Archives for category: 3d printing

image from the film The Princess Bride

closeup from the movie

This is a prop I’ve wanted for a while. There’s been so-so reference material for a long time pieced together by other hopeful collectors. We’ve got some great close-ups and screen captures from the film itself. The goblet wasn’t created for the film, so there are a bunch of them out there in the world. They’re rare, but they exist. A member of posted some pictures recently of the one they happened to have already owned by chance. Another member posted pics they found from an eBay and Etsy auction they found while browsing Pintrest. Judging by their descriptions of the items, and a private message to one of them, none of the sellers knew they were used in the film.

Using these pictures as reference material, I’ve been able to sculpt a CAD model using AutoCAD and Meshmixer, and have printed out a goblet in plastic using the DittoPro at the Baltimore Node Makerspace. The final plan is to cast these goblets in lead-free pewter.

I first created a sketch of the model to get a good feel for it and to see if I ran into any trouble making the model. After creating a rough sketch, I got some feedback. We know the base is a bit bigger, 3.5″ instead of 3″ like I originally drew. The circles around the top are different sizes, 3 big and 3 small. The lower three ovals are way more squished than I had originally draw. I went back and recreated the goblet, making these changes.

Some times I run into a feature that’s too difficult for me to model in AutoCAD and I have to use a mesh modeling tool, like Meshmixer. This usually happens if it’s a very complex curve, or a variable radius fillet. Sometimes there’s also a feature that ends up being way easier to model in Meshmixer anyway, or something that is so simple to do in Meshmixer it’s not worth fiddling with it in AutoCAD. Since it’s very difficult to go backwards from Meshmixer back into AutoCAD, I want to be sure to save all of these Meshmixer tweaks for the end, and do as much as I can directly in AutoCAD first.

I was able to create the entire goblet in AutoCAD except for smoothing out a seam I created on the petals, and the complex curve where the fins blend into the neck of the goblet. The part where the fins curve into the goblet cup is easy, but where they start to smooth out near the bottom is hard so I just go over that area with a smoothing brush in Meshmixer.

Based on the seams seen on the screen-used prop, I can tell that it was originally cast from a rigid multi-piece plaster or rammed sand mold. We can see that there are seams down the sides of the goblet in three places. I personally don’t want to use plaster to cast the mold for this, and would prefer to create it from a split rubber mold for simplicity. It’s much easier for me to make the rubber mold than the multi-piece plaster mold, and much safer. I’ve read for plaster molds you have to make sure they’re completely dried out or else they’ll explode when you pour molten metal into them because the water expands so fast. I have experience with silicone rubber mold making but not plaster, and for this project I want to use the silicone. I do still want the seams to be in the part as if I cast it in a multi-piece mold, so I’m adding them to the CAD model. My hope is that the final cast piece will appear to still have come from a multi-piece mold instead of the split rubber one.

closeup shot showing top circles are different sizes

reference image from an etsy seller

reference image from an etsy seller

excellent reference shot
CAD model after making several adjustments
wireframe view of initial sketch

initial sketch model

I’ll be updating this blog post as I work on the project. Please check back soon to see my progress.

A neat thing happened today. Valve, a popular videogame company(makers of Half-Life, Portal, DOTA, Counter Strike, and more) and Shapeways, a 3D printing service bureau that also lets you have a 3D-print-on-demand online storefront, reached and agreement where we can upload any design to Shapeways and share 10% royalties with Valve when sales are made. It’s very exciting to me, and certainly the way licenses should go, in my opinion. It’ll be neat to see if other storefront sites like Etsy take on this approach.
So, here you are, an officially licensed pair of Portal earrings, designed by me, up on Shapeways:
Here’s a few parts I made for Oriole Park at Camden Yards. I headed down to the stadium today to test the fit, and it went perfectly. All 9000 padded seats between the dugouts behind home plate use two of these. It’s a self adjusting bracket to hold the seats to the rest of the chair. Since each seat is angled differently to face the field, a simple bracket doesn’t do the job. This bracket has a rotating ball in a socket to accommodate for any angle. I unfortunately don’t get to sell them the entire 18,000 sets of parts that they use for the whole lot because these are just replacements for when the originals break, but it’s still a nice order and pretty need to add them to my client list! The black one is the original, mine is the all white set. The company who made these originally got bought by a different stadium seating company, and they no longer offer these parts. They don’t break too often, but when they do, the seat is unusable, so it’s always good to have a few on hand.
 My white 3d printed ones next to the black originals

These three parts are the original pieces.

This is the 3d printed nylon part. The central ball rotates freely to any angle.

You can see the metal bars that stick out from the arm rest.
The ball part fits right on the metal bar, and adjusts itself to the angle as needed.
I tested the screw hole placement in their repair shop on a newly refinished padded seat. They replace the vinyl on these all the time.

It works! Installed in section 30 row 29 seat 11

Here’s a great article shapeways put up about it:

I didn’t blog about this when I created it but it looks cool so I figured I’d share the pictures. I gifted this to my friend Jenny. It’s a pretty neat competitive puzzle game. I just used birch plywood, cut it at the node, stained it, and hand painted the symbols. The raised walls are important to the gameplay, and the off-the-shelf version of the game just has printed graphics instead of a physical wall. The pieces are 3D prined and are the Make: robot. I downloaded their .stl file here and you can download my mishmash of source files I created for the game here. It’s poorly documented but should provide a great starting point for anyone interested in making their own.

Han Solo’s going away party! Tuesday April 22nd, 7:30pm.

Unfortunately, the 1:1 replica of Han Solo in Carbonite isn’t going to be coming along with us to the new Baltimore Node location. It was a commissioned project anyway and I didn’t want to move it and then move it again, so I decided to finish it up and get it read to ship it off to my customer in Alabama later this week. Don’t worry though; I’ll be building another one once we get into the new space.

I decided to run an event where people can come and learn what I learned fromthis project. I’ll talk about everything I can remember and show pictures and let you ask questions. I’ll discuss laser cutting, CNC routing, silicon molding, resin casting, 3d printing, 3d model packing optimization, microcontroller programming, sculpting with bondo, and much more. I’ll be discussing the project on April 21st, and you can see the completed sculpture then.

Please bring some food to share if you want, or $5 for pizza. RSVP on meetup

Shapeways is changing their pricing for SLS nylon parts and the metal parts too.  A big change is that each part inside a model now has its own $1.50 startup fee.  If you group them all together and put ’em in a box, then they count as one part, saving you that hefty cost.  The example I’m using is the 30-pack of tiny nipples used as detail pieces for Han Solo in Carbonite side panels.

Before the price change, these cost me $7.18 and I was selling them for $13.  That same file will now cost $62.37 after October 7th, and even more of a gap for polished parts. Something needed to be done if I want to continue offering these parts at a reasonable price.

After I put everything inside a cage, I was able to bring the cost back down to a fair number. It was $7 before and now it’s about $11.  The more time you spend optimizing the layout of your parts, and the tighter together you can get them, the more money you’ll save.  Making them hollow isn’t the #1 most important thing anymore.  A lot of factors contribute to the cost, so optimizing all of them will give you the best result.

Not all parts go up in price; some actually go down. Realize that shapeways used to charge $1.40 per cc of material you used, and now they’re only charging $0.28.  That’s a big drop.  They’re also charging for ‘air’ at 21 cents per cc, but that’s just to reflect the work they need to do for that part. This means that many objects which are not so wispy and wiry, and instead are tighter and denser actually go down in price.  Check out a simple solid sphere for example.  Hollowing it out would only save you $0.07 per cc, whereas before it’d save you the full $1.40. This means that under the new pricing structure, this particular shape would go down in price from $12.95 to $7.07.  Not bad.  I wonder now though how well shapeways will compare to other 3d printing service bureaus.

In the video I showed example of lowering prices by putting things in cages.  You can also lower the price by linkin things together.  The above earrings show that the system only charges you for objects which are actually separated.  The linked rhombic dodecahedron earrings (available in my store here) are successfully detected as two parts, not 6 like you might expect.  Same goes with the Claw earrings, which aren’t even solidly connected by a closed loop.

You can also link objects together with a 3d printed string so that they’ll be detected as one part.  It does a very good job of this on my tests so far. This is like putting beads on a necklace. 
If your parts are spaced out far away from each other, it’ll cost a lot since you’re using a lot of space inside the machine. The photo to the left shows the bounding box of the parts. It’s the overall rectangular box that fits around all parts in a model.

The overall bounding box isn’t exactly the same as the machine space.  If you have a big enough hole inside your part, (40x40x40mm) then shapeways won’t charge you for that as machine space since they’ll fit someone else’s part inside of yours.

By the way, have you seen my kickstarter campaign? It’s currently live at

Check out the campaign!

Share it and back it if you can.  Thanks!

If you helped me with this project, give yourselves a hand! In fact, why not head over to e-NABLE at and print out a hand for someone who needs one! They’re hosting an event here in Baltimore on Sunday the 28th, and are looking for about 150 more 3D printed hands for the people who need them.  
It took about a month, including the scanning, slicing, uploading, distributing, checking in pieces, tracking down missing ones for reprints, and the final assembly.  I’ll be giving a talk about the process and the platform on Thursday this week at Ignite Baltimore.  Thanks to all of you who helped by printing pieces, to Marty McGuire who helped glue and did website admin, Direct Dimensions for the scan, and Tinkerine Studios who sponsored some of the shipping costs.  If you’ve got a printer or know someone who does, head over to WeTheBuilders and sign up for alerts for our next project!
Here’s the build video:

More photos:

Today we launched our second crowdsourced 3d printed sculpture at

The way it works is that people who own a 3d printer can log in and download one 4″ chunk of a sculpture, 3d print it, and mail it back to Baltimore where they’re all getting glued together.  George was made earleir this year and has travelled allover the US to different shows and landmarks. Our next project which just launched today is Distributed Benjamin Franklin is our second supersized crowdsourced 3D print. Earlier this year, a new site, was launched by me and a team of 5 other makers at the ArtBytes Hackathon in Baltimore. With the help of 70 makers and their personal 3d printers worldwide, all 110 pieces were 3D printed and mailed to Maryland where they were all glued together to create the one-meter-tall sculpture called George Crowdsourcington.  

George has been traveling all over the country to over a dozen events. He debuted at the NY 3D Print Show, went to Inside 3D Printing Conference in New Yrok, did a residency at the Adafruit office in Manhattan, traveled to South-by-Southwest, Direct Dimensions, The Baltimore Node Hackerspace, several mini-Maker Faires, 3D/DC, Public Knowledge’s HQ, ConFabulation Symposium and 3D Printed Art Gallery at Prince George’s Community College, the US Science and Engineering Festival in DC, and even visited the Capitol Building and the National Mall.  Next month he’ll be at the Silver Spring Maker Faire and World Maker Faire in NY.The hope is that with your help we’ll be able to print all 198 pieces needed to bring Distributed Ben Franklin together in time for the Silver Spring and World Maker Faires.  The parts are available at and anyone with a 3d printer is free to get started! The project just launched on Monday, August 4th and we’re hoping to get all the parts shipped to Baltimore by September 1st, in time for us glue. The model of Ben was sliced up into pieces using netfabb. Any piece which was totally inside the sculpture and a perfect cube was removed, since there’s no need for us to 3D print that.

Tinkerine Studios (the makers of the DittoPro 3D printer) sponsored both the George Crowdsourcington and Distributed Ben Franklin projects by reimbursing shipping costs for makers who participate by 3D printing and mailing in their parts.

Direct Dimensions of Owings Mills, MD did the 3d scanning for us.

We The Builders is a fun free project to be a part of.  It’s great to use your 3D printed to create something awesome with other makers from all over the world. You can find all the details and get started over at

My friend Ryan Kittleson likes to post videos of him doing modeling in Sculptris and they’re always fun to watch.  I decided to try it on a project I did yesterday.  It’s about 3-4 hours of work in AutoCAD, playing back at 7x speed.  Let me know what you think!