|Antimatter.. It will help us in the never ending quest to build the bot under the weight limit. - Elgin Clock [more]|
This zip file contains all the information about the 2002 ThunderChickens CCT. It contains a Word document describing the theory and math, an Excel document with the Bill of Materials, AutoCad 2000 drawings, a STEP solid model of the assembly, SolidWorks 2001Plus native solid models, assemblies, an...
This zip file contains all the information about the 2002 ThunderChickens CCT. It contains a Word document describing the theory and math, an Excel document with the Bill of Materials, AutoCad 2000 drawings, a STEP solid model of the assembly, SolidWorks 2001Plus native solid models, assemblies, and drawings. This is the version of the CCT that we used at the Championship and has all the bugs worked out. I hope this information will prove useful to anyone who tries to use it.
Have fun with it!!
EDIT: The CCT is now U.S. Patent Pending as of 2/20/2003
12-29-2002 01:31 PMPaul Copioli
I apologize for taking so long to get back to you, but I have been on vacation. I will answer in the order you asked your questions.
1. The reason is interference. There are 3 different interference conditions to deal with when designing planetary gears. Bosch chose to allow more backlash to avoid the interference, because backlash is not that big of a deal for the kind of drill it was designed for.
2. Your gear ratio is slightly off, but here is the exact number:
(44/9) * (48/16) * (43/13 +1) = 63.18:1
That ratio is the speed ratio of the output if the ring gear is not spinning. At max speed of Chip that would be 88 RPM. The sun gear will spin at 375 RPM at max chip speed. The speed at 30 amps is important, because that is the speed you will be going at max pulling time. That speed is 64 RPM at output and, as you said, 276 at the sun gear.
3. Oh, but we do spin the ring gear faster than the sun. Our design speed for the ring gear is 600 RPM. That would make our total speed ratio range from 63.18:1 to ~8:1. You are right that there are some limitations when spinning the ring gear that fast.
4. I will try to address all your concerns here. All of the things you said in question 4 are true, however, we never spun the ring gear while at design load!! The worm gear setup I chose gave us pretty good backdriving resistance. What I mean is that the FP and Drill did not have to work when sitting still (most of the time). The point at which a worm gear set is backdriveable is highly dependent on 2 things: lead angle (number of threads) and Mu between the gear mesh. Every once in a while we would start backdriving the worm gear and it would look like we were in neutral. We weren't in neutral, but we were in a no move situation. You could stop us with just a pinky finger, but you could not move us back with 200 lbs pushing force. We did not like this condition, so we had a "pulsing" button on one of the joysticks that would pulse power to the FP and drill to stop backdriving. After a few trials (and burnt motors, as you eluded to), we got the pulsing routine to work. We did not have any further gear reduction, just used the backdriving resistance of worm gears.
Now, our design intent for the CCT had 2 main objectives: Max pushing force of 280 lbs and max speed of 10ft/sec. We knew you had to get to the goals fast and once you had them it did not matter. What we found is that with 1 goal, we could spin the ring gear at 600 rpm and move quite easily (drained the battery), but once we had 2 goals and transferred the weight ... no more ring gear spinning except for anti-backdrive. We had it designed for a specific purpose, but if I were to design it so I could use the spinning gears at all times no matter what, then I would spin the ring gear at about 75% of the sun gear speed at max speed (contrary to our 200% of sun gear speed). This only gives you a 3:1 jump in gear ratio as opposed to the 4.31:1 jump you talked about. The advantage of the 3:1 jump is that you get max torque available at all times, even at high speed; but you do not get the variation in gear ration that you require.
The one thing I will do differently this year is LOCK the ring gear in place when I am slow/torque mode, probably with pneumatics. You can do it with the CCT as designed, but need to add some holes. Another thing, do not directly couple the motors to the worm shaft as I did. Put the motors below the worm shaft and add another gear stage (1.5 or 2:1), it will make replacement easier.
I hope I helped.
01-03-2003 10:59 AMSimon G
Thanks for the detailed reply. It was very helpful. We actually looked at a similar design here at Sikorsky a while back for large scale helicopter applications...!
One more question. Your current worm lead angle is very close to 10 degrees (9 deg, 24' to be exact), which would give you backdriving in extreme conditions, especially with your smooth, ground worm...
Why didn't you use a single thread worm and a 30 tooth gear instead? (thus cutting your lead angle)
Was it a geometry (packaging) issue, or am I missing something critical? Your worm shaft thrust would go up with a smaller worm gear, but the torque requirements would be the same, right? Does the efficiency drop with a smaller lead angle? Is there a limitation on RPM with a single thread worm too?
01-03-2003 11:12 AMSimon G
Sorry, but one more quick question...
It's unclear how you actually attach the drill ring gear to your housing that holds the worm gear. I saw there was a note ref. on your 111-000 assembly drawing, but I didn't see the actual note. It's all steel... are you welding it or brazing it together?
PS... did you actually spend ~ $130 per bearing for the Kaydon reali-slim part?! (we have a very slim budget) Or do you have a contact who might donate something like that?
01-03-2003 01:18 PMPaul Copioli
The worm lead angle has a huge impact on efficiency. I would really want to use a quad thread worm, but I needed some backdriving resistance, so I chose dual. A single lead was just too inefficient.
The ring gear is press fit into a steel part (TC-2002-111-005)
that the worm gear (TC-2002-111-001) is bolted to. We used liquid Nitrogen (just because it is fun to play with) to shrink the ring gear and then pressed it into the steel part.
FANUC Robotics (the company I work for) is the largest non-military customer for KAYDON. I know our sales representative personally and he gives me any standard KAYDON bearing for free. In return, KAYDON is placed as a sponsor to our team. So, to answer your question; no we do not pay one penny for the KAYDON bearings. If you decide to use the CCT and need a couple of those bearings, I can probably get you a couple.