Newbie questions that Googling has not sufficiently answered

So, I've wanted to get into 3D printing for a long time now, but haven't had the spare funds until lately to do it. I've been looking at various objects and printers, and have a few questions I'm hoping people can answer.

1) What ensures the best print quality? From what I can tell, the primary influences are extruder nozzle width, size of each layer, and the g-code generator. (Ideally, I want prints like this - [www.youtube.com] )

2) How difficult is building a printer vs. ordering a pre-constructed model, especially for those who are less mechanically inclined? And is there a sufficient price differential to make building it worth attempting?

3) A follow-up to 2, how difficult is owning/maintaining a 3D printer going to be without technical expertise? I have an engineer friend I can bug, but I'd prefer not to.

I've been looking at some of the Mendel models on The Future is 3D, primarily because I want a large build envelope to make mech-like costume parts. Anybody who has experience using one of those, your experiences would be welcome.

I want to know how to get 0.1mm Z slices to be able to print like that first Yoda. What size orifice in the nozzle, I wonder, did they use?

Replicator ships with a 0.4mm orifice.

1) Nozzle width isn't critical for quality. Layer height is a function of the G-code generator. The most important factors in the quality are in how well you set up the g-code generator and how rigid your machine is. The threaded rod machines - prusa etc - are limited in how rigid they can be. Mendelmax and Mendel90 are your best bets here.

2) Easier than you might think with more recent designs. If you can drive a screw you can put together a reprap.

3) If you build it yourself you'll learn everything you need as you go.

1. Everything.
2. Even if you buy a prebuilt machine you will need to maintain it which is not necessarily trivial. Have you seen the UP!? It's not really meant to be user serviceable, but I get the impression that it is quite reliable.
3. Your machine will work until it does not. Troubleshooting is what DIY 3dprinting is all about.

As for size. The larger you go the more trouble you are asking for. Costume parts would be especially difficult due to size, being mostly thin shells, and would require a ton of supports.

I believe the default nozzle on the Replicator is 0.4mm.

I disagree that rigidity is important for quality prints. Consistent flow and consistent layers heights are the most important factors IMO. Neither requires a rigid frame.

Edited 1 time(s). Last edit at 07/23/2012 02:25AM by billyzelsnack.

I agree it isn't about rigidity, it's not much of an issue in printers, a MendelMax or a Mendel90 is about as likely to have say ZWobble as an original Mendel despite being massively more rigid.
Both are much easier to put together square though and that does matter.
Coming from CNC mills rigidity being an issue was my first thought looking at a Mendel, but although Mills and Printers both have 3 axis they experience very different loads doing very different jobs.
In DIY printers quality is about identifying and removing mechanical issues, electrical issues, then tuning the printer settings.

billyzelsnack Wrote:
-------------------------------------------------------
> I disagree that rigidity is important for quality
> prints. Consistent flow and consistent layers
> heights are the most important factors IMO.
> Neither requires a rigid frame.


Frame rigidity affects how accurately the affector is positioned, so it certainly does affect print quality.

Quote

MendelMax or a Mendel90 is about as likely to have say ZWobble as an original Mendel despite being massively more rigid.

Why? The X ends are constrained by the Z bars, which are rigidly attached at both ends to a flat sheet on a Mendel90 so the flexible coupling should absorb the wobble without disturbing the carriage.

---

[www.hydraraptor.blogspot.com]

Acumen Wrote:
> Frame rigidity affects how accurately the affector
> is positioned, so it certainly does affect print
> quality.

Sure, but you have to get everything else worked out first before its effects start showing up. What most people would consider quality prints are completely attainable without having a rigid frame.

nophead Wrote:
-------------------------------------------------------
> MendelMax or a Mendel90 is about as likely to have
> say ZWobble as an original Mendel despite being
> massively more rigid.
>
> Why? The X ends are constrained by the Z bars,
> which are rigidly attached at both ends to a flat
> sheet on a Mendel90 so the flexible coupling
> should absorb the wobble without disturbing the
> carriage.

Actually I don't have a 90, so I take that back, I do have a Max, mine is a 1.0 but I've relocated the motors to the bottom, and even with 1/4 inch acme lead-screws, with flexible coupler and open top, I still get some slight Z-Wobble. Actually I'm at a loss on the actual cause at the moment. My current guess is it's a function of the self aligning Igus bushings being too willing to self align to even the forces from the 1/4 inch lead screw which are actually pretty straight. One of the oddities in my case is the wobble is more visible in one axis that the other, when I get around to it I'll print some new X-Ends, but I'm done taking it appart and putting it back together for a while.

My point was more that the wobble is a function of the design of the Z axis, and the straightness of the parts and not the rigidity of the frame.
It's easy to come from CNC mills, look at printers and assume you should apply the same set of criteria to a printer you would for a mill, but they are very different machines with very different optimization criteria.
Clearly they need to be rigid enough for accurate positioning at whatever speed you run at, but you can get excellent print quality out of a Mendel derivative or a Huxley derivative, I don't think a well assembled Mendel would have lesser quality than a well assembled Mendel Max for example.

I have had nothing but slop with Igus bearings. They seem to have a 0 - +X tolerance and the specified rods have a 0 - -Y tolerance. Therefore worst case you have X + Y slop. This can be about 0.05mm, not good if you are using layer heights below 0.1mm.

LM8UUs are preloaded so there isn't any slop with the correct size rods, at least when they are new.The downside is more friction, which increases backlash.

---

[www.hydraraptor.blogspot.com]

I have some LM8UU's here and for that matter brass bearings, I just need to print the XEnds, but recently I've spent more time messing with the printer than using it and the wobble is very minor, so it'll have to wait a couple of weeks.

Hi. I am in a similar position and looking for a kit or fully built printer. I came across these guys today [www.nextdayreprap.co.uk].

They have a really good manual which shows the build process better than i've seen anywhere else.

[www.nextdayreprap.co.uk]

Every reprap I have seen pre made seems so wimpy and so bloody small too. Now get a 900mmX900mmX203mm then we are talking about something that could handle just about anything.

The trend is for them to get smaller, not bigger. Everything gets harder when you make the machine bigger in square law or higher power ways and a lot of people don't have the room or money for a machine that size.

---

[www.hydraraptor.blogspot.com]

For my needs I don't wish to 3d print keychain bobs I need something that can handle bigger items else, for me, it would be nothing but a toy. I get what you are saying but to be really viable as an end product we need something that will do more than knick-knacks and the dimensions I mentioned could handle the one offs, knick-knacks, small items and give the room to make something more valuable. Don't know what the Nema sizes would be but a 25mm or 30mm rails should be able to handle it.

Edited 1 time(s). Last edit at 07/24/2012 10:54AM by Dark Alchemist.

The biggest problems are combating plastic shrinkage over such large distances. ABS shrinks about 0.5%, so over 900mm that is 4.5mm. The machine would need to be an oven and then even the expansion of metal over those distances becomes an issue.

Also the price of filament to make something that big is huge so you would need to be looking at granules. The print times would be massive unless you use big layers.

Achieving both precision and speed over those distances is not cheap. It becomes a totally different machine and there is no way it will compete with a small machine for printing small objects with fine detail quickly and cheaply.

The current size is viable for lots of projects hence its popularity.

---

[www.hydraraptor.blogspot.com]

See, there has to be a way of making it bigger. I suppose you could make a mold with a CNC and inject the bloody thing but limiting a 3d printer to just knick knacks means it doesn't have as much viability as one thinks.

What is the biggest size you can go with our limited capabilities right now?

Yes it can be made bigger because there are commercial FDM machines that size but the cost of the parts and the engineering involved puts it outside most hobbyist's reach and it wouldn't be able to make many of its own parts!

The biggest things I have printed have been up to about 200x100x30, see [funwithelectrons.blogspot.co.uk]. Click on the pictures to see them full size as the scaling makes them look crap.

There are bigger reprap machines but I haven't seem many large objects printed, especially in ABS. It is much easier to print things in sections and weld, clip or screw them together as with that laser tag gun.

---

[www.hydraraptor.blogspot.com]

The most common size of a repraps print area is roughly 200mmx200mm about 8x8 inches, as much because there are available PCB heated beds in that size as anything else.
My MendelMax has larger travels in both axis, but I use a Mk2 heated bed on it, so they go unused.

I think you need to take a step back, build a printer and understand what the challenges are, I have a CNC mill and a lathe, and I know the first thing I tried to do with my printer was design and build a better one, while I think there is significant of room for improvement, especially if you relax the self replication constraint, for what they need to do and the constraints of the materials they work with, they are really pretty good designs with pretty well thought out compromises.
For small reasonably priced CNC Mills 8 inches would be a pretty large Y travel, the cheaper Tormach which would cost you ~$10K with tooling, has what 14x7.5inches of travel, so I'm not sure your at a disadvantage over anything but industrial level mills. Assuming the part can be practically manufactured by FDM.

Personally I'l prefer to be able to reliably and easily print dissolvable support before I'd care about anything in the 3ftx3ft size range.

Welds make a weaker part, right or is it like welding metal?

I think solvent welding gives full strength or very close to it. I use solvent cement normally used for plastic piping.

Thermal welding should also be full strength. The 3mm filament we use originally came from the plastic welding industry.

---

[www.hydraraptor.blogspot.com]

Well, I don't know about this thermal welding of plastic only the solvent but if I had a part (say a wing of an airplane model) that is 6 feet across how would I weld it together so everything is as if it were all one piece?

Best to make the parts interlock in some way so they register. Then you put some solvent cement along the surfaces that join and press them together and hold it in place until the solvent evaporates.

---

[www.hydraraptor.blogspot.com]

Oh, nice I hadn't thought about that but for such a big item would 2 per piece that are, say, 1.5-2" round a piece be better or .5 inches of 4-6?

I don't think it matters much as their job is done once the solvent has dried.

---

[www.hydraraptor.blogspot.com]

Well, just wanted to make it more structurally sound as I wouldn't want a wing to fall off in mid flight.