Steel and Flesh

Wherein Chris Jurney, a video game programmer, attempts metal projects and severely damages his body and you watch in morbid fascination.

[steelandflesh]

Waterjet Toolpaths

Cutting parts on the waterjet takes some prep work.  I already posted up the vector art of the parts, but cutting those shapes out of 16 guage stainless steel requires converting to a toolpath.  That takes place in 2 steps.

First is importing the Illustrator file into FlowPath.  FlowPath is a tool that lets you decide which parts of your vector art are inside the part vs outside.  It also lets you draw enter / exit paths on each cut to avoid ugly marks at places the cutting head starts and stops.  Finally, it lets you set traversal lines between your cuts to make sure the fragile head doesn't bump into anything you cut out when moving to the next cut.  Here's what the file looks like after pathing:

The arrow next to each line is showing which side will be cut.  The jet of water has a width of ~0.045 inches, so getting the side wrong will result in an incorrectly sized cut.

After the path is generated, next up is to cut it.  Before running the file on your actual metal it's a good idea to do a preview run in the cutting software (FlowCut) to make sure the tool will do what you expect it to.  Here's an action shot of the test cut mid run.

Red lines show cuts, greens traversals.  Circles are places where the jet of water and garnet will start or stop.  You can see some places mid-line where there are stops... those are called tabs.  Those are small gaps in the cut left in intentionally to prevent the part from lifting off the table from the splash while the cut is running.

One more thing FlowCut will do for you is give a time and cost estimate.  Because of the expense of maintaining this huge tool, running the pump costs $2 a minute.  The part should be done in about 35 minutes.  Next step is to lug the epic sheet in and do some damage.

[steelandflesh]

Reviving the Sammie Wagon

It wasn't looking too good for the Sammie Wagon.  After coming back from Burning Man, having served 800 people or roughly 1.5% of the event, it was looking rough.  The playa dust didn't do much to it while at the event, but once that dust got wet back in the moist San Francisco air, it rusted every inch of the frame solid orange.

All that rust combined with the fact that Carrie and I won't be on the playa this year meant that I had plans to turn the wagon into a project moving push cart to get back and forth from Burning Man.  Enter Matt Franklin.

I floated the idea of some surrogate sandwich supervisors taking over the operation of the cart this year, at Matt took the bait.  The first step in fixing it up was skinning it and removing the rust.  We're about halfway done in this shot of Matt using an angle grinder on the frame.  You can see how badly rusted it was before on the rear vertical bar.

After stripping the frame and cleaning it with some nasty chemicals, I powder coated it clear to protect from future damage.  After 10 minutes in the oven to cure the powder into a hard candy shell, here is the result:

[steelandflesh]

Vector Art

The next step in making the Divekick controllers is to waterjet out the pieces for the skin from my big piece of stainless steel left over from the surgical cart project.  Layouts for waterjet are generally done in 2d vector formats, so I worked in Illustrator.  Dave Lang, chief hype man of Iron Galaxy, was kind enough to send over some vector logos to use as a basis.  Here's the front plate pattern.

It looks simple now, but when it's done, the Divekick logo cutout will be backlit through acrylic, and there will be 2 big 3.5" backlit buttons sticking through those 2 holes.  The plates will also be riveted down to the frame.  The little nubs on the sides of the button holes are the registration marks that keep the buttons from spinning.

That panel is part of a full sheet layout containing 5 sides of the controller.  I'm leaving the bottom open to give access to the electronics and the socket for the stand mount.

[steelandflesh]

Finishing the Frame

Second day of the Divekick controller build was spent finishing up the frames for the body of the two controllers.  I started welding them together with TIG again, but there are a lot of inside corner welds to do with the frame in the way, so fitting the torch, the filler rod, and my head all around each point was enough of a hassle that I switched to MIG halfway through the evening.

Here's the fancy clamp arrangement I used to hold everything in place for welding.  Everything flat got clamped to the table.  Everything vertical got clamped to right angle bars and pinned to the table with welding table pins. Those are the things in the bin at the top right.  You push them through the table, then tighten an inner allen bolt to force ball bearing to expand out holding the pin down, then you spin an outer knurled nut tighen your work down to the table.

After all the welding came the grinding to smooth it all out.  Here's the finished pair, ready to be skinned in shiny shiny stainless steel.  Just behind you can see where the next step will be performed using the power of 55,000 psi water (almost 4x more pressure than the bottom of the Pacific) and garnet powder.  Next time: art.

[steelandflesh]

Divekick

New project time.  The most epic fighting game ever made is going to launch this summer, and I want in on the ground floor with a custom controller.  The game is hugely epic and absurd, so I wanted to make a controller that would match.  Thus was born the steel Divekick controller project.

My plan is to make a pair of sturdy mild steel frames out of 3/4" square tube, sit them on mild steel pedestals, and skin them in waterjet cut 16 guage stainless steel.  I'm guessing the final products will end around 30 or 40 pounds each.  The pair will be lit from the inside, and will daisy chain together and connect up to a PC.  Iron Galaxy willing, they will also include a build of the game, so all you'll need for a tournament is these 2 controllers and a gaming PC and you're ready.

Here are some progress shots from day 1.  The big metal cutting bandsaw at TechShop was down, so I had to use a circular saw to cut down my tubing.  The circular saw was missing the length stop to let me measure consistent repeat cuts, so I had to improvise.

Once the bars were all cut to length, including a 14 degree angle cut on the top pieces, I did a test layout to make sure everything fit together.

Next, I clamped all the pieces down and TIG welded them together.  Why TIG when TechShop has MIG welders setup for mild steel handy?  MIG is boring and easy.  TIG people tell me it takes 1000s of hours to become an expert, so might as well start now.

Here are the final side frames, welded and ground.

So far on this project, the toll is: 30 feet of steel tubing, 0 injuries.

[steelandflesh]

Cart Final Welding

Time for final assembly on the cart.  The last piece to weld on is the railing around the top.  This time I had help in the for of Zack.  Zack is larger than the available safety equipment:

This time on Steel and Flesh, a pro tip!  TIG welding stands for Tungsten Inert Gas.  That means you are welding with a rod of tungsten that is not consumed by the welding process, and your weld is shielded with an inert gas like Argon to keep it from instantly oxidizing from exposure to air under high heat.  The tungsten is mounted into a very delicate little handle that usually doesn't heat up unless something has gone wrong.

MIG stands for Metal Inert Gas welding.  MIG is like a hot glue gun.  The electrode is the same as your target metal, and is consumed into the weld in spurts as a spool of wire feeds through the gun.

It turns out that if you accidentally pick up a piece of steel that has been clipped to just the same length as your tungsten rods, and then you mount that into your TIG torch, you can do MIG welding with it for about a quarter of a second.  INNOVATION!

(don't do this)

Here, Zack is getting excited because we're almost done.  We cleaned up the welds first using a 36 grit zip wheel and then a 120 grit zip wheel.  There were a few seams that I wasn't able to weld cleanly, so we filled those with epoxy to keep fluids out once it's used in surgery with blood flying all over it.

Here's the finished cart.  It's massively over structural, and cost about 3 times more than a premade cart would have, but that's not really the point, is it?

Finally, here's a glamour shot of the cart seeing the lovely sights of Howard street at midnight on the way to pick a ZipCar to take it home.

[steelandflesh]

Finishing Stainless Steel

So the way to properly produce a super shiny object made with stainless steel is to purchase it pre-shined from a supplier, and leave on the protective plastic for the finish side so it's still shiny when you finish.  I did not do this.

As always with metalworking, plan B is grinding.  I recruited an assistant, Aaron, for some double cart grinding action.  He's using a sheet of 120 grit sandpaper mounted on an angle grinder to get it vaguely smooth.

In addition to the grinding, I did some of the finish welding.  Since I punched circles in the shelf plates for the posts to pass through, the leftover from that hole is the perfect size to cap each pipe.  Here's the before...

Here's the after.  A little rough, but sealed and functional.

Grinding will fix everything.

[steelandflesh]

Assembly Welding Begins

Ok, I expected to next be posting about the teardown of the Sammy Wagon into a project transport cart, but instead I kept lugging this cart back and forth by hand.  So in lieu of that, here's my progress on the surgical cart.

Next step is to weld the vertical posts into the lower shelf frame.  I'm slowly learning what all the fancy clamps and work-holding gear for the welding tables do.  Here's some simple table clamps holding the shelf down before I start welding on that post of the right.  The post is clamped to a large right angle bracket with 2 angled cups mounted do it.  The rig is keeping the post at a perfect 90 degree angle, and in just the right spot so it doesn't scoot around or warp while I weld it in place.

How did I get the post to the exact right height to produce a lower shelf exactly 12 inches off the ground?  SCIENCE!  Check out this badass rig:

So on the left side is the angle bracket the post I'm about to weld is mounted on.  Since the plate is upside down, and my casters are 2 3/4", that means I need 9 1/4" above the table to get a 12" high shelf.  But how to get it to exactly the right height and clamped in with no assistance?  A second rig.

Just to the right of the post is an angle bracket mounted under the table, with a second bracket sticking out to catch the post.  Above the table, I've got a ruler mounted perpendicular to the table as my scale.  The bottom bracket is slightly loose.  When the post is too high, I tap it down, too low, I tap it up, until I'm at exactly 9 1/4".  Then I can tighten the post to its angle bracket on the left, and then move the post into the place it needs to be welded.

Once welding was done, a new problem.  I wanted to roll the cart back home, but the sleeves that the casters mount into are super loose.  To tighten them up, I shoved a paper shim into each tube to hold it in place.

So after another evening of welding, here's the progress.  It's starting to look like a cart.

[steelandflesh]

The Iron Price

As promised, here is the bill so far on the stainless steel cart.

When welding as a non-professional, occasionally a spot in a weld goes poorly.  Maybe you blasted a hole in your metal.  Maybe you let go of your TIG pedal too quickly and a crater formed from the metal rushing away from the center of your weld pool.  Maybe you crashed your tungsten tip into the piece in an epic flash of sparks and metal.

Regardless, your semi competence allowed you to repair your mistake, but the downside is that your weld is now a hulking pile of metal.  "No problem!", you think.  "Grinding time!"... and you'd be right.  The 10,000 rpm angle grinder with a flap wheel or grinding disc can turn amateur hour into professional in no time flat.  Just don't raise your index finger off then handle when holding one without gloves.

Pretty gross.  Probably a centimeter long and just shy of the bone.  Luckily scaping off that much flesh takes with it all the ouchy parts, so it hurt not at all.  A quick cleaning to get all the steel powder out, and a wrap of duct tape, and back to grinding!

Now that your welds all look beautiful, it's time to cart them back down the nicest part of San Francisco come midnight:  6th Street.  Conveniently located along the route from TechShop to my project storage are a needle exchange, 2 missions, an hourly hotel, and a halfway house.

I'm using the Sammie Wagon to move projects around (more on that later) attached to my wife's electric bicycle, so luckily I can do the high speed tour of the local fauna, but on arrival I have to detach the cart.  Here comes the second problem: sticky cart attachment spring.

Pro tip for the day: when yanking violently on two pipes connected end to end to disconnect them, do not allow your body to come between them.  Thinking on this project has caused me to resign myself to a fate of ending life with fewer fingers than I began it.  Either that or I will buy a pair of gloves.

Next time: an old project is destroyed so that a new one can have life.

[steelandflesh]

Stainless Steel Cart Progress

Inaugural post!  I am a game programmer at Supergiant Games, and I often work on metal projects.  Right now, I'm building a stainless steel surgical cart for my wife's hospital.  This first shot is it mid-weld.

"Why are you welding sheet metal for a simple cart, Chris?  Why didn't you just fold the metal into a box like you planned in this shot of the template you cut?"

Nice try, smartass.  Turns out machines like a finger break (for bending sheet metal) rated for 16 guage mild steel won't necessarily handle stainless, so I had to cut my sheet and weld it back together into a box.  This is tricky for stainless because as it heats up, it warps quite a bit.  This shot is me welding with a pool of water and wet rags on top of my sheet to try and keep it cold and flat with partial success.

You might wonder "Man, that looks heavy.  How did you carry all that 16 guage stainless steel around downtown San Francisco's colorful and character filled SOMA 6th street area?"  Why, on my back using duct tape, of course.

Enough questions.  Next time: why you wear gloves when using and angle grinder, and why should be careful with your fingers when mounting a custom bike trailer.