FRC 95, The Grasshoppers, Build Thread

[soul777toast]

Originally Posted by JamesCH95

The shooter is now wired up and functional!

Here is the firing stage of the shooter spooling up to full-speed. It is powered by 2x RS775Pros with 1:1 versa planetary transmissions. All belts are designed 'perfect' center to center. Shafts are 1/4in steel D-shafts supported by bearings rated to ~50krpm.

Kovaka firing stage spool-up

When I showed this video to my father (Lewis Sussman, he was our head coach for a few years back in the mid 2000's), his response was (and I quote!) "It sounds like a giant hive of killer bees!"

[JamesCH95]

Originally Posted by soul777toast

When I showed this video to my father (Lewis Sussman, he was our head coach for a few years back in the mid 2000's), his response was (and I quote!) "It sounds like a giant hive of killer grasshoppers!"

FTFY

[soul777toast]

Originally Posted by JamesCH95

FTFY

"A plague of [killer robot] locusts o'er the land!"

[Chris Endres]

Originally Posted by JamesCH95

Started our drive-train burn-in only to have the burn-in go a little too literal for us...

Side plate off to investigate:



Yeah... had some rubbing chain tensioners...



Milled them down, reassembled, and tried out some defenses.

Moat

Low Bar

Did some crossing exercises, everything looked good, until...



Weirdest chain failure I've seen, and the only one I've ever had on a 95 robot. Master-link was good, fractures are on opposite sides of the chain loop, tensioners were installed and slack(ish), and our nominal FoS is around 4 or 5. Bizarre...

This was my biggest concern, along with cracking wheels, when designing this year. I would suggest getting a lot of spares for competition. Going over the obstacles and slamming into the ground will definitely break chain, as you saw first hand. At least it happened now and not at a competition.

[JamesCH95]

Originally Posted by Chris Endres

This was my biggest concern, along with cracking wheels, when designing this year. I would suggest getting a lot of spares for competition. Going over the obstacles and slamming into the ground will definitely break chain, as you saw first hand. At least it happened now and not at a competition.

What about slamming down causes chains to break in your estimation? (I mean this as an honest question, not a sarcastic one)

[JesseK]

Originally Posted by JamesCH95

Moat

[.URL="[.img]http://i.imgur.com/FQOvLiV.jpg[/img]"]Moat[/url]

(I added the '.' to the tags)

Getting an 'about:blank' on this URL. Looks like a vBB IMG tag is embedded inside the URL tag. The IMG tag links to the picture just prior to the link. The other youtube link works fine though.

[JamesCH95]

Originally Posted by JesseK

[.URL="[.img]http://i.imgur.com/FQOvLiV.jpg[/img]"]Moat[/url]

(I added the '.' to the tags)

Getting an 'about:blank' on this URL. Looks like a vBB IMG tag is embedded inside the URL tag. The IMG tag links to the picture just prior to the link. The other youtube link works fine though.

Fixed, thanks. Clearly clicked the "image" shortcut when I meant to click the link shortcut.

[Chris Endres]

Originally Posted by JamesCH95

What about slamming down causes chains to break in your estimation? (I mean this as an honest question, not a sarcastic one)

When pivoting on a wheel, driving, and slamming down on the ground with 120+ lbs will exert huge amounts of not only tension, but shock to the chain. There were times in 2012 where teams that went over the middle bump would shear a chain link after slamming into the ground. This is especially prevalent with #25 chain used on drive wheels.

That is my understanding and experience (wheelies in 2013 broke our chains sometimes), if I'm not completely right, sorry.

A way you can help prevent it is by switching to #35 chain, but honestly, just bring 5 or so spare chain lengths to competition (just in case).

[JamesCH95]

Originally Posted by Chris Endres

When pivoting on a wheel, driving, and slamming down on the ground with 120+ lbs will exert huge amounts of not only tension, but shock to the chain. There were times in 2012 where teams that went over the middle bump would shear a chain link after slamming into the ground. This is especially prevalent with #25 chain used on drive wheels.

That is my understanding and experience (wheelies in 2013 broke our chains sometimes), if I'm not completely right, sorry.

A way you can help prevent it is by switching to #35 chain, but honestly, just bring 5 or so spare chain lengths to competition (just in case).

Interesting. I believe your empirical evidence for sure, and we always bring an extra 25ft spool of chain and dozens of master links! Cheap insurance...

[JamesCH95]

Last night we got our shoulder joint hooked up, but found that we couldn't hold tension properly in the chain, so we'll be working to improve that tonight.

We also got our long-awaited pneumatic wheels in, which helped immensly with crossing barrier like the rock wall.

However, we snapped a chain again! I re-did my tension calculations and found that I had made a mistake and that our FoS was not 4-5 ish, but closer to 1.8! That's no good, clearly. So, we'll be upgrading to #35 chain for our drive.

Pictures and video to follow.

[JamesCH95]

It's been a long couple of days...

Re-chained our drive pods with #35 chain in a night. Remarkably we were able to make everything fit.



We discovered that some spacers had been installed improperly, which likely contributed to chain breaks, but installed the #35 chain anyway as insurance.

We did a bunch of driving around and everything went smoothly for once! We're able to conquer the low bar, moat, rock wall, and rough terrain (all in low-friction flavors).



Our "W-Drive" is proving to be quite smooth over a variety of obstacles. It seems to be effectively mixing the great carpet performance of Colsons with the good cushion of pneumatic wheels when going over obstacles.

On the other hand we've blown up a couple welds, thrown tons of belts, and are just starting to work on PID drive/arm tuning... so... it's going somewhat rough at the moment, but are working to correct as much as we can as quickly as we can.

Going on hour 11 today... wish us luck!

[soul777toast]

I'm sure James is totally burnt out after this weekend, so I'll post a quick update- after a monumental final push (just under 12 hours on Saturday and another 9 on Sunday ) by our heroically dedicated students and coaches, we've ironed out most of the last bugs and wrinkles with our design. We have now successfully tested and proven functionality on all of the static defenses, ball pickup, and ball shooting! Of course, anyone who's done this before knows all too well that a robot build is NEVER "finished", so we'll likely be tinkering and tweaking right up to the last minute before bagging and tagging, but things are definitely looking good.

Most of what's left is code and drive practice, and our crack programming team is chugging away on drive tuning, vision, auto move, and awakening sentience in our maniacal artificial intelligence (It's passed a turing test, now we just need to implement the "Destroy All Humans" protocol, currently it's stuck on "Annoy" instead of "Destroy" >_< )

Last bit of good news for now, we had a friend of the team who's previously acted as an inspector at competitions come in and give us a going over, and we've passed robot inspection with flying colors!

I'm sure once James has awoken from his coma he'll be back to post pics/vids and to chime in, but for now good luck to all and see you at the competitions!

[JamesCH95]

An end-of-build update.

We bagged a nominally functional robot. All of our features work, except for vision tracking (due to inadequate LED illumination, mostly).

Here we are testing with a ring light that did illuminate well enough, but did not package well into the robot.



Down the mouth of the beast. You can see all of the belts and rollers, running on 1/4in D-shaft with set screws. The surgical tubing in the back bowing in the side-panels is absolutely key to our consistent shooting. It centers the ball in the back of the shooter and prevents any drag from the sides on the ball as it exits the later stages. It acts like a funnel of sorts. The barber poles help ensure that balls enter the shooter/collector properly and prevents the ball from rolling belts off of their pulleys.



Crossed a lot of defenses to work out our #35 chain upgrade and get a little driver practice in. Also, we are doing header preservation with an ardupilot, so the driver pulls the joystick trigger and the robot keeps itself going straight. A nice little aid for crossing.



Disclaimer: we generally include our b-roll stuff rather than editing down to the highlights/best performance. Partly because editing video takes time, and partly because it can be disheartening to compare ones own holistic experience to another teams' highlights.

Some videos:

Self-righting with the arm.

The magic of the W drive! Here's the rock wall.

One shot from shooter consistency testing. We just skim the 8-9ft ceiling at roughly 20ft range, landing at about 30ft.

Crossing the moat/trying to get stuck on the moat. We can come to a complete stop anywhere on the moat and either back out and/or continue through.

Crossing the ramparts would be easier if we made sides for the defenses... I think we'll do okay on a real field. Note: heading preservation was NOT working for this trial

I don't have a good video for the rough terrain, but we can drive over, and even turn, on it without issue.

Overall... a long build season for the Grasshoppers, but a rewarding one. We got way outside of our comfort zone on a lot of different aspects of this robot. But, that's how you grow! We look forward to competing with our fellow NEDErs in Windham, UNH, and UMass!

[JamesCH95]

One of our programming coaches has come up with an interesting side-project he calls Pit Hub. He has setup a wired network using a Raspberry Pi to run Git Hub in the pits, so programmers can operate as normal at competitions. I'm excited to see how it stream-lines our competition coding efforts!