30-31.1.12

Monday and Tuesday have been relatively uneventful. I now have the correction plates for the design screwup in the drive plates fully fabbed and have started installing them.

I had fun today (Tuesday) teaching myself to power tap. I saw Mr. Van Dyke power tapping PVC over the weekend and was captivated. After asking him about it, and doing some reading at home, I decided to try it with a larger tap (1/4-20) and soft (6061) material. While I was relatively shy at first and wasn't making it all the way through the .125" plate, I did eventually get the hang of the basic idea. The process of power tapping consists of using a drill press or mill as a framework for tapping instead of attempting to line things up by hand, as well as doing so under the power of the drill or milling machine motor. The way this is done is by placing the tap in the drill chuck and hand-tightening it; the reason for hand-tightening only is that it causes the chuck to act as an artificial torque-clutch slipping before breaking the tap. Then (with copious amounts of cutting fluid) one spins the tap up to speed by turning the drill press on, brings the tip down to just above the surface, quickly shuts off the drill press (or mill) and plunges the tap into the hole, letting it coast through (the whole process is done with very little pressure). Overall, I'm excited to keep improving my tapping skills.

Last night and today I did some work on my own time on the materials of Project ÜberTank, and discovered that this program will require a good bit of funding to manufacture at the level required for an FRC robot. The shear amount of tread required for a robot that is 37"x27" is quite large and rather expensive. A preliminary run-through of the parts and materials on McMaster-Carr ended with a total price for just the tread being close to a thousand dollars. While I was slightly deterred at first, Ms. Lindborg told me to go forward and not let money stop me. Whether this is a true commitment from the power that controls team money to go through with it, or simply her curious as to what I am capable of, is something I do not know. Time will tell, I suppose...

The current BoM (Bill of Materials) for a single linkage:

• 3.25" 6061 aluminum round-stock of a 0.25" diameter (McMaster #8974k31)
• 1.5" W1 water-hardened tool steel of a 0.125" diameter (McMaster #8890k18)
• (2) 1" long nylon spacers of a 0.5" o.d. (Outside Diameter) and a 0.25" i.d. (Inside Diameter) (McMaster #94639A506)
• (1) 0.75" long nylon spacer of a 0.25" o.d. and a 0.14" i.d. (McMaster #94639A305)
• 2" latex tubing of a 0.75" o.d. and a 0.5" i.d. (McMaster #5234K85)
• (2) self-locking retainer clips for a 0.125" shaft (McMaster #98430A116)

With these parts/materials the cost for one 1.25" link is $5.71, and with a rough estimate based on the size of our robot the number of links needed will be approximately 165 links. Do the math and you get well over $900. This number of links does not account for the lot size of parts on McMaster-Carr (it will not be possible to buy the exact number of parts/amount of material needed), nor does it consider spare parts or extra material for manufacture error.

On top of all this comes nearly every significant problem that we had with the first "standard" tank drive (machining the Drive Plates, et cetera...), as well as the necessity to fabricate some sort of sprocket system for holding/powering the tread.

While I have not given up (yet), this is proving to be an incredibly costly and in-depth project. We will see what more is to come.

21.1.12 Sunday CAD

I believe I have spent a cumulative hour or two away from the computer starting the minute I woke up at 8 this morning. I have been glued to SWX, organizing team CAD files, designing a solution to the misplaced gearbox conundrum, and reworking old, outdated, and unfinished CAD work. The result is, among other things, one sexy looking chassis....

26-28.1.12

Things are finally starting to come together! Thursday meeting I saw some people beginning to really work on their own on the robot and Friday it took off. Groups were autonomous, fabricating and designing like I've never seen this team work. Friday night I brought home the drive plates I spent so much time designing, and Mr. Reif machined during the week. It was not easy to design, and not easy to machine. The hardest part is probably that after the dozens of hours that went in to designing and fabricating these plates, they have some serious design flaws/issues. First of all, even though I spent a lot of time on them, I was still rushing through and did not CAD all the parts of the drive modules. Because I couldn't see how all the parts would go together in 3D, I placed the mounting points for the gearbox/motor assemblies higher than is feasible. The current placement interferes with the frame and tread.

I worked last night after our meeting, and will continue to work today on a solution. I need to design a sub-plate that essentially moves the mounting holes for the gearbox down. I have the material here to fabricate these plates assuming I am able to finish. Also on the list of things to do today is spending 2 hours or so putting up flyers for our upcoming ewaste fundraiser on February 11th.

Here are a few pictures of the Drive Plate in CAD and the fabrication process...

21-24.1.12

I am home getting some school work done (and writing an entry for my engineer's log) after a long 4 days of robotics work alternating locations between the Robolab and the home/shop of our main technical mentor Adam Reif. While I thought I was finalizing designs Sunday, little did I know. I spent all Saturday night, all Sunday, all Monday night, and all Tuesday working on that same set of CAD files (among a little bit of other big-picture CAD work). I ended with a set of files and drawings that I am very happy with. Mr. Reif pushed me to fix every single bug, design flaw, SWX (Solidworks) getaround, screen-blip, and assumption in my design until I had something I can both be proud of having created, and present to potential employers as prior design work. After revision upon revision, and printing half a dozen differently new/updated drawings, I am not only not sick of working with SWX, my passion is newly stoked. I have spent the evening reading, watching videos, learning more about using SWX to its fullest.

Unfortunately, due to the shear amount of time it took to complete the drive-train "Drive Plate" design, I will not be able to machine the actual part. When I left Mr. Reif's house, we had discussed every nuance of the design, and had cut and laid out the material to be machined. Even though two plates are needed (a right and a left), only one needs to be machined because they are identical so the machinist is able to stack the two and machine them at the same time. While this seems completely straightforward, and even obvious, I had not considered the idea until Mr. Reif mentioned that that would be the most efficient way of going about it.

While perusing the interwebs I came across an incredible reference for all things FRC. From how to weld and where to buy tools, all the way to a hyperlinked list of other robotics competitions; it's all on the Hauppauge Robotics Eagles website under files.

The next robotics meeting is Thursday 26.1.12 from the end of school (the team draws students from 6 different schools so the times are not the same for each person) until 8. I will try to remember to export a jpg or two of the weekend's design work.

21.1.12 Saturday Meeting

Written Sunday 22.1.12 morning:

Saturday morning my mother snuck out of work for 30 minutes in order to get me to the Robolab. The meeting started later then I thought, however I had a jacket and little was lost. Saturday was quite productive for the chassis and drive-train teams. Aron Linker worked on finishing the designs in detail for the drive-train modules, while I finished designing and then fabricated the inner frame. We now have a complete frame, and I am writing this post off and on throughout the same day (Sunday) that I am finalizing CAD drawings of the drive-train for fabrication. I hope to finish machining the parts needed for drive-train today, and have a complete robot chassis Monday 23.1.12.

Here are some pictures I never managed to post of the newest version (1.2) of ÜberTank...

A single link separated from the chain.

A full, 36 link assembly with a sprocket at one end (~8" diameter sprocket).

16-20.1.12 Monday-Friday Meeting(s)

**I wrote this Saturday (21.1.12) morning waiting for the Robolab to be opened and Saturday meeting to begin.

Well I have really been slacking off with regards to logging my robotics work. Almost a full week has gone by. I'm not going to pretend I remember everything that has happened, but I'll do my best to sum up where we've gotten. The Chassis and Drive-Train team (Aron Linker and me) have designed a highly modular drive-train which will enable us to switch between a standard treads and our own proprietary omni-direction treads (Project ÜberTank). The basic frame is complete, with some inner structure for mounting robot componentry the only thing that remains.

14.1.12 Weekend

Wow! It has been a long, roboticcy weekend. After the meeting Friday night ending at 10, my dad and I got up early Saturday morning and hit home depot for emergency prototyping materials. We spent the morning fabricating a proof of concept for Project ÜberTank. The results were rough, but worked. We ended up with a three-link chain, each link with two rollers perpendicular to the axis of tread rotation. The chain/tread has incredible traction, however rolls sideways with very low friction exactly as intended.


At the Saturday robotics meeting, I presented the prototype to the team and spent the day working out design issues with the robot chassis and drive-train layout. All of our major issues were resolved, and we ended with a modular design that will allow for the possibility of both traditional treads, and ÜberTreads.






Here are some CAD screen captures of both version 1.0 and 1.1 of ÜberTank.

 "Linkage Plate" from version 1.0: A plate of aluminum with slots machined for attachment of the roller axles, as well as depressions to prevent slippage on contact with a drive (or floating) sprocket.

Single version 1.0 link.

 Alternate view of version 1.0 link.

 Full assembly of six version 1.0 links.

 Single version 1.1 link.

Alternate view of version 1.1 link.

Assembly of six 1.1 links. In 1.1 the drive or floating sprockets have concave teeth on which the crossbar between each linkage bar can rest. This design should reduce weight and wear due to the fact that the round crossbar will wear significantly less than the filleted depressions in version 1.0. Also, due to the decrease in the amount of material, it is possible to use a lower-wear material for the crossbar such as stainless steel (as opposed to aluminum).

While I don't have any pictures of the fully assembled prototype yet, I plan on taking some tomorrow (Monday the 16th) in the robotics lab. Here are a few photos of the prototyping process Saturday morning.

Dad working on rollers for prototype Saturday morning.

Me drilling assembly and axle holes for prototype Saturday morning.

Stacks of two different aluminum parts for prototype.

Mr. Reif enjoying a cupcake to its fullest at Saturday meeting in the lab.

Up we go

Early start to a long day: its 5:45. During Build, if I want "spare time," this is how it happens. I'm going to go make stuff.

Until tonight...

13.1.12 Friday Meeting

Long day. We had a ~2-10 meeting in which we made some good progress. I was working on chassis and drive-train with Aron Linker, and had some very serious design issues. We had problems due to maximum dimensions and creating the ability to go over the 4"x6" bump in the middle of the field. We are still quite unsure as to what to do, however we settled (pending...) on a very short, wide chassis design with a high-strength, dual purpose "arm" for both tipping the bridge, and for assisting in balance when crossing the bump.

Not much more tonight. I want to get to bed early as I have early-prototyping plans for the morning. More on that tomorrow.

Also hoping to load some screenshots of the CAD files from Project ÜberTank, as well as some truly inspiring footage of our main technical mentor Adam Reif.

12.1.12 Thursday Meeting

I do not have a lot to report on from today's meeting. I started out doing some  CAD work for Project ÜberTank, mostly cleaning up/reworking existing design drafts from Tuesday. I have most of the designs cleaned up so they are not only modeled, but well modeled. I have never had much problem making things in Solidworks, but thinking ahead towards more complicated assemblies etc... has always been a challenge for me. This project is forcing/enabling me to take my CAD skills to the next level by creating a well done, complete set of designs for the project.

After spending some time on ÜberTank I went on to work with Aron Linker on some fundamental design decisions regarding the chassis and drive-train.

I left 2 hears early for a quasi-emergency doctors appointment so I look forward to being brought up to speed tomorrow.

10.1.12 Tuesday Meeting

Today we met from 3 to 6. As a whole, the team spent a good deal of time discussing design (even more), however I was not as involved in that as previously. I focused on Project ÜberTank with Sam Sedgwick. ÜberTank will be the operating name for the development of OmniTreads. We spent most of our time (apart from design discussions) CADing the preliminary design, under special orders of the team coordinator Mrs. Lindborg. We were able to complete a model of both a single tread link, and a working flexible chain of several links. This is a preliminary design and will be refined greatly, however a first pass in under 48 hours from concept to full CAD model is better time than I could have asked for. Due to my involvement in such a covert (sideline) mission, I am hoping to be issued a Solidworks capable computer to enable me to work on such things, as well as other (more central) team related CAD assignments from home. Unfortunately I was unable to prepare screenshots/renderings of the model. I hope to do that during our meeting Thursday (no meeting Wednesday), however I will be leaving early so I don't know whether I will have time. When I have them, I will show them here.

The design (as seen in a basic, incomplete sketch in yesterday's post) is as simple as could be managed while maintaining ease of fabrication and relatively low cost.

No meeting tomorrow, I will post Thursday night after my appointment (for which I am leaving early).

9.1.12 Monday Meeting

Today was my first day at Shenandoah High School. School went by quickly enough, and I was able to get on with robotics. We spent another entire meeting researching and discussing design, we are several days behind schedule now. While this does not sound "good," at the some time it is to be, to a point, expected. If we are able to get through and finish the big picture design tomorrow so that we may come back Thursday and hit the ground building, I believe we will be in good shape. We talked about different types of ball collection devices, ball launching devices, and we dabbled for moments here and there in different possibilities for the drive train. My opinion with the 4"x6" rectangular-profile obstacle running across the middle of the field, treads may be our best option. In the 3 regional competitions I've been to, I don't remember seeing any group of 5 or 6 winning teams in which there were not at least a pair of treaded 'bots. It provides incredibly effective traction, mobility, and in the case of the obstacle in this years game protects any components mounted to the underside of the robot. For my homework tonight I did a couple of sketches. I sketched the parts needed in a semi-fantastical idea I had to create OmniTreads: tank treads with the holonomic aspect of omniwheels (not mecanum wheels). I also diagrammed a few key points regarding a collector consisting of a 2" spinning "roller' with sticky, compliant foam wrapped around to suck in any basketballs that are unfortunate enough to come it contact with said device.

The closest things to decisions we made today were eliminating about 95% of the possibilities for different types of collectors, elevators (a mechanism to bring the ball from ground level where the collector is up high to the launcher), and launchers. Tomorrow we still have to finalize decisions regarding those, as well as discuss our drive train.

8.1.12 Sunday Meeting

This morning I woke up to my mother saying, "Avery, maybe you should get up..." When I asked what time it was and she answered 10 (the time robotics was starting this morning) I amazed both of us with my speed. In the lab we decided upon a final game strategy. Personally I am very happy with where the team ended up. While our decision is far from the strategy that I initially envisioned, that is far from what is important. What I believe is important is the fact that the team has, with some time and effort, unanimously agreed on a game strategy that is (once again, subjective here) simple enough for the team to successfully build a robot that could potentially dominate. We decided on a robot that will have the ability to rapidly gather (foam) basketballs from the ground and fire them to the other side of the field, as well as have some significant thought put into balancing on the appropriate alliance bridge in the middle of the field.

Tonight's "homework" was to consider and sketch some different ideas/designs for a gathering mechanism and a firing mechanism. I skimmed through a book on the 2006 FRC game which in some aspects was similar to this year's and settled on a firing mechanism that was discussed today in the meeting as well. I sketched that up very quickly, but spent most of my time brainstorming the strengths, weaknesses, and possibilities with a system like that. I also put a few ideas on paper all together in a detail-free semi-comprehensive sketch of a robot chassis with a gathering mechanism roughed out.

Monday school starts and after school we have a meeting from around 1 until 6.

Kickoff

Well, today was the big day. We got up at 5am and trucked all the way down to St. Francis to discover the nature of this year's FRC. I am VERY excited about the game this year(Rebound Rumble). A derivative of basketball, I believe the 2012 competition will be an incredibly intense and difficult season for even the most seasoned veteran teams. It seems after 2 decades of hard work, FIRST wants to push us even harder!

Actual Kickoff went quite well, and the brainstorming session held in the 3189 RoboLab also went well for the most part. Tho good about today's meeting was that a LOT was said and the structure that was put in place by last year's Task Force did a great job of forcing the team to learn the rules and understand the game before we even started discussing possibilities for the season and robot. Unfortunately I feel some of the other work done by the Task Force is being partially ignored by the majority of the team. The conclusion made by the Task Force over last summer was that the greatest source to date of 3189's failure is our tendency to try to create a robot that is capable of doing everything on the "checklist" of tasks FRC gives us each year. The Task Force also concluded that the "solution" is to choose a select few functions for the robot to perform, and perform well. At the end of today's meeting the team was split with the majority leaning towards a multi-purpose robot: exactly what the Task Force recommended against. The final discussion will happen tomorrow morning when people are rested and less irritable. Time will tell what game strategy the team chooses.



Without watching the Game Animation (above) or reading the Manual, it is difficult to really understand how the game works. It is played on a 27ft by 54ft field, with 4 basketball hoops on each end of the field at 3 different heights. The MS Kinect sensor seems to be legal in virtually any use or configuration. From usage on the 'bot to as a controller, FIRST really wants to let team creativity rule all with this particular donation. Apart from the Kinect, the Kit seems to be pretty standard. It appears the included wheels have changed this year, and the First Choice program (used by teams to purchase additional parts with points allotted by FIRST) is back for a second year in what appears to be a much more streamlined version 2.0.

Tomorrow (Sunday) all-day meeting will be used to
     1) Finalize game strategy (was supposed to happen today).
     2) Inventory KoP (Kit of Parts) in order to ascertain whether a missing parts request must be sent in.
     3) Develop (Discuss, evolve and finalize) overall robot design