All posts by David Black

Day #40 – Off to powder coat!

I arrived back at the lab just a short 4.5 hours after I left, and loaded the elevator frame, intake support frames, intake arms, and carriage plates into the Suburban.  After 45 minutes in traffic, I arrived at the powder coater and dropped off the parts, turned around and sat in 45 minutes of traffic back to the school.

This afternoon we received the orders of additional bearings for the elevator, BaneBots green 4 7/8 wheels for the upper portion of the intake, and air cylinders for pivoting the intake.

Due to the really late work session last night, not having the upper frames today, and the exhaustive list of work remaining to be done over Thursday through Tuesday, we took today completely off to catch up on rest, and we did not perform any FIRST Robotics work other than what is mentioned above.  We’ll come back and hit it hard tomorrow.

Day #39 – 2/10/15 – Round up the ponies

Lab opening time: 1:30 PM

We began the day by unpacking newly-received parts orders and organizing the components according to subsystem and hardware type.  Organization now is critical to ensuring an efficient assembly process over the next few days.

After sorting the parts, we programmed, set up, and ran the drilling and tapping routines for the clevis pieces for the intake arms.  We then machined the intake radius arm bearing holes and drilled the holes in the crossmember for the VersaPlanetary gearbox.  Finally, we machined the four four-hole plates to mount the pneumatic cylinder to the intake, and proceeded to weld together the device while measuring very precisely.  We noticed a critical design flaw in the intake this year, that is something we didn’t catch originally.  The gear mesh depends on a weldment, and powder coat thickness.  Those two things alone are tough to hold a decent tolerance on, and together, give concern that the gear mesh will not be optimal.  We did our best to weld it in a very accurate location while accounting for powder coat thickness.  We’ll see how it turns out.  Additionally, after welding each half of the intake, we determined that the rotating frame pivots do not fit into the stationary frame clevises when assembled.  We remedied this by putting the whole welded intake frame into the Haas Mini Mill and remachining 30 thousands off each of the machined pivots.  It now fits with plenty of free play to allow for easy assembly.

We rounded out the night by test fitting the elevator carriage with all the bearings and shoulder bolts installed.  After tack welding the carriage crossmember in place, and performing a second test fit, we noticed a lot of slop, amounting to approximately 1/16 of an inch, which would cause rather severe side-to-side rocking of the assembly.  Rather than attempting to weld the crossmember to a wider width, we decided to test fit again after powder coat, and laser cut a shim if needed to tighten the fit of the carriage in the elevator frame.

During test fitting of the elevator, we made frequent use of our new “Pony” brand C-clamps, which are quite nice actually.

During the fitment test, one concern that arose was that the elevator frame may bow outward slightly in the middle.  To remedy this issue, we are considering adding a cross brace in this area.

Today’s work was so fast paced and extensive that unfortunately, we did not bother spending even a just moment to take any pictures.  Everything took us much longer than anticipated, as indicated by the closing time below.  The good news is, everything is ready to go to powder coat in the morning.

Lab closing time: 3:30 AM on 2/11/15

Day #33 – 2/4/15

Lab Opening Time: 3:00 PM

Today again was not an official work day, but a few of us proceeded to see how much progress we could make in the absence of others while we were able to focus on a few specific tasks at hand.  Significant achievement was seen in wiring, assembly of the swerve modules, and manufacturing of the elevator bearing plates.  The modules are approximately 70% assembled, the 12ga zip cord is installed from the PDB to the Victor mounting areas, and the left side elevator bearing plates are complete.  We did face some difficulty in manufacturing the bearing plates.  The operation involving two nested parts comes within 0.5 inches of maxing out the Haas Mini Mill’s travel at 12×16 in the XY plane.  Additionally, an accumulation of small oversights and errors in depths and feed rates led to the 1-hour manufacturing process on these pieces to take several more hours than they should have because of frequent re-posting of the G-code with minor edits each time.

Here are some nice pictures we took of the wiring of the PDP

Best

IMG_1543

Lab Closing Time: 1:00 AM on 2/5/15  (Late yes, but we were determined, and having a good time making chips on the CNC!)

Day #32 – 2/3/15 – Exothermic Event!

Lab opening time: 3:00 PM

Although today was not an official work day,  a few members of the team were present to continue work progress in preparation of the frame’s return from our powder-coating sponsor.

While today started with modest ambitions, it ended in what could be described either as a lucky save or a devastating loss.  Further analysis in the coming days will determine whether the outcome is the former or the latter.  Whilst laser cutting the Victor SP mounting plates from 1/8″ Delrin, the material ignited, and flames quickly grew to a point where a determination was made to discharge a fire extinguisher on it.  Unfortunately, the only type of extinguisher available was a dry-chemical ABC type extinguisher.  The the fire-suppressing powder within is not a favorable substance to deluge into an expensive machine containing precision optics, as was made quickly evident by the outcome of the event, as shown in the photo below.

Lazer Cutter Aftermath

The expedient efforts to extinguish the flame prevented potential further damage due to combustion, but we now had quite a mess on our hands, and potential damage to the machine due to the fire suppression effort.  Holding knowledge that the dry-chemical discharge from fire extinguishers is corrosive to many types surfaces, work began immediately to clean the machine and room of all powder and residue, inside and out.  This task required a tremendous amount of effort and attention to detail, but students handled the unexpected cleanup task with a professional attitude.  We began with a wipe-down of all surfaces with a damp cloth, and partial disassemble of the machine.  We proceeded to vacuum out remaining powder in areas the vacuum could reach, and blow out the remainder first with compressed air, then with a mix of compressed air and Windex.  The machine appears to be in decent shape after such a traumatic event, but while the cleanup is going well, further work will be needed to remove some remaining residue.  It’s not yet certain how the optics including beam window, lens, and gold surface mirrors fared.

We attribute the incident to a combination of factors: the air assist needle valve was closed, and the fume extraction pre-filter was partially clogged.  These two factors in combination allowed for vapors to accumulate in the direct vicinity of the ignition source.  Furthermore, Delrin is not self-extinguishing.

After sufficient cleanup efforts, we held a brief discussion on how to proceed with the rest of the week.  We determined that we would not complete the elevator and intakes before Friday, so a new goal for completion was set for Sunday, with anticipation of sending those systems to powder coat on Monday 2/8/15.   So, instead of starting on the elevator, we began to assemble the swerve modules and install the main 4ga wires from the Anderson SB50 connector to the power distribution board.

Lab Closing Time: 10:00 PM

Day #19 – 1/21/15 – Small Team Meeting

Lab Opening Time 2:00 PM

Although today was not an official work day, Jack, Cynthia, and I spent a great deal of time working on the robot’s frame layout, electronics arrangement, and wire routing.  We determined that it would be very difficult for the Victor SP motor controllers to live near the power distribution board, and it would be in our best interest to move them closer to the swerve modules.

Additionally, we finished both versions of the swerve top plates.  We manufactured a quantity of four of each, which is double the quantity needed for one robot.  A picture is shown below:

Swerve top plates, made from 6061-T6 plate.  Eight tools.  Cycle time: 13 minutes each.
Swerve top plates, made from 3/16″ 6061-T6 plate. Eight tools. Cycle time: 13 minutes each.

Lab Closing Time – 6:00 PM

Day #18 – 1/20/15- A Trip to Autodesk

While 1/20/15 was an official day off for Team 696, I’ll share  my summary of the day.  My day started quite early, heading out to LAX at 4:00 AM, where I caught a 5:45 AM flight up to SFO with our school’s principal Doug Dall and animation teacher John Over.  The goal for the day: visit the Autodesk Gallery.  I must say, it’s quite an impressive place.  I’d highly recommend arranging a visit if you’re ever in San Francisco.

On display were several applications of various Autodesk software products in areas ranging from art to biomedical engineering to architecture and automotive design.  A picture of a  recyclable and sustainable concept car from Mercedes is shown below:

Full-scale model of a Mercedes concept car.  As I understand it, the car's structure would be "grown" from a bio-fiber material.
Full-scale model of a Mercedes concept car. As I understand it, the car’s structure would be “grown” from a bio-fiber material.

 

A favorite of mine at the Gallery was a high-reduction gear train, with a ratio that I estimated to be approximately 250 million to one.  At the rate the motor was spinning, it would take over two years for the vise handle to rotate once.  Needless to say, it will be quite some time before this mechanism crushes itself.

12 stages of gearing at approximately 5:1 per stage.  For every one rotation of the first gear, the last gear moves approximately only 0.0000015 degrees.
12 stages of gearing at approximately 5:1 per stage. For every one rotation of the first gear, the last gear moves approximately only 0.0000015 degrees.

Even more productive than the gallery tour was our meeting with Autodesk’s educational director.  We were very pleased to learn about the incredible software licenses and training opportunities being made available to students and schools.  We are currently in the process of determining which particular software titles to include in the system images for the new computers to be installed in our animation and engineering labs.  Our hour-long meeting is hopefully just the start of great things to come from our partnership with Autodesk.

All I can say is, students, you’re really doing yourself a disservice if you haven’t at least taken a look at the incredible value in software made available to you for FREE through the Autodesk Education Community.

As a final note, I thought I’d leave you with these two posters I saw in the BART Station.  They send a strong and important message.

What_sets_you_apart

Day #12 – 1/14/15 – Hats Off to CAM/CNC

Lab opening time: 1:30 PM

While today was a non-scheduled day for most of the team, the CAM and CNC groups met to program, set up, and run the first machined component on the Haas CNC mill, the swerve drive “Top Hat”.  While the programming took approximately two hours and the setup took about 30 minutes, the part was set up and run without issue, and completes in approximately 4.5 minutes.  Tomorrow, the team will program, set up, and run the second operation to face off the back, verify bearing fitment, cut bar stock, and machine six more.

Claire offsets a 1/2" endmill to a dial tool-height setter on the Haas CNC Mill.
Claire offsets a 1/2″ endmill to a dial tool-height setter on the Haas CNC Mill.
Swerve drive "top hat" - 1st operation.  3.5" bar stock, 4 tools, 4.5 minute run time.
Swerve drive “top hat” – 1st operation. 3.5″ bar stock, 4 tools, 4.5 minute run time.

Lab closing time: 7:00PM

Day #11 – 1/13/15 – Rest Day

Tuesday January 13th was a full day off for the entire team.  Nonetheless, we accomplished some minor objectives such as determining stock sizes for parts to be machined later in the week, ordering USDigital MA3 encoders, and visualizing workholding and operations for the machined components.

Day #5 – 1/7/2015- “Secret” CAD Meeting

Lab opening time: 2:00 PM.

After an intense four days of prototyping, field building, strategizing, CAD design, and brainstorming most of the team took today off with the start of second semester at school.  However, a select few members still worked on furthering the CAD design of the swerve module and assembling the VEX Swerve drive prototype, using  parts that arrived today.

On the actual robot’s swerve module design, after three revisions, we got the bearing blocks properly modeled and finalized.  Also, the axles are now modeled properly.  We revised the design of the large steering gear to allow mounting of the bearing blocks, using low profile socket-head cap screws.  We discussed and analyzed the mounting of the “tube” piece to the 84T steering gear, and determined that six #4-40 screws would be the best we’d get given the space constraints with all other factors considered.  It’s the smallest size screw we’ve ever used for something semi-structural, but “back-of-the-napkin” calculations indicate the assembly should hold together without issue.

Everything in the swerve drive module is a very close (<1/16″) fit in relation to everything else.  It’s a little scary, but so far it’s working.  However, we don’t yet have sufficient confidence in the design to make the parts order.  We’ll take at least another day before we get to that point.  We have our fingers crossed that nothing goes out of stock.

On tomorrow’s agenda is design of the holes to attach the tube to the steering gear and the top-hat, and revision of the tube and center axle for mounting of R4A radial bearings.  Hopefully we can also design the top-hat piece.

Friday’s CAD work will include design of the static mounting plates that couple the swerve module to the frame, and provide mounting for the Mini-CIM drive motor and the VersaPlanetary for steering.

We hope to finish design of the swerve modules by the end of Saturday.  At this point, we are still confident in our ability to finish the swerve drive completely by the end of the month.

Construction of the VEX swerve robot currently stands at one complete module assembled to the frame.

Lab closing time: 6:00 PM 

After-dark work: FIRST Choice Order Priority List.