St. Louis and be available for any team that needs it. It will have a deployment plate with it already set up, all that is needed will be a set of rails to send it out on and a battery. Her name is Miracle.

Now, just so you don't get confused, Miracle still works just fine, but we on Eagle Force don't like to sit by when we know we can do better.

Meet Miracle 2 v2.1a.


Miracle 2 v2.1a was inspired by a post by Sanddrag, along with several other designs we have seen.

When we originally set out to build a new mini to take to St. Louis. we were shooting for a sub 2 second mini. That is a far cry from the first estimate I made of ~7sec. way back when this thread started. So how does Miracle 2 v2.1a perform? Lets just say, < 1.2sec is pretty darn close to acceptable.
There is one issue we still need to solve, see if you can identify it.
http://www.youtube.com/watch?v=nvXV8_8UQGk

BTW,
I have this Minibot all set up in SolidWorks. There are a couple minor items missing from the CAD, like the nylon screws, but otherwise, it is fairly complete. If you want a zipped version of the project, just shoot me a PM.

slijin 13-04-2011 00:18

Re: Minibot climb rate
 
Wow, that's a pretty frightening suicide leap. Maybe adding a couple magnets would help rectify that.

SM987 13-04-2011 00:19

Re: Minibot climb rate
 
Looks awesome. We have several videos with kids making sure to catch the minibot after hitting the plate. Our solution was a piece of PVC a bit larger in OD to the pole cut to around 200 degrees so that the chaotic "landings" were minimized.

billbo911 13-04-2011 00:38

Re: Minibot climb rate
 
Quote:

Originally Posted by slijin (Post 1052276)
Wow, that's a pretty frightening suicide leap. Maybe adding a couple magnets would help rectify that.

Quote:

Originally Posted by SM987 (Post 1052278)
Looks awesome. We have several videos with kids making sure to catch the minibot after hitting the plate. Our solution was a piece of PVC a bit larger in OD to the pole cut to around 200 degrees so that the chaotic "landings" were minimized.

Well, the plan for us is to reduce weight even more, not add more. So, we are going to try a little physics first.

As far as we can tell, the battery is the main issue. It's momentum is causing the back of the mini to continue moving up, even though the frame has stopped against the top plate. To remedy this, we are going to flip the battery over and move it up until it is parallel with the top of the frame. That way, it has no where to travel once contact with the plate is made. (If that isn't enough energy into the sensor to trigger it, nothing will be.) Maybe some padding on top will help as well.

R1ffSurf3r 13-04-2011 11:30

Re: Minibot climb rate
 
Quote:

Originally Posted by billbo911 (Post 1052292)
Well, the plan for us is to reduce weight even more, not add more. So, we are going to try a little physics first.

As far as we can tell, the battery is the main issue. It's momentum is causing the back of the mini to continue moving up, even though the frame has stopped against the top plate. To remedy this, we are going to flip the battery over and move it up until it is parallel with the top of the frame. That way, it has no where to travel once contact with the plate is made. (If that isn't enough energy into the sensor to trigger it, nothing will be.) Maybe some padding on top will help as well.

What is the diameter and material of those wheels? That sucker is quick!

billbo911 13-04-2011 12:01

Re: Minibot climb rate
 
The axles are .375" and made of Polycarbonate. The tires are a silicon based surgical tubing material.

If you notice, this thing is accelerating almost the entire time. With slightly smaller axles, say .036", it might accelerate a bit quicker, but could reach max velocity too soon and increase the overall time to climb. The only way to tell is to try it, and we are happy right where it is.

From my experience with R/C car racing, I find the best approach is to be accelerating all the way until just before you start the next turn. In this case, that point is the top of the pole. So, it looks like this design is pretty close to optimal with the weight we currently have. Reducing the weight further will only improve things.

Richard Wallace 13-04-2011 12:59

Re: Minibot climb rate
 
Quote:

Originally Posted by slijin (Post 1052276)
Wow, that's a pretty frightening suicide leap. Maybe adding a couple magnets would help rectify that.

Minibot suicide is only one of the concerns about that leap -- the other is human safety. UL Safety Advisors at MSC made pit announcements, and distributed printed notices, requiring the use of gloves to catch minibots that regularly fell from the many practice poles teams had brought. That safety guidance is wise; at two events, I saw students in the first-aid stations being treated for bad cuts they had sustained by trying to catch falling minibots.

Kevin Sevcik 13-04-2011 13:25

Re: Minibot climb rate
 
Quote:

Originally Posted by billbo911 (Post 1052292)
Well, the plan for us is to reduce weight even more, not add more. So, we are going to try a little physics first.

As far as we can tell, the battery is the main issue. It's momentum is causing the back of the mini to continue moving up, even though the frame has stopped against the top plate. To remedy this, we are going to flip the battery over and move it up until it is parallel with the top of the frame. That way, it has no where to travel once contact with the plate is made. (If that isn't enough energy into the sensor to trigger it, nothing will be.) Maybe some padding on top will help as well.

Have you considered shifting the impact point behind the battery? If you're correct in your guess that inertial load from the battery is pulling you away from the pole, then you could just shift your pivot point at the top of the minibot. If your impact point is farther from the pole than the battery, then the inertia of the battery will push you more into the pole. It might not be the optimal solution, but it might be quicker to implement. Plus it'd preserve your current weight distribution.

If none of that made sense, sketch up a free-body diagram of the minibot as it makes contact with the pole. Assign an arbitrarily large upwards force at the center of mass of your minibot, balanced by a downwards force at your point of impact, and a horizontal force centered on your magnets. See what happens to the horizontal force as you move the impact point farther from the pole.

wireties 13-04-2011 13:42

Re: Minibot climb rate
 
Quote:

Originally Posted by billbo911 (Post 1052385)
From my experience with R/C car racing, I find the best approach is to be accelerating all the way until just before you start the next turn. In this case, that point is the top of the pole. So, it looks like this design is pretty close to optimal with the weight we currently have. Reducing the weight further will only improve things.

Provided it is reaching the same max velocity (no matter how long it accelerates), doesn't reaching max velocity sooner yield the fastest time up the pole? If it is linearly accelerating the whole way up, the average velocity is only 1/2 the max velocity. RC cars don't have gravity to consider I reckon.

HTH

Ether 13-04-2011 14:00

Re: Minibot climb rate
 
Quote:

Originally Posted by billbo911 (Post 1052385)

From my experience with R/C car racing, I find the best approach is to be accelerating all the way until just before you start the next turn. In this case, that point is the top of the pole. So, it looks like this design is pretty close to optimal with the weight we currently have.

...

If you notice, this thing is accelerating almost the entire time. With slightly smaller axles, say .036", it might accelerate a bit quicker, but could reach max velocity too soon and increase the overall time to climb. The only way to tell is to try it...

I could be wrong, but I don't think in this case the time-to-climb is minimized by making the wheel diameters so large that the minibot is accelerating all the way to the top.

Since you have a working minibot with a measured time-to-climb, you could easily plug your weight and wheel diameter into this model and adjust the friction value until the model matches your measured time-to-climb. Then change the wheel diameter in the model and see what diameter results in minimum time-to-climb.

A side benefit of smaller wheel diameter is less motor current.


Quote:

Reducing the weight further will only improve things.
Absolutely.



billbo911 13-04-2011 14:36

Re: Minibot climb rate
 
Quote:

Originally Posted by Kevin Sevcik (Post 1052415)
Have you considered shifting the impact point behind the battery? If you're correct in your guess that inertial load from the battery is pulling you away from the pole, then you could just shift your pivot point at the top of the minibot. If your impact point is farther from the pole than the battery, then the inertia of the battery will push you more into the pole. It might not be the optimal solution, but it might be quicker to implement. Plus it'd preserve your current weight distribution.

Kevin,
Your suggestion is actually quite easy to follow and might be the final solution. Currently, the initial impact to the sensor plate is the tip of the actuator arm of our Stop/brake limit switch. That is followed by the impact of the frame of the minibot.
To shift the initial impact point behind the battery should be as simple as placing a small amount of surgical tubing across the back of the top of the battery and then moving the battery up so that it is flush with the top of the frame. This will do two things:

1) Pad the battery with a bumper that is made of legal material, and reduce the possibility of damage to the battery from being slammed into the sensor plate.
2) It will shift the pivot point outside the center of mass of the minibot and thus cause it to rotate into the pole instead of away from it, as you suggested.

Quote:

Originally Posted by wireties (Post 1052419)
Provided it is reaching the same max velocity (no matter how long it accelerates), doesn't reaching max velocity sooner yield the fastest time up the pole? If it is linearly accelerating the whole way up, the average velocity is only 1/2 the max velocity. RC cars don't have gravity to consider I reckon.

HTH

There is a minor problem with this. To get the minibot to accelerate faster, you have to go with smaller "wheels". This is direct drive with no transmission or gearing to deal with.

As you stated, average velocity can be determined by max/2. But, with smaller wheels, max velocity goes down. (Here is where practice makes perfect. If we really wanted the absolute fastest minibot possible, we would have to make several different sizes of axles to find which gave the best time.)

Let me try a little back of the envelope discussion to help explain this.

Let's assume that max velocity is reached just as you reached the top of the pole. Max velocity reached was 8 ft./sec. So average would be 4 ft./sec and the 10 foot pole would have been climbed in 2.5 seconds.

Now, we reduce the "wheel size" so that we accelerate to max velocity quicker, but in doing so, also reduce our max velocity. Now max velocity is 6 ft./sec and is attained at 6 feet up the pole (ball park for discussion). The average velocity for that 6 feet ( (6 ft./sec)/2) = 3 ft/sec, and therefore, 2 seconds to travel. The remaining 4 feet of distance is traveled at a velocity of 6 ft./sec and takes .667 seconds.

So, 2.667 seconds to cover the same distance instead of 2.5 seconds. That is why you "should" accelerate the entire way.

These numbers are for discussion only and do not represent the actual values.


Either,

I will definitely run these values through your model and see what the result are.

As I said, the real test would be to test with multiple different "wheel" sizes and measure the actual result.
There are so many variables in this project, battery voltage, wheel tackiness, wheel friction against the pole etc., that sometimes trial and error just works out better.

Kevin Sevcik 13-04-2011 14:51

Re: Minibot climb rate
 
And now I'll turn around and doubt your back of the envelope calculations. I keep seeing people in this thread declaring that the average velocity is half the maximum velocity if you reach the maximum at the top of the pole. But this is only true if you're accelerating at a constant rate.

This is not the case with a permanent magnet dc motor. With a pmdc, as speed increases, produced torque decreases. As torque decreases, your acceleration decreases. So you're going to spend longer accelerating from max/4 to max/2 than you did accelerating from 0 to max/4. This is going to make a significant difference in your calculations. I'd recommend playing with the model Ether linked to and seeing the effect of changing wheel diameters, because it's not as straightforward as you think.

billbo911 13-04-2011 15:24

Re: Minibot climb rate
 
Quote:

Originally Posted by Kevin Sevcik (Post 1052440)
And now I'll turn around and doubt your back of the envelope calculations. I keep seeing people in this thread declaring that the average velocity is half the maximum velocity if you reach the maximum at the top of the pole. But this is only true if you're accelerating at a constant rate.

This is not the case with a permanent magnet dc motor. With a pmdc, as speed increases, produced torque decreases. As torque decreases, your acceleration decreases. So you're going to spend longer accelerating from max/4 to max/2 than you did accelerating from 0 to max/4. This is going to make a significant difference in your calculations. I'd recommend playing with the model Ether linked to and seeing the effect of changing wheel diameters, because it's not as straightforward as you think.

I believe you are spot on, but, for the sake of discussion, I believe the example I used is close enough to understand the principal.

Now, for complete accuracy, or theoretically accurate results, then Either's model should definitely be considered.

I will do a couple more measurements for accuracy tonight and then run them through the model and see what happens. (Honestly, with Spring Break being next week and the school being closed, I am fairly certain the only changes we will be making are those that keep the bot on the pole :yikes: )

Based on the values I recall for weight and wheel diameter, the Either Model shows that a wheel diameter of .436" is optimal. Currently we have axles that are .375" in dia. and then there is a very pliable Silicon Surgical tubing over that. Uncompressed, the entire wheel is fairly close to .5" in dia. So, under the load of compression into the pole, we are fairly close to the given number. Again, I will verify this tonight.

Ether 13-04-2011 16:15

Re: Minibot climb rate
 
Quote:

Originally Posted by billbo911 (Post 1052451)
Based on the values I recall for weight and wheel diameter, the Either Model shows that a wheel diameter of .436" is optimal.


Did you adjust the friction until the model agreed with your time-to-climb, and THEN adjust the wheel diameter to find the optimum? See excerpt below:

Quote:

Originally Posted by Ether (Post 1052426)
plug your weight and wheel diameter into this model and adjust the friction value until the model matches your measured time-to-climb. Then change the wheel diameter in the model and see what diameter results in minimum time-to-climb.



billbo911 13-04-2011 16:35

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1052466)
Did you adjust the friction until the model agreed with your time-to-climb, and THEN adjust the wheel diameter to find the optimum? See excerpt below:




I set up the model by entering in values for weight that are fairly close, but probably not exact. I then entered my best guesstimate for our current wheel diameter. I set the time to 1.2 sec, which is our measured time. Once these values were set, I adjusted the friction value until the distance value matched our climb distance.

Your Excel file does not modify the TIME value, TIME is an input, but it does modify the DISTANCE value as each input parameter is modified.

Once I found a friction value that matched our climb distance, I left that value alone. After that , I re-iteratively modified the wheel diameter until I maximized the distance traveled in the same time of 1.2 sec.

wireties 13-04-2011 16:44

Re: Minibot climb rate
 
Quote:

Originally Posted by billbo911 (Post 1052438)
These numbers are for discussion only and do not represent the actual values.

Its all about the numbers I reckon. You are trying hit the sweet spot on the motor curve. The fastest minibot I've seen is from FIRST Team 148. I didn't measure it but it seemed to be in the 1s range. It sure looked like that thing hit max V quickly.

Good luck in St Louis!

Ether 13-04-2011 17:01

Re: Minibot climb rate
 
Quote:

Originally Posted by billbo911 (Post 1052476)
I set up the model by entering in values for weight that are fairly close, but probably not exact. I then entered my best guesstimate for our current wheel diameter. I set the time to 1.2 sec, which is our measured time. Once these values were set, I adjusted the friction value until the distance value matched our climb distance.

That should be OK.

Quote:

Once I found a friction value that matched our climb distance, I left that value alone. After that , I re-iteratively modified the wheel diameter until I maximized the distance traveled in the same time of 1.2 sec.
That should get you in the ballpark. If you are game, try this instead: once you've found the friction value as described above, modify the wheel diameter and observe the graph to read the climb-time* for the given (fixed) distance. Find the wheel diameter which minimizes the climb-time for that given distance.

If you don't want to mess with that, would you mind posting your present observed values for bot weight, wheel diameter, and climb-time so I can run them?


* you can change the value in cell A7 to change the scale of the graph if necessary

billbo911 13-04-2011 17:12

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1052488)
That should be OK.



That should get you in the ballpark. If you are game, try this instead: once you've found the friction value as described above, modify the wheel diameter and observe the graph to read the climb-time* for the given (fixed) distance. Find the wheel diameter which minimizes the climb-time for that given distance.

If you don't want to mess with that, would you mind posting your present observed values for bot weight, wheel diameter, and climb-time so I can run them?


* you can change the value in cell A7 to change the scale of the graph if necessary

I don't mind at all. I believe the values I used were as follows:
2.35LB.
.436 dia (final), .5 original to find friction.
1.2 sec. climb time.

Ether 13-04-2011 17:40

Re: Minibot climb rate
 
Quote:

Originally Posted by billbo911 (Post 1052492)
I don't mind at all. I believe the values I used were as follows:
2.35LB.
.436 dia (final), .5 original to find friction.
1.2 sec. climb time.

Oops, forgot to ask: what did you use for the climb distance (ie the distance climbed in 1.2 seconds). Also, just to be sure: you didn't use a "ramp" launcher did you? In other words, the minibot started out from a dead stop and climbed in 1.2 sec correct?



billbo911 13-04-2011 17:50

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1052501)
Oops, forgot to ask: what did you use for the climb distance (ie the distance climbed in 1.2 seconds). Also, just to be sure: you didn't use a "ramp" launcher did you? In other words, the minibot started out from a dead stop and climbed in 1.2 sec correct?


I should have known that was needed :o
95 inches.

Ether 13-04-2011 18:22

Re: Minibot climb rate
 
3 Attachment(s)
Quote:

Originally Posted by billbo911 (Post 1052503)
95 inches.


Given your inputs of 2.35 lb minibot weight, 0.5 inch shaft diameter, 1.2 seconds climb time, and 95 inches climbed distance, here's what I came up with:

friction: 0.83 pound

optimum diameter: 0.43 inches

... which tends to support the conclusion that you are pretty close to the optimum diameter. The only puzzling thing is that the graph in the model looks like your bot velocity with your original .5 inch diameter has pretty much reached its peak value half way up the pole. Was your assessment that the bot continued substantial acceleration all the way to the top a subjective judgment or did you analyze data from a video?

(see attachments)


billbo911 13-04-2011 18:30

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1052513)
Given your inputs of 2.35 lb minibot weight, 0.5 inch shaft diameter, 1.2 seconds climb time, and 95 inches climbed distance, here's what I came up with:

friction: 0.83 pound

optimum diameter: 0.43 inches

... which tends to support the conclusion that you are pretty close to the optimum diameter. The only puzzling thing is that the graph in the model looks like your bot velocity with your original .5 inch diameter has pretty much reached its peak value half way up the pole. Was your assessment that the bot continued substantial acceleration all the way to the top a subjective judgment or did you analyze data from a video?

(see attachments)

Very much subjective. I based the comment on visual and auditory observation.

The bottom line is, we are really close. Lightening will only help, and deployment will be the biggest key to consistent success.

Mark Sheridan 13-04-2011 19:03

Re: Minibot climb rate
 
Hey Bill

Perhaps there is wheel slip in the beginning of the climb, thus prolonging the acceleration in the beginning of the climb. 0.83 lbs is a lot of friction for the minibot. My guess is the the friction is lower and the minibot can accelerate quicker, but slippage losses in the beginning of the climb slow down the time. It overcomes the slip when motor toque is lower and the rpm higher, thus higher up on the pole. From here is starts accelerating with its full potential. Overall it appears to accelerate the entire length. It would be like pushing (or distorting) the velocity curve to the right.

On our revised minibot, it appears to slip a bit in the beginning but accelerates quickly after it. We run a 2.5 pound minibot with a 3/8 od wheel. It climbs in about 1.5 seconds. We are defiantly running higher magnets forces then yours. By increasing the magnet force, you can prevent wheel slip and prevent the robot from flying off.

I will see if your numbers compare well to my calculator later tonight.

Ether 13-04-2011 19:20

Re: Minibot climb rate
 
Quote:

Originally Posted by Mark Sheridan (Post 1052527)

Perhaps there is wheel slip in the beginning of the climb

I think you may be right:

http://www.chiefdelphi.com/forums/sh...76#post1045576




boomergeek 13-04-2011 19:35

Re: Minibot climb rate
 
We finished our last regional and did not implement our secret weapon:
(although we did tell the Boston Regional judges about it)
Rubber on rubber (1.2 CoF sliding friction) for the first 3 inches of vertical climb before switching to rubber to steel (0.3-0.6 CoF sliding friction).
We projected that would save us 0.2 to 0.3 seconds and still keep normal force low.

The rubber on rubber is minibot contact to a hostbot ramp that runs vertically parallel to the tower pole and then switches onto the pole before the minibot crosses the 18" line.

billbo911 13-04-2011 21:41

Re: Minibot climb rate
 
OK, more info and a video.

Kevin,
Shifting the battery up to the top and placing some surgical tubing at the back of the battery did the trick. Now our pivot point is the farthest point from the pole. This video is taken with a fairly depleted battery.

http://www.youtube.com/watch?v=P89k0hlHd1E

Either,

More accurate measurements:

Weight is 10052g, or, 2.32 lb.
Wheel diameter is effectively .45" when compressed against the pole.

So, based on our earlier discussions, it is really close to ideal.

crazyStone 13-04-2011 22:53

Re: Minibot climb rate
 
2.215 lbs right now but we are a little slow at the beginning of our climb at about 1 second, we are going to tinker with our magnets to try and decrease the amount of time. We shall see how well this works in the next few days.

sumginazu 14-04-2011 00:03

Re: Minibot climb rate
 
billbo911 - Team 1351's minibot had the exact same problem with falling off the pole. I'll try to implement changes similar to what you did; hopefully we can avoid catastrophe at Calgames. :) We already came close to destroying our extremely fragile Minibot at the Silicon Valley regional. It fell off of the pole twice during matches, where nobody could catch it - the .090 aluminum that forms its chassis barely survived.

Mark (or anyone else) - Other than increasing magnet strength, how can one reduce the wheel slip? I know there are obvious options like changing the tread material, but I'm wondering if any other techniques could also work.

billbo911 14-04-2011 00:23

Re: Minibot climb rate
 
Quote:

Originally Posted by sumginazu (Post 1052635)
billbo911 - Team 1351's minibot had the exact same problem with falling off the pole. I'll try to implement changes similar to what you did; hopefully we can avoid catastrophe at Calgames. :) We already came close to destroying our extremely fragile Minibot at the Silicon Valley regional. It fell off of the pole twice during matches, where nobody could catch it - the .090 aluminum that forms its chassis barely survived.

Take a close look at the top of the battery. That "bump" is just a couple pieces of surgical tubing under some electrical tape. The battery is also raised up so that it is level with the robot frame. By doing this, the point that hits the sensor plate first is the bumper on the battery. The tubing compresses helping to dissipate some energy and also forces the minibot into the poll instead of away from it.


Mark Sheridan 14-04-2011 01:01

Re: Minibot climb rate
 
Quote:

Originally Posted by sumginazu (Post 1052635)
Mark (or anyone else) - Other than increasing magnet strength, how can one reduce the wheel slip? I know there are obvious options like changing the tread material, but I'm wondering if any other techniques could also work.

This sounds like the best solution:


Quote:

Originally Posted by boomergeek (Post 1052538)
We finished our last regional and did not implement our secret weapon:
(although we did tell the Boston Regional judges about it)
Rubber on rubber (1.2 CoF sliding friction) for the first 3 inches of vertical climb before switching to rubber to steel (0.3-0.6 CoF sliding friction).
We projected that would save us 0.2 to 0.3 seconds and still keep normal force low.

The rubber on rubber is minibot contact to a hostbot ramp that runs vertically parallel to the tower pole and then switches onto the pole before the minibot crosses the 18" line.

I like this idea a lot. It is unfortunate that you did not have time to implement it. I hope your team will be able to try it out at an off season competition.

Bill our calculations show that your minibot should climb in about 1.3 seconds. I think its a little off, I am going to tinker with it a bit. Its set up to have drive-train efficiency different at low rpm than high rpm. Just to be clear, this is a separate efficiency from motor efficiency.

jacob9706 14-04-2011 02:02

Re: Minibot climb rate
 
You can also modify the transmission or just take it off all together. Do the calculations with no transmission and a 3/8 shaft with surgical tubing as a wheel. I got under 1.5 seconds at max efficiency.

AlDee 14-04-2011 03:46

Re: Minibot climb rate
 
Quote:

Originally Posted by wireties (Post 1052061)
Late in the process we got a 1.7s minibot working with 1/4" tubes covered with surgical tubing. I never understood why everyone thinks magnets are necessary. If you center the minibot's mass left to right and balance the top to bottom mass difference by angling the minibot a bit, magnets are not necessary. Sure the minibot is not trying to climb straight up but it is also not trying to pull the magnets up with it - and its simpler.

At the beginning of the build season, someone had suggested magnets and we quickly dismissed the idea, thinking with the potential weights involved, magnets would not be practical, except perhaps to hold a latching mechanism closed. As the build season progressed we experimented with many differnt concepts and prototypes. I favored a balanced approach as you describe. I found that anytime we would get close, one tweak would throw the whole thing off. When someone put together a prototype with a rare earth magnet, it became very simple. No muss no fuss, just get it to the pole and climb. That was the direction we went and it worked very well.

Mark Sheridan 14-04-2011 14:12

Re: Minibot climb rate
 
Quote:

Originally Posted by AlDee (Post 1052671)
At the beginning of the build season, someone had suggested magnets and we quickly dismissed the idea, thinking with the potential weights involved, magnets would not be practical, except perhaps to hold a latching mechanism closed. As the build season progressed we experimented with many differnt concepts and prototypes. I favored a balanced approach as you describe. I found that anytime we would get close, one tweak would throw the whole thing off. When someone put together a prototype with a rare earth magnet, it became very simple. No muss no fuss, just get it to the pole and climb. That was the direction we went and it worked very well.

Yeah my team was in the same boat as you. I honestly thought magnets would not work. Now my students are amused that mentors can be wrong too.:)

Bill,
I fixed the calculations on my end, I am getting a climb rate of about 1.10 seconds. So I think slipping is defiantly slowing you down.

I have also been comparing my minibot to the the polycarbonate one here:
http://www.chiefdelphi.com/forums/sh...&postcount=191

Ours is the same mass but defiantly slower. I think another variable to consider is the accuracy of shaft placement. Ours is defiantly poor, the shafts are not perfectly perpendicular to the pole, so we are defiantly loosing efficiency. I don't know if this is common for others, but when we attach our un-powered minibot to the pole, it won't roll down unless we move it a bit. An accurate minibot should just slide down when not powered. Ours still slides down after the climb because it bounces off the top plate.

billbo911 14-04-2011 14:33

Re: Minibot climb rate
 
Quote:

Originally Posted by Mark Sheridan (Post 1052824)
Yeah my team was in the same boat as you. I honestly thought magnets would not work. Now my students are amused that mentors can be wrong too.:)

Bill,
I fixed the calculations on my end, I am getting a climb rate of about 1.10 seconds. So I think slipping is defiantly slowing you down.

What?? Mentors wrong!! Seriously, never happens.:D

I've been thinking about the slipping. Though I know we have some really grippy "wheels", it is possible. It could also be that the shafts are spinning in the "wheels". We will try a couple experiments today to see if we can resolve it.

The reality is, we are going from a dead stop to max torque instantly, and fighting gravity while we are at it. The best option to overcome that is a curved ramp. I seriously doubt we have time to redesign out deployment system now. So.... stopping the slip is our best bet.

Ether 14-04-2011 15:01

Re: Minibot climb rate
 

Given:

0.5 inch diameter wheel

0.09 Nm stall torque

0.4 coefficient of friction between pole and wheel (estimated, YMMV)


Find: required normal force to prevent slipping


Solution:


1) calculate the motive force at the wheel/pole interface produced by stall torque:

(0.09 Nm)/[(0.5/2 inches)*(0.0254 meters/inch)] = 14.17 Newtons = 3.19 lbs


2) divide the motive force by the coefficient of friction:

(3.19 lbs)/(0.4) = 8 pounds normal force required for no wheel slip


Edit:

To be clear, the above calculation is for the normal force on ONE wheel (shaft) powered by ONE motor.

If the wheel (shaft) is powered by two motors, then the number would double.



JesseK 14-04-2011 17:34

Re: Minibot climb rate
 
Different excercise from Ether's:

Using the second derivative of Ether's equation x(t), the acceleration versus time is:
a(t) = -D*e^(-B*t)
To find out when there's no more acceleration, set a(t) = 0 -- since ln(0) is negative infinity, we basically see that the minibot is always technically accelerating. HOWEVER we can see when it gets "close" to its max. speed by substituting in a very small number for a(t). In my case I chose 0.01 m/s^2 since that's small enough to become "unnoticable" IMO.

Sovling the equation for t, to see when the minibot is no longer accelerating gives us
t = ln(0.01/-D)*(1/-B)

Then we can plug that 't' back into the original equation to figure out the height at which the minibot is no longer accelerating.

Presuming:
  • 0.40" Wheel Diameter
  • 2.35lb Weight
  • 0.8225lbs of friction (anecdotally determined)
  • 2 motors, drive driven to the wheels
  • No slip
The time of acceleration = 0.86 seconds
The distance of acceleration = 5.57 feet

crazyStone 15-04-2011 00:23

Re: Minibot climb rate
 
Our frame which we are mounting our motors is actually holding up even better than we had expected even though it is only .040" thick. But we tinkered with the magnets and are now running even faster and have now begun to launch off the pole as we approach <1 second....

waialua359 21-04-2011 03:48

Re: Minibot climb rate
 
Our team is pretty excited about our new and improved minibot.
It is currently performing between 1.1 - 1.2 seconds from dead start (no ramp).
Our old one was good enough in weeks 2 and 4, but knew that it had to improve to stay with the mix of teams that are hovering above, at, or right below the 1 second mark.

jblay 21-04-2011 04:13

Re: Minibot climb rate
 
Quote:

Originally Posted by billbo911 (Post 1052570)
OK, more info and a video.

Kevin,
Shifting the battery up to the top and placing some surgical tubing at the back of the battery did the trick. Now our pivot point is the farthest point from the pole. This video is taken with a fairly depleted battery.

http://www.youtube.com/watch?v=P89k0hlHd1E

Either,

More accurate measurements:

Weight is 10052g, or, 2.32 lb.
Wheel diameter is effectively .45" when compressed against the pole.

So, based on our earlier discussions, it is really close to ideal.

It seemed to me in the initial video like the plate on the top of your tower was slanted which may have caused your minibot to fly off. Its a problem we had before connecticut and then I realized that was the reason and we never had the issue again.

I don't know if you tested with a full battery or not but the battery being full would change the speed of the minibot I imagine and cause it to go slower and could be part of the fix for the minibot flying off the pole which isn't an actual fix.

wireties 21-04-2011 07:01

Re: Minibot climb rate
 
Quote:

Originally Posted by AlDee (Post 1052671)
At the beginning of the build season, someone had suggested magnets and we quickly dismissed the idea, thinking with the potential weights involved, magnets would not be practical, except perhaps to hold a latching mechanism closed. As the build season progressed we experimented with many differnt concepts and prototypes. I favored a balanced approach as you describe. I found that anytime we would get close, one tweak would throw the whole thing off. When someone put together a prototype with a rare earth magnet, it became very simple. No muss no fuss, just get it to the pole and climb. That was the direction we went and it worked very well.

I surrender and am ordering magnets from K&J to try before the off-season events!

Brandon Holley 21-04-2011 08:23

Re: Minibot climb rate
 
Quote:

Originally Posted by jblay (Post 1055097)
I don't know if you tested with a full battery or not but the battery being full would change the speed of the minibot I imagine and cause it to go slower and could be part of the fix for the minibot flying off the pole which isn't an actual fix.

I'm a little confused...you are saying that a full battery will cause the minibot to go slower??

-Brando

Ether 21-04-2011 08:57

Re: Minibot climb rate
 
Quote:

Originally Posted by waialua359 (Post 1055096)
Our team is pretty excited about our new and improved minibot.
It is currently performing between 1.1 - 1.2 seconds from dead start (no ramp).

Could you tell us more about it?

- weight?

- wheel (shaft) diameter?

- wheel (shaft) tread material?

- magnets or something else?

- have you measured the friction?



Alpha Beta 21-04-2011 09:26

Re: Minibot climb rate
 
Quote:

Originally Posted by Brandon Holley (Post 1055115)
I'm a little confused...you are saying that a full battery will cause the minibot to go slower??

-Brando

Well, of course. A fully charged battery has all of those extra electrons to lug around. Oh wait...that's not right. They're still there just switching between the anode and cathode. I guess I'm confused too. :rolleyes:

Team 288 21-04-2011 10:43

Re: Minibot climb rate
 
Quote:

Originally Posted by Ether (Post 1055122)
Could you tell us more about it?

- weight?

- wheel (shaft) diameter?

- wheel (shaft) tread material?

- magnets or something else?

- have you measured the friction?


I can probably answer for this person, by inserting 931's minibot, instead of theirs.

931 has a minibot that climbs justbelow the 1 second mark.

Weight : 2.2 lbs

Wheel shaft diameter : .45 inches

Wheel shaft material : Electrical tape(standard)

27 lb pull magnet

No. =)

We'll have 6 of these at worlds and are willing to let other teams borrow (teams in Archimedes get priority)

Here's a video with a 7 foot pole. We got .7 seconds for total climb time.

http://www.youtube.com/watch?v=cUSCBbhu4JM

jblay 21-04-2011 12:07

Re: Minibot climb rate
 
Quote:

Originally Posted by Brandon Holley (Post 1055115)
I'm a little confused...you are saying that a full battery will cause the minibot to go slower??

-Brando

Didnt see that I wrote that backwards the first time. I meant the battery not being full would slow the minibot down cause before the video he says that the battery was not full.


All times are GMT -5. The time now is 10:25.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi