Friday, June 22, 2012

PENSA D.I. Wire bender

The D.I. Wire Bender by PENSA llc is an arduino-controlled CNC machine that bends metal wire to produce 2D and 3D shapes - an interesting take on a 3D printer. The D.I. Wire Bender can read vector files, OBJ files, text commands, and coordinates.

This is one of very few low-cost machines I've seen that can do rapid prototyping in metal - and it is open source! The Google Code project page is here. You will need tougher motors if you want to use tougher materials than 1/8" aluminum wire/rod.

I find the D.I. Wire Bender exciting for the following reasons:
  • Rapid Prototyping in Metal is typically expensive; this could be a lot cheaper.
  • CNC Rapid Prototyping is even better, because it removes a few chances for human error
  • If we can do this, we can make a CNC pipe bender - Which would open up doors for rapidly prototyping and manufacturing new vehicle designs. For example, a CNC pipe bender would make it easy for the MakerPlane team to print out structural components for future non-composite aircraft designs.

Monday, June 18, 2012

Print Circuit Boards with a Laser Printer

According to, if you have a Laser Printer, some Printed Circuit Board (PCB) blanks, and a few other easy-to-find items, you can Print your own circuit boards in about 30 minutes.

You can get a detailed step-by-step instructions with pictures and a great bill of materials (and where to buy them if needed) here, via, but here's the gist of it:

  • Design your circuit (here's a directory of free software for that)
  • Print it on photo paper using a laser printer
  • Tape the paper onto a PCB blank
  • Heat-transfer the ink from the paper to the PCB blank
  • Soak the board + paper in cold water
  • Peel off the paper, scrub off remaining paper leaving nothing but toner in the shape of your circuit on the copper surface of the PCB blank
  • Etch off the copper that is not covered by the toner (lots of methods available; soak it in acid, wipe it off with ferric chloride solution)
  • Wipe off the toner with a solvent like acetone, leaving only the copper circuit

The speed and relative simplicity make this a great option for rapid prototyping, and testing new circuit/Open Source Hardware designs prior to production runs.
Lots of variations have been tested on the process above. Tom Gootee talks about his process for it and experiences with various tweaks, like using a clothes iron on the linen setting instead of a laminating tool.

Friday, June 15, 2012

Project Falcon: Free Interactive Wind Tunnel Simulation Software

Project Falcon is wind tunnel simulation software brought to us as a free technology preview by Autodesk Labs. The goal of Project Falcon is to help designers intelligently consider the aerodynamic properties of their designs without first having to learn computational fluid dynamics (CFD). This could be a great thing for vehicle designers.

Windows is the only supported operating system, but the minimum hardware requirements look pretty achievable.

To me, the exciting features of Project Falcon are:
  • Free to download; no Autodesk license required
  • Broadly accessible user-interface (Probably less intimidating than Elmer for your first foray into Aerodynamic design)
  • Easy installation (register with Autodesk, download the project falcon installer, run the installer, double click the icon, then open the STL file you want to check the aerodynamics for)
  • Blog/forum support via Autodesk Labs
  • Users develop an intuitive grasp of aerodynamic principals using Project Falcon's rapid visual feedback in response to design and parameter changes
  • No specialized knowledge required to study the aerodynamic properties of your 3D designs
If anything could motivate me to mess around with Microsoft Windows, it's probably Project Falcon, owing to the 2nd and last bullets on the list.

Here is a brief video introduction to Project Falcon via Autodesk Labs.

One thing the Project Falcon overview page emphasizes is the high speed with which this software calculates and displays results. A prominent complaint in the wind power optimization research papers I've run across is that CFD is computationally expensive and time consuming. So is Project Falcon sacrificing quality for speed? Probably. But this Project Falcon Validation paper shows that Project Falcon at least calculated the the correct coefficients of drag for a sphere, a cube, a cone, and an odd polyhedron that looks like a brick with 3 corners cut off and something small sticking out the bottom. I can't help wondering if Project Falcon has potential as a tool for optimizing the shapes of home-use wind turbine airfoils.

In the Aero Challenge, Local Motors has invited members of its open source vehicle design community to to use Project Falcon to help create a more aerodynamic design for the next Peterbuilt big rig.

Wednesday, June 13, 2012

Interactive Intro to Aerodynamics/CFD by Local Motors

If you are interested in automotive design and looking to develop a better intuitive grasp of aerodynamics, today is your lucky day.

PeterBuilt is sponsoring a design competition for the next Big Rig via the Local Motors Forge, a place where interested members of the online open source community can go to work and play at vehicle design.

The Local Motors website is encouraging visitors to download Project Falcon, and use it to play around with edgy new Big Rig designs for the Peterbuilt design competition.  Project Falcon is a new and at least temporarily free piece of wind tunnel simulation software that allows for interactive investigation of the aerodynamic performance of designs, and is intended to be used early in the conceptual design phase. Project Falcon reads .STL files, so you have a lot of options for CAD software to model your ideas. The goal is to find a design that will look good and increase fuel economy by achieving greater aerodynamic optimization of the vehicle. The motivation?  10 Awards, and $15,000 in prizes.

The design submission period is June 5th - June 26th.

Here's the video via Local Motors:

Project Falcon may also be a great tool for those of you looking to get into Open Source Aircraft design.

Look out for a future post with more details on Project Falcon software.

Monday, June 11, 2012

Let's Build an Optimization Tool for DIY Wind Power Airfoils

While I absolutely love the DIY accessibility of home wind power generation projects like the Chispito Wind Power Generator, The DIY aviation nut in me is screaming that we could all get significantly more power out of rigs like this if we had an optimization tool that would ask us our motor specs and what the wind is like where we're mounting our generators, then spit out .stl files of the right shape airfoils to get the most power out of the wind. I don't know what percent difference the average builder could expect to see from optimized blades...but based on the research paper linked at the very bottom of this post, I think it would have to be huge. The difference between an aircraft-optimized airfoil and a wind turbine optimized airfoil can be as much as 50% in normal wind conditions, and neither of those airfoils seem to have much in common with the simple   blades we DIY types make out of cut up PVC pipe.

CNC hot wire foam cutting technology is a good start for rapid prototyping custom airfoils based on .stl files. With this technology in play, I could see the production of custom wind generator blades becomming a great little microfactory business.

The past couple times I started to post something along the lines of "Dear Santa or Jesus or open source community members, next I would pretty please like the following open source tool to exist" I found what I wanted in a Google search. The most recent two examples were free, open source computational fluid dynamics code and an inexpensive, open source stereolithography machine.

This time, the closest I have found to an airfoil optimization tool for DIY wind power generation are these research papers:

  • Aerodynamic Shape Optimization of Vertical Axis Wind Turbine Using Differential Evolution: Summarizes the preliminary results of a UT Arlington Aerospace Engineering group's efforts to create an automated airfoil optimization code. Bonus: if you want to learn the basics of wind power theory, read the introduction to this paper. It'll be a great vocab lesson even if math isn't your thing. The group published this paper under the creative commons attribution license...cross your fingers that they will be just as generous with the source code they're working so hard to create.
  • Study of the Performance and Robustness of NREL and NACA Blade for Wind Turbine Applications: This study predicts that major power gains (~10-50% over the wind speed range of 3-9mph) would result from building small home-use wind turbines using the airfoils designed for horizontal axis wind turbines by the National Renewable Energy Lab (NREL) as opposed to the currently common practice of using airfoils NASA designed for aircraft back when the agency was still called NACA. As you can see in table 1 and in figure 3 (click here, scroll down), the NREL and NACA airfoils look almost identical. I suspect that using either type would yield a vast improvement over the current DIY standard of cut up PVC pipe.
The top one looks like a great start...but I'd like to see the open source community run with it and start making better wind turbines.

Friday, June 8, 2012

B9Creator: An Open Source Stereolithography Solution for < 3% of the Price

The B9Creator is an open hardware project brought to us by Michael Joyce.

This Wikipedia article claims that stereolithography machines typically cost in the range of $100,000 to $500,000, and use resin that costs between $80 to $210 per liter. The B9Creator delivers this functionality (rapid prototyping using light to solidify resin) for <3% of the price, conservatively, using resin that costs about ten cents a gram. For $2,375, backers on Kickstarter can get a complete kit that can theoretically be assembled in an afternoon. For $3,375, backers get a fully assembled and calibrated machine.

As you can see in the video below, Michael Joyce, the B9Creator's inventor, is committed to the development of open source software and hardware, and is looking forward to the innovations that will be inspired by his creation.

The B9Creator is offers unusually high resolution for a low cost 3D printer (0.05 - 0.1 mm for the B9 vs. 0.2-0.3 for the Makerbot Replicator). The B9Creator starts by slicing a 3D object data file into a stack of 2D images, and projecting the first 2D image onto a thin layer of photo-initiated polymer resin long enough to cure a .05 - 0.1mm layer, which attaches to the build platform behind it. The B9Creator then moves the build platform to break the bond between the cured resin and the projector window, re-positions the build platform above the projector, and projects the next 2D image. The B9Creator repeats this process until the 2D images have been stacked up to produce the finished 3D object.

The B9Creator can build 3D objects at 12-20 mm/hr independent of the object's density. RepRap project and Makerbot 3D printers use fused deposition methods, wherein plastic is melted, extruded through a small nozzle, and 3D objects are built by fusing melted plastic from the nozzle onto the layer below. Because this method (called Fused Deposition Modeling, FDM) relies on the relatively fixed rate at which plastic is melted and extruded through the nozzle, denser objects take significantly longer to build using FDM than more fluffy ones. The build speed of the B-9 creator is dependent on the layer thickness set by the user, but does not depend on the density of each layer.

This video is from the B9Creator's kickstarter pitch, which as of this writing has more than quadrupled its funding goal and still has over a week to go:

Also via the kickstarter pitch, here is a video showing the B9Creator prototype in action, printing the Metatron:

Have you seen the B9Creator in action?

Tell me about it in the comments! I am especially curious how sturdy the resin objects produced by the B9Creator are, and what, if any, surface prep is required to clean the models of any un-cured resin film.

Wednesday, June 6, 2012

Plans for a Simple DIY Wind Generator

The Chispito Wind Generator is a small, DIY wind power generator capable of generating 100 Watts in a 30 mph wind. It starts charging a 12-Volt battery in a 7-10 mph wind, and you can build one with the relatively short list of inexpensive parts and tools found here. The Chispito Wind Generator uses an old treadmill motor for the generator.

If your electricity use is average (around 11,500 kWh/year for US households) and you lived somewhere with 30 mph winds 24-7, this thing could provide about 7.5% of your power needs.  Nobody I know lives anywhere that is consistently that windy, but the Chispito Wind Generator could be a fun and educational DIY project in a lot of places. Here is an 80-meter wind resource map of the US that may give you some idea of the wind speeds in your area, although 80 meters is a bit high for a back yard wind generator.

For pictures and build instructions, check out "How to Build a Chispito Wind Generator" page.

You make the blades on this baby using some cut up PVC pipe and sand paper. I love that the tools and materials needed to build these blades are so close to universally accessible. If you follow this blog, bets are probably good that you could finish a significant percentage of this wind power generator project before making a trip to the hardware store.

All that said, the  DIY aviation nut in me is screaming that we need an inexpensive, open-source way to optimize airfoils for domestic-use, DIY wind power generators. Look out for a future post on the huge increases in power generation that can be obtained by using well-optimized airfoils, and my ideas about how we can make such airfoils inexpensive and widely available.

    Monday, June 4, 2012

    Kit Built CNC Mill/3D Printer

    What do you get when you cross a MakerBot, a dremel, and a kit built CNC router from

    A WhiteAnt CNC Mill/3D Printer

    Video stolen from the white ant product page, via

    Although the cost of the WhiteAnt kit strikes me as similar to other open source 3D printer kits, and I already have a Makerbot thing-o-matic in the house, there are a few things that catch my  interest about the idea of building a WhiteAnt:
    • I am a proponent of versatile, low cost manufacturing equipment, and the WhiteAnt looks like a 2 for 1 deal since the user can quickly swap the extruder for a Dremel and have a CNC mill without taking up extra space, or investing the time to build another frame and set up a second set of electronics and software
    • The WhiteAnt frame looks a lot sturdier than the Makerbot
    • Building a WhiteAnt is essentially a practical, guided lab exercise for this textbook on 3D printing in plastic, and I am old school enough to like textbooks and formal labs.
    • produces good videos about how to assemble their various kits. (To see what I mean, you can watch an instructional video on the WhiteAnt Dremel mount assembly here, or the video instructions for connecting the WhiteAnt electronics.)
    The WhiteAnt is built using the arduino, a single-board open source microcontroller, replicatorG, an open source 3D printing program, and the generation4 electronics and tool-head available from

    If the ability to do 3D printing via fused deposition modeling in extruded plastic is unimportant to you, and you need to use a mill more than you want the experience of building your own, you may be better off to sacrifice the cool factor and buy a low-cost mini-mill like this one from It comes fully assembled, has enough torque to mill steel, and has a similar price and x/y/z travel to the White Ant kit.

    Friday, June 1, 2012

    Eureka CNC: A Microfactory for Airplane Parts, Among Other Things

    Eureka CNC is a microfactory that uses a CNC hot wire foam cutter to produce specialty aircraft parts, among other things. According to the Eureka CNC website, the owner, Stephen James, has a (very impressive) day job in the USAF, a family to provide for, and an awesome mental problem called project ADD...and he has still managed to single-handedly produce a wide variety of useful and cost-effective products and build a few airplanes of his own.

    Exciting features of Eureka CNC:
    • The ability to rapidly and precisely turn a 3-D CAD file into a foam airfoil core ready for the next step in the airplane build project (covering it in fiberglass)
    • Extreme versatility and efficiency: products include custom crown molding, race car fairings optimized for structure and Reynolds number, and (most exciting of all) wing cores for a wide variety of home-built composite aircraft including the Long-EZ, Cozy MK III, Cozy MK IV, Berkut, E-Racer, Quickie Q2/Q200 with LS1,
    • Although building it did not sound easy, the Eureka CNC hot wire foam cutter does sound like it's based on technology that is well within the reach of the open source community
    •  Now that there are open source aircraft design projects in the works (click here and scroll down for a list), we will probably soon see rapid prototyping processes like Eureka CNC's make new aircraft design ideas a reality in record time.
    • This technology could be applied to designing, creating, and selling some awesome fiberglass kit car bodies
    I would love to see a higher level of integration between the outputs from conceptual design and mesh creation software like this, computational fluid dynamics optimization codes like this, 3D geometry output files, and affordable CNC rapid prototyping technology like the Eureka CNC hot wire cutter. Anything to shrink the currently huge amount of time between having an aircraft design idea and seeing it in prototype...

    On a side note, I am a happy customer of Eureka CNC. My husband and I bought wing cores from Eureka CNC for our airplane build project, the Cozy MK IV. The average build time for Cozy MK IV projects is around 3000 hours, which amounts to a year and a half of 40-hour work weeks. Today, we are in the neighborhood of 10% done. Stephen James' microfactory-built CNC wing cores saved us a big chunk of time by completing several steps of the build project for us, so maybe that figure is more like 12-15%.