Through the Flames and Fire (Pt II)

(Wait, that doesn’t seem quite right…)

Late last year, the young Combat Robotics Team made a mistake and had a minor “unexpected thermal event” (FIRE).  Until now, we had not posted anything about the event, because we are not particularly proud of what happened.  However, we learned a lot about responsible design practices and how to react in an emergency.  I am writing these posts in the hopes that you can benefit from our experience without having to undergo your own robot-ravaging accident.

Part II of III : What We Learned

This unexpected thermal event taught us several important lessons.  Obviously, we were doing some things wrong, and those caused the accident.  However, the fire also revealed several things we were doing right without realizing it.  By drawing on both of these, we have learned and grown, and do things differently now.

The things we did wrong are probably obvious, but they are worth restating.  The condition of The Blender’s electrical system had been allowed to deteriorate to a dangerous level.  A horde of zip ties and quite a bit of electrical tape held the wires in place, but goodness only knows the last time a proper inspection had been performed.  If we had found the damaged wire earlier, the fire may have been prevented.  Additionally, we should not have been operating The Blender in such tight spaces with the shell attached – an active shell obscures the manual kill switch.  As we have now learned, the shell has the potential to take on a life of its own.  If the drive motors had also activated, a costly situation could have become a deadly one.  This would have been compounded by our final error: standing far too close to The Burning Blender.  Ten feet, or even 20 feet, afford only seconds to react if control of the drive motors is lost.  If the shell is also active, severe injury can result.  Most of these mistakes originated from too much faith in The Blender performing as instructed.

While our mistakes were significant, we also handled ourselves quite well once the fire began.  First, we remained calm.  There was no screaming and running in circles.  We remained calm, and assessed the situation.  Due to UTDesign’s online training courses, Haley knew where the fire extinguisher was and retrieved it.  Thanks to the same courses, she knew how to use the fire extinguisher to maximum effect.  Meanwhile attempted to minimize the danger to ourselves by backing away and elevating ourselves (protecting our legs from The Blender’s shell).  Once we had contained the incident, we called the building manager, who contacted the appropriate emergency services.  By staying calm and taking appropriate actions, a bad situation was prevented from becoming much, much worse.

Since the incident, we have learned from both what we did right and what we did wrong and adjusted our practices to prevent such an event in the future.  For example, we use a multimeter to check the integrity of the electrical systems before every time powering up The Blender.  During power-up, The Blender is carefully restrained to prevent unintentional movement.  During demonstrations and testing, we never drive The Blender with the shell on.  This prevents the dangerous situation that exaggerated the danger of the fire – a spinning shell prevents access to the mechanical killswitch.

More than anything, we have learned to respect and fear the power of The Blender as an electromechanical system.  When watching videos of the arena or performing maintenance, it is easy to forget the danger The Blender poses to us and other people.  However, the fire taught us otherwise, and we have adjusted accordingly.

I hope you have found this post informative.  In the next post, I will relay some information that you should know to help prevent such disasters happening to you.  If that fails, I hope the information helps minimize damage from such unfortunate events.

Through the Flames and Fire (Pt I)

(Wait, that doesn’t seem quite right…)

Late last year, the young Combat Robotics Team made a mistake and had a minor “unexpected thermal event” (FIRE).  Until now, we had not posted anything about the event, because we are not particularly proud of what happened.  However, we learned a lot about responsible design practices and how to react in an emergency.  I am writing these posts in the hopes that you can benefit from our experience without having to undergo your own robot-ravaging accident.

Part I of III : What Happened

IMG_20150923_185949

On September 23, The Blender was going for a test drive.  Our Blender Promotional Video was set to be recorded on the next Saturday, September 26, and I had been assigned to determine several statistics on The Blender for use in the video.  While max rotational speed, max kinetic energy, and max tooth speed were somewhat simple to calculate, I was hitting a roadblock on max linear speed.  I decided that a direct test would be simplest, so Haley and I cleared some space to drive The Blender around.  Jack manned the controller, and we hit full speed ahead with the spin motors deactivated (for safety!)

After a few runs, we were getting results around 15 mph, but we wanted to do a longer test to be sure.  On a longer run, The Blender veered off to the right and bumped into a table.  All things considered, it was a light hit – the table took no damage.  However, a small flame emerged from inside The Blender’s case.  Just as fast as Haley could retrieve a fire extinguisher, the flame grew larger and the shell spun up to full speed.  We climbed on chairs, and Haley regularly and liberally doused The Blender with the fire extinguisher until, after several minutes, the shell slowed and stopped.

The clock was now ticking.  The footage for the promotional video was going to be recorded in two days.  The Blender had experienced serious fire damage, and many components had been either destroyed, damaged, or filled with fire extinguisher dust.  Over the course of those two

wire responsible for fire - offending post has been removed

days, in the midst of midterms, our team of five people completely disassembled The Blender.  Everything had to be stripped down to its smallest components, cleaned, analyzed, and repaired or replaced as necessary.  The reconstruction was done in time for the video shoot – but the team was left with no spare parts and no funds – a state in which we have remained for several months.

Once the dust settled, we figured out what had caused the fire.  A loose wire had scraped against an interior post, eventually removing all of the insulation.  The slight bump from hitting the table swung the wire until it touched that post.  The chassis had been accidentally grounded, so 24 volts from the batteries rushed through the negligible-resistance wire and into the shell with extremely high current.  High current caused high heat – enough to instantly weld the copper wire to the shell.  As the wire remained on the shell, current continued to pour through, the temperature continued to rise, and the fire started.  The extreme current caused the batteries to vent oxygen, exaggerating the situation.

The extreme heat caused several other components in the robot to fail, including the motor controllers.  When our particular kind of motor controllers fail, they fail “closed,” meaning the circuit is complete and current passes through.  Because of this, the spin motors activated, and The Blender sprang to life even as it was consumed by the fire.  Thankfully, there were no teeth on the shell and the drive motors did not activate.  Otherwise, an unfortunate and costly accident may have become something else altogether.

I hope you enjoyed the story of our accident.  In the next post, I will discuss some of the ways we do things differently since the fire.  While The Blender itself has changed little, we have developed a variety of cautionary measures for maintaining, inspecting, and displaying The Blender.

Outreach: FLL North Texas Regional Championship

This past Saturday, our team had the opportunity to volunteer at a FIRST Competition.  FIRST, For Inspiration and Recognition of Science and Technology, is an organization which encourages K-12 students to explore robotics.  The FIRST Lego League is for grades 4-8 and involves the construction of Lego robots to complete various tasks.  The students are given a Lego Mindstorms set and are responsible for the design and programming of their robot.

The event, which was hosted at Parish Episcopal School, was the FIRST Lego League North Texas Regional Championship.  This event marked the end of the season for many teams, and exhibited the height of development for their designs.  We were invited to bring the Blender along as an exhibit for the young roboticists.  Once we set up the exhibit area, Haley, Jeremiah, and Morganne were selected to be judges in the competition. I was left to explain the Blender to the bright young minds.

I was amazed by the amount (and depth!) of interest that the FLL teams showed.  For 12 hours, the elementary school students kept me on my toes.  I explained everything from how our competition works to the inner workings of the Blender to why attaching flamethrowers to our shell isn’t the best idea ever. Every hour or two, I would perform a (shell-less, for safety,) test-drive of the Blender.  The crowds that gathered would easily keep me busy for the next 20 minutes.

Haley wanted to say a little bit about the rest of the team’s time judging:

As someone who was part of various FIRST teams for five years, I was really curious coming into this what judging would actually be like.  It turned out to be a lot of fun and something I’ll definitely do again.

I was assigned to Core Values judging, while Morganne and Jeremiah went into Design judging.  Every fifteen minutes for the first half of the day, I’d meet a new team who would show off how well they work together with a team activity and then answer questions about Coopertition, Gracious Professionalism, and teamwork.  I was blown away by their enthusiasm and I loved seeing that FIRST was as awesome for them as it had been for me.

One thing that really stood out to me was the fact that through working together, a lot of the teams emphasized that they had made new friends that they wouldn’t have known otherwise.

20160206_173927

Morganne would like to remind you that you too can get involved! FIRST is always looking for volunteers to help with a variety of events. To learn more click here to sign up. Both technical and non-technical volunteers are appreciated!

Finally, a huge thank you to Parish Episcopal School for hosting the event.  It was a wonderful facility, and the staff was super helpful.  Thank you to all of the excited FIRST kids.  Additionally, thank you to FIRST for inviting us.  We had loads of fun and hope to do this again soon!

 

Thank-You to Our Sponsor: Dassault Systemes

Thank you to our sponsor, Dassault Systemes!

Our software request applications have been accepted, and Dassault Systemes has donated 6 Solidworks licenses to our team! Solidworks is the CAD program of choice for our team. This incredibly powerful design and simulation tool makes managing the Blender much simpler, and expedites forward progress. We look forward to making the most of this generous donation. Whoosh!

Make-a-Thon

Over this past weekend, our workspace at UT Dallas (the senior design building) hosted an event known as the Make-a-Thon.  Students are given 26 hours to complete a challenge, and then after those hours are up their project is assessed by a panel of judges.  This leads to some interesting all-nighters and a really good opportunity for some team bonding, so that’s what we did!

The challenge was to improve the lives of senior citizens.  We chose the idea of a chair that helps you sit up, since that was one of their complaints.  This way, the seniors in our community won’t end up like poor David:

20160129_195507

We planned our design, and started out with a really terrible 10$ chair that the guys picked up from surplus a few months ago.  Then we ran to the store and bought the necessary wood and jacks which became an important part of our design.  Lance built a really sturdy base for the bottom of the chair to stand on at the correct height, and then David attached the chair to the jacks and the jacks to the base.  This is what the mechanical stuff on the chair looked like, once it was done:

image

We completed that portion of the project in the early hours in the morning.  We had yet to put on the sprockets, chains, motor, and other electrical components which Jack had been working on.

image

Jeremiah worked hard to get the chain and sprockets to fit together onto the jacks.  This was not a trivial task, since he was literally trying to fit a rectangular peg into a circular hole.
image

Meanwhile I tried to get the motor to stay on the same plane as the constantly-moving shafts on the jack.  I ended up making a shelf that attached to the shafts themselves and zip-tied the motor onto it.

image

Lance, meanwhile, finalized our presentation which we would soon show to the judges.  A panel of senior citizens and engineering professors would judge us against several other teams with other ideas.

image

Finally, with about two hours to spare, and almost everything attached except for the chain and the control panel, I decided perhaps it was time for a short nap before we had to present.  David and I each wandered off to find couches in the building, since neither of us had slept yet.  (Lance went home for a while around 3am, Jeremiah took a few minutes on a couch around 5, and Jack fell asleep while taking a shower that morning.)  The other three teammates were putting on those finishing touches.

About 30 minutes later, I got up and wandered back to our work table.  Apparently, while I was asleep, the components had been completed and we finally tested the chair.  And the motor burnt out, so we were without chair power.

This is because we tried to use an old drill motor on it to essentially lift a person.  It was geared down significantly, but still not enough.  We were now stuck with a chair on jacks, and that’s about it.

So, for the presentation, we ended up using a more powerful handheld drill to move the chair down with a person in it, then pulled off the sprockets and used the hand cranks to lift it back up.  It wasn’t nearly as impressive as it was meant to be.

Despite our last-minute demise, we still had a great time and I’m glad we signed up for this. Shout-out to Neel and Morganne who were already signed up for different teams.  Congratulations, also, to Morganne’s team, that took first place with their 3D printed pill organization system!  Also special thanks to Rod Wetterskog and Andrea Turcatti for hosting and organizing this event.

 

 

Thank-You to Our Sponsor: Texas Instruments

Texas Instruments recently donated 75 MOSFETs for our custom motor controllers designed by Jack, our Electrical Lead. We’re excited about this new partnership and would like to thank TI for their generosity!

IMG_9781

 Woosh!

How to Charge the Batteries:

IMG_9747

  • Check voltage:
    • if >20.5V:
      • Set charger to NiCD – 0.1A – CHARGE
      • Charge and watch until voltage is <20.5V
    • Set charger to Li-ION – 3.0 – CHARGE
      • Charge until <24V minimum, 28-30V preferred
      • Max voltage: 32.8V, Caution: if voltage is exceeded, may cause fire
      • rinse, repeat. The chargers are picky and will turn themselves off and beep. Ignore this, and continue charging.
    • During charging, periodically press to monitor cell voltages.
      • A cell less than 2V is bad and should be recorded on the battery label
      • A cell greater than 4.1V needs to be discharged to prevent cell failure
  • Because our batteries suck, they cannot be balanced. Just, be careful, watch the voltages, and try to keep them from getting too disparate. If they do, set the chargers to discharge and bring them down to 24V and re-charge them up.

 

Created by Jack Doan, Revised 3/23/15

Motor Controllers

This weekend, Jack and Jeremiah’s kitchen became a workstation for the new motor controllers! They’re really starting to come together.

image

Then Jack and David needed to connect the setup to the oscilloscope on the desk. The obvious solution was to string the oscilloscope probes across the hallway to the front door.

image

Personally, I think they’re doing a great job and admire the dedication it takes to work on the same project for basically an entire weekend. I owe these guys a pizza for sure.

Dallas STEAM Explorers

P1240263 (2)

One thing our team values a lot is getting middle school and high school students interested in robotics! We had the opportunity to present to the Dallas Steam Explorers on October 5th. We appreciated your enthusiasm and your questions!