The print starts out fine until it moves over to the second area (where it forgot to build a raft) and starts spewing plastic out into the air since there's nothing in lower layers to print on top of.
> Modeling is kind of a pain.
Unfortunately, it's much more than "kind of". It's one of the biggest barriers into 3d printing. It's all happiness and flowers when your stuff is simple or on Thingiverse/STL/model website. But once you venture out that, modeling is very hard. Not only that, but you have to abide by 3d printing modeling standards, which are learned by trial and error.
> Filament isn't food-safe
Actually, it depends. First, you need a natural filament. ABS and PLA by themselves are safe. The additives and plasticizers aren't always. And the worst is that companies will hide behind "trade secret" when you ask them for the MSDS for all the components in their plastic mixture. Or worse yet, you get Chinese filament and fun things like lead and cadmium. Yum!
Now, say you find natural ABS, great. ABS MSDS is clear and concise. Now, your second issue with food safety now is the voids in a FDM printing technique. You have to heal the holes, somehow. This can be done a few ways. The best way is to use an acetone vapor bath and melt them together. It works, and works well. Another way is to use a food-safe epoxy and paint it to fill the voids.
But no, it is not directly print->drink from.
> It is extremely slow to print anything.
It is slow to print. However what you should be comparing it to is injection mold turn arounds. There, you would be looking at a mold change and 2 weeks absolute minimum turn around, versus a few hours. And the cost per model is really pretty cheap. A mold usually costs between $10,000-$25,000 because of the CAD time and 4/5/6 axis metal milling of said mold. I know a few companies bought a cluster of 3d printers and print out what they need, and have minutes-turn around for changing and fixing parts.
> Model prep (slicing and supports) took a surprisingly large amount of time....Supports are horrible to deal with and remove.
I use slic3r to do most of my GCode generation work, and it's usually considered one of the fastest. On models that have tons of detail, yeah, it's slow. But not that bad for most models. I think this is the case of being stuck with garbage proprietary tools with no easy way to choose the best software for the job.
Also, there's a bit of work with using the water soluble supports with a 2nd hotend. That plastic's expensive, but works well. You have have to tune the 2nd hotend appropriately (and tuning initially can be painful). Aside that, standard supports using ABS or PLA do indeed suck. Try to print without using them if you can.
> but then at the top it would get all stringy.... I had several prints getting stringy
That's a consistent sign that your hotend was getting too hot for the plastic. The correct way to do this is print at a hotter temp at the base and bottom few layers, and cool down to a 'non-stringy temp', usually found by slowly turning down the hotend until you get low/no ooze.
> having to replace the nozzle
MakerBot's Replicator 2 nozzle was known defective. Worse yet was that they sold 3 packs for $500 knowing that about after 20-30 hours of use will destroy it. They're currently defending a class action lawsuit precisely because of this issue.
And also, with my 3d printer, my nozzles are $8 each, and all metal. Mine can print PLA, ABS, HiPS, Nylon, and polycarbonate. If you had a RepRap, you can look on aliexpress and find them, and the rest of the parts to build stuff cheaply.
- I make CAD software for a living, so I've got a leg up when it comes to modeling. And I still absolutely find it a pain.
- Interesting feedback on being food-safe, thanks for the clarification.
- Yes, I realize that it's much faster than other options. I guess I was just a bit naive. Still, I'd love to have a 3d printer that is even faster than, say, a Carbon3d printer. It would really speed up the cycle time.
- When I first got started on the Rep2 makerware was using makerbot's C++ slicer for most models, and a Python-based slicer for high quality. The situation is much better now than it used to be. I've heard good things about slic3r.
I have gotten better at designing so that supports aren't needed. I've also had some good luck with meshmixer's editable supports.
- The rep2 nozzle is okay, it's the 5th gen replicator (after they were bought by stratasys) that has the horrible problems and is super expensive. Based on what I've heard, I wouldn't recommend buying a 5th gen makerbot.
No, I would love a 3d printer that would only take seconds to make a part. However, with ABS, we're effectively limited to about a 200mm/[email protected] is about the top end of printing. Obviously we can add in more hotends and other tricks, but the flow of ABS itself is the limiting factor.
And about the supports, I've usually approached them more from a metal casting way. Making a sprue network isn't difficult; the only weirdness is to design the 'infill' in the STL, rather than let the slicer handle it. Because usually, I want a 5% rectilinear infill for the sprues and something better (say, 30% honeycomb) for the part. Ideally, I'm working on a way to delineate areas that I want different infill on slic3r. That would be pretty spiffy.
Also, have you seen Antimony on the CAD front? http://www.mattkeeter.com/projects/antimony/3/ Pretty neat idea, similar to the system in Blender.
Difficulties I was aware of:
- Modeling is kind of a pain. Thingiverse and other 3d model repositories are helpful, but anything custom modeled can be a pain. TinkerCAD is pretty simple.
- Single color limits the kind of ornamental models you can print
- Filament isn't food-safe
Difficulties I was surprised by:
- It is extremely slow to print anything. The simplest model is maybe 10 minutes. For larger models the time can increase to 8-10 hours easily.
- Model prep (slicing and supports) took a surprisingly large amount of time. This has been improved, but still not great.
- Supports are horrible to deal with and remove. There are other slicing packages that offer better supports, but nothing fantastic.
- The hardware was definitely not consumer grade. I was expecting something like a microwave, where I open the door at the end and a model comes out. I have to know far too much about how the 3d printer works.
So, here is a short history of issues with my replicator 2:
- First few prints worked great.
- The nut and bolt print was misaligned by layers - the nut was skewed into sort of a parallelogram. This was due to one of my belts needing to be tightened, causing slippage each layer.
- After a few prints, I started getting good prints at the bottom of the part, but then at the top it would get all stringy. I tried various things to fix this like decreasing the distance between the build plate and the print nozzle, increasing it, and so forth. Eventually I needed to open up the print assembly and tighten something that controlled how tightly it pulled filament through.
- At one point I had some filament break off inside, blocking the nozzle. I needed to remove the heat sink to get at it, but when I put it back together, I accidentally squished the wire that sent temperature information, so it stopped being able to heat. Took a while to figure out the problem.
- I also had a problem where one of the circuit boards was misaligned, so when it was trying to go to its home position it wasn't touching the sensor and would just keep trying to move in the one direction even though it had reached the end.
- I had a file with the normals flipped, so it wouldn't print supports properly. It was a minor problem, really.
- I had lots of problems where the filament would either be too far away and not adhere properly to the build platform, or too close and jam the nozzle. I also had a problem where some prints were very difficult to remove, I ended up destroying them trying to remove them. Using blue 3M painter's tape solved the issue. If you get the 2" rolls, 3 strips cover the build plate.
- I had several prints getting stringy, I ended up having to replace the nozzle and thermal tape entirely