Sep 12, 2015 | By Don Foley
Built in Savannah, Georgia in 1862, the CSS Georgia was designed to be an ironclad warship. However the Confederate ironclad didn’t quite work out to the specifications the designers had hoped. Instead of being a mighty ironclad, it was relegated to being a gunship at anchor to protect Savannah.
The steam engines couldn’t push the CSS Georgia up the Savannah River, so they decided to drop anchor and make it a marine battery. Near the same time, General Sherman was working his way to Savannah and the Confederate army decided to scuttle the ship so it wouldn’t fall into the hands of the enemy. The CSS Georgia never fired a shot in battle.
Fast-forward to the Summer of 2015. The U.S. Army Corps of Engineers, while widening the Savannah River to make room for the new generation of Panamax ships, had to fit into their $700 million budget a $14 million line item to salvage and restore the wreck of the CSS Georgia (the resting spot of the CSS Georgia has been known since 1968.) The Army Corps of Engineers is slowly bringing up the artifacts of the CSS Georgia. This project caught the attention of Don Foley, both a history buff and 3D print designer. He decided to re-create what was known about this ironclad ship.
Many of Don’s builds are based off of illustrations that he does for his clients (National Geographic, Popular Science, Discover, Scientific America, etc…) but this one he did for himself. His builds tend to be space related (he’s covered the space program since 1981, having grown up in Cocoa Beach) or sea related (he’s a sailor who has sailed the east coast several times). Toss in a pinch of history, and this new build is right up his alley.
Section by section, the craft is being raised. (photo by US Army Corps of Engineers)
Back in the day, ironclad craft were high-tech. Upon research, Don found that the sides of the CSS Georgia were interlaced rail sections. This was a ‘make-do’ type of ironclad. The Army Corps of Engineers are working around the craft to raise portions of the craft and with each dive, more is learned about the ship.
CSS Georia Cannon being raised this Summer (photo by CNN)
A cross-section of the CSS Georgia bulkhead. You can see the overlapping rail sections. (photo by US Army Corps of Engineers)
When Savannah was under threat, the ladies of Savannah rallied and came up with $115,000 dollars (Confederate dollars) to help build the CSS Georgia to protect their city. The ironclad floating battery never fired a shot in battle. When Sherman came closer to the city, the CSS Georgia was scuttled. The City of Savannah didn’t want to be destroyed, like many other southern cities, and came up with a financial ‘arrangement’ with Sherman so the city wouldn’t be sacked. (We’re glad they did, because Savannah remains one of our most beautiful southern cities.)
Given the current events of the salvage of the CSS Georgia, Don wanted to do something for this small slice of history so he built his version of the CSS Georgia. Only a few drawings of the craft exist. No plans survive. ‘Almost’ one photograph of the confederate gun boat can be found. Almost one photo? In the 1980s, a spirited youth and his brother took a photo of a model they had built and attempted to pass it off as a ‘photo of a photo’ of the ship. Historians are calling this a forged photo, but it is not a copy, so Don calls it a ‘faked’ photo.
That brings us to Don’s 3D print of the CSS Georgia. Much like the CSS Hunley that he covered for National Geographic, this story had several things to pique his interest. It was in the U.S. South, historic as well as maritime based. And it was new technology for its time. Given these factors, Don built a 3D printable version of the CSS Georgia. It was a bit challenging due to the current lack of information. Based on the information available (Sept. 2015), Don decided that the railroad ties were used vertically, given that the current findings were in 24′ sections. Several, but not all, off the illustrations of the time supported this.
Don built his CSS Georgia in Lightwave 3D. Many historic notes had four cannon ports on each side with additional ports forward and aft. They also note that there were most likely only 4 cannon on the ship. Don put one cannon in each available port. If you build this and want to be historically correct (or as close as possible) you can delete some of the cannons.
Screen grab of the main deck in Lightwave 3D
Lightwave 3D modeling supports 3D printing as well as Don’s illustration and animation demands. Lightwave 3D might be overkill for most people interested only in modeling for 3D printing, but it offers a solid suite of tools for complex modeling and is Don’s tool of choice.
Section 2 of the 3D print of the CSS Georgia in Simplify 3D
The print is in 4 STL parts: the first is the main gun deck; the second is the stand and all the varied parts; the third is the stern hull; and the fourth is the bow hull. All the STL files were imported into Simplify 3D, Don’s favorite print processing software.
Don’s studio has seven 3D printers. His hand’s down favorite producer is the Wanhao 4S. The parts and pieces (the second print) was the most complex print he’s done to date, with 55 separate pieces. To ensure the print was viable, Don printed it two times and both worked perfectly.
Final print of the second print, the parts and pieces, over 50 objects
Don’s notes on 3D printing
Hundreds of printers are on the market now with several design programs and a host of slicer options. Many of these can produce great 3D prints. Below is what Don Foley uses to produce his prints. He is not saying this is the ‘only way’ or this is the ‘best way’ but simply these are the tools he uses to create the prints that he produces.
1) The 3D printer
This print was done on Don’s Wanhao D4 and D4S 3D printers. The bed is similar to the Makerbot printers and many of its clones. So Don builds most of his prints to fit in this popular form/size. Most printers have several thousands of hours on them and just keep churning along.
2) The design
Initially concerned that the design on this project would work, Don didn’t hold out a lot of hope for it. Hundreds of railroad ties would need to align perfectly. Lightwave 3D gave him the tools to pull it off. The model was fairly forgiving, thankfully.
3) The software
This is the step where many good ideas fail. Until Don started controlling his printer with Simplify3D, he didn’t get the results that he demanded. Today, this is the ONLY software he uses to create his prints. Why buy an expensive 3D printer and not spend a little more to get good results?
4) The material
Since Day One, Don has been buying his material mostly from MakerGeeks (although he tried a dozen other sources.) Earlier this year MakerGeeks bought their own extrusion machine and started making their own home-brew material. After working with MakerGeeks material for over a year and having great success with the material they supplied, Don crossed his fingers and figured they would not supply a filament that was inferior to what they were already providing. They didn’t disappoint. The entire build is done from MakerGeeks new home-brewed material. While Don has tested MakerGeeks home-brew material, this is his first full print using their material on one of his builds. Curling was a big concern, considering the large flat surfaces that needed to align. As you can see from the build photos, curling wasn’t a problem at all. The PLA material was printed at 205° at 2500 mm/m with no heat on the bed. There was no curling from the bed.
5) Bed prep
If your bed isn’t ready, your print will fail. The first thing, of course, is a perfectly level bed. The next is the bed surface. Don has found that the ‘paper’ method works fine. Once you level your bed using a sheet of paper, you’re good to go. Next. just tweak adjustments as you print. Don puts down Scotch Brand Pro Painter’s tape and wipes it down with rubbing alcohol. If the head digs into the tape, the bed is to high. If the material prints in the air and not on the tape, the bed is too low. The key is to find that perfect spot. To make this determination, he prints a brim for all of his prints. If the brim prints fine, the print will be fine.
6) Zero your Z
If your model’s base points aren’t zeroed out on the base plate, nothing above matters. You can do everything above correctly, and still have a failed print if the base points of your model are not flush with the print bed. This may seem obvious, but even Don has this happen (and it happened on this print’s first test). In a complex print, parts may be floating above the print platform and you won’t see it until you’re printing spaghetti.
Close-up of the print of the bow section. No curling.
This photo is a good example of how the print did not curl. This print would not work if anything curled. The bow and stern need to fit together (they are printed upside-down.) The key gun deck is printed flat. The only way this print works is if there is 100% NO CURL so all these parts fit together. And they do.
3D Printer specs
Every printer and every material need different specifications. For Don’s purposes, he clings onto the following specs for one single reason…they work for him EVERY time. These are the key specs he puts into Simplify3D. Bottom line, it works for him.
- PLA (from MakerGeeks)
- Bed temp: None
- Head temp: 205°F
- Print Speed: 2500 mm/m
- Extrusion multiplier: 0.90 mm
- Extrusion Width: 0.45 mm
- Retraction Distance: 1.00 mm
- Retraction Speed: 1200 mm/min
- Layer Height: 0.1500 mm
- Top Solid Layers: 6
- Bottom Solid Layers: 6
- Outside/Perimeter Shells: 3
- Interior Fill: 15%
- No Supports
Successful Printing (Don’s take)
Don completely understands that 3D printing is still in its early days. He’s not pretending to have it all figured out. With seven 3D printers in his studio and the numbered steps above, this is how he cranks out successful prints on a daily basis. Your 3D printer and your formulas may be different. The message here is to keep track of all the variables, change one at a time, and keep experimenting until you find out what works for you. When Don hit the ‘magic’ variables for him, he kept printing. Over one year later his ‘numbers’ allow him to print EVERY single time, even with complex prints.
Don prints slow. Painfully slow. And this is possibly why his prints work. Almost everything he prints is around 2500 mm/min. This print is no exception. That means that the bow hull print took about 10 hours, the stern hull took about 10 hours, and the parts print took about 11 hours. The main gun deck took 20 hours to print. Yup, that’s 50 hours of print time or just over two days of printing. The print may be fine if set at a faster time, but Don has learned that a slow successful print is better than a fast failed print. So he ALWAYS prints slow. It certainly could be up for debate, but Don has 50 small parts printing perfectly on this print multiple times (and thus recorded) with no problem.
A few things may be different in this image from the earlier images. Don had a hard drive crash in-between back-ups, and rebuilt the model because he liked the subject so much. This was the first time he built a model looking at the model he had just built and printed sitting in front of him. His vow was to make the second version better and it is. (Don also has a better back-up scheme now too.)
After printing, the first thing to do is to glue the oval sleeves and hull together. Apply Control Gel super glue to facing parts (not too much so it spills over), then turn it over so the deck is flush down on your table. This will ensure a flat surface for your gun deck to be placed on.
Above: full build after editing
Above: the parts after printing
Don applies a primer coat for most of his models. Usually this is for his Metal Effects paints, but he finds the primer coat helps ‘forgive’ some of the ridges created by 3D printing. He buys bamboo skewers from the grocery store to elevate the model from the newspaper ‘drop cloth’.
Final painted CSS Georgia
The long boats are 3D printed as part of the ‘parts print’ and can be attached to the davits using thread. Don uses a beading needle to pull off this trick. The needle is like a think wire that is divided in two. Separate the wire, throw your thread into the gap and then pull the beading needle through the small oar holes on the long boats. Even if the holes on the boats fill in with paint, the metal needle should poke through with no problem.
This 3D model will be free to download until Sept. 14. You can purchase/download this model here.
Posted in 3D Printing Applications
Maybe you also like:
- Printhese, Roel Deden's reimagined 3D printed arm prosthesis wins Dutch James Dyson Award
- Fake 3D printed security seals used to mask pharmaceutical heist
- Dutch company Scanologics releases new mobile 3D scanner: the ScanLounge v2.1
- Make a 3D printed camera slider for just $20
- Maison 203 presents a nature-inspired clutch made entirely by 3D printing
- Ballantine's designs 3D printed 'space glass' for drinking Whiskey in zero gravity
- Finnish friends use miniFactory 3D printer to complete cool Ghostbusters outfit for summer festival
- More revealed about 3D printed Acetabular Cup formally approved for hip surgery in China
- Haier unveils new standing 3D printed air conditioner with gorgeous pattern inspired by fish scales
Jason wrote at 9/15/2015 2:08:31 AM:
People don't realize that they can make just about anything with a 3D printer these days. With the one I have, I've already made lots of models and have xmas taken care of. Blueprints are easily found online.
Don Foley wrote at 9/14/2015 12:32:06 AM:
Note: Model is now free till Sept. 20th