More tinkering than modeling? Expecting/hoping for too much?

[joatrash 1]

Hello everyone.

Rest at ease, while this is my first post, it's not yet another "which printer should I get" query. Rather, it's more along the lines of "Should I bother getting one at all or keep using commercial vendors?"

Here's the thing. I'm a bit of a prop and modelmaker. For the last few years I've been printing stuff for my kits on high-end machines through various vendors ( high detail MJM and SLS) but it become frustrating. It's horribly expensive prototyping stuff that is mainly for my own use, I don't like waiting two weeks and the restrictions some vendors have on printing even the most unrealistic scifi weapon is a big hassle. I've got projects lined up right now that will definitely cost more to print at a commercial vendor than if I got an Ultimaker 2 (or even a Form 1).

Naturally I've been considering getting my own printer for a while. My concern however, is that all of the prosumer machines available right now are either more of a hobby in themselves (you have to LOVE the tinkering process) or will just not do what I want them to do.

I've uploaded photos of a few models in my gallery:

http://umforum.ultimaker.com/index.php?/gallery/album/336-models-done-on-high-end-machines/

Ultimaker 2 seems to have the most detailed prints out there (aside from the resin printers like the Form 1 which I am also considering). I've seen the 20 micron Yoda and the devil head. But what does it actually take to get those results? I've been scouring forums in seach of answers and am none the wiser. Am I looking at lots of failed prints or drastic changes in setup every time I print a new model? Am I going to spend more time getting the machine to do what I want than modeling?

I realize that it will take some more work if I want to get the same kind of detail I need, that it will take a bit of time to learn the printer, and that I might have to send out for SOME super-detail jobs. But the question is- how much? Currently, even the most high-end machine will turn out parts that needs SOME cleanup but my fear is that these prosumer machines will need so much that keping things like perfect symmetry in parts will be impossible. I'm more interested in the modeling and finishing the model than the printing aspect.

One thing I can say however, is that print time is not a factor I care about- I'm used to waiting 10+ days for a package to arrive, so waiting 15 hours for the printer to finish is not a problem.

A couple questions, if you'll excuse the newbness of them:

 

• Will an Ultimaker2 print 20 microns (i.e. golden yoda and devil head quality) with what comes in the box (aside from perhaps upgrading software) or will I need to experiment and mod hardware like crazy to get there?
  
• If the UM can do HIPS (as I have seen indicated here and there), why are so few using it? I mean... it's fantastic material to work with if you're used to gluing models together... and it sands very well.
  
• When modeling for one of these printers, does the mesh have to be CLOSED if it is orientable for it to print? Often when I export my models (from a high-end prog) Netfabb says they are not closed even though I made triple-sure that they are before exporting. It usually isn't a problem when I send out for prints.
  
• Can I have intersecting geometry (i.e. one closed rectangle stuck halfway into a closed sphere to create ONE solid model?) I ask because I sometimes do very complex models made up of dozens of smaller pieces that intersect. The "big" print vendors can handle these types of models.
  

So what do you all think- looking at the stuff I'm used to doing, will I have any real use for a machine like the UM2 or should I wait?

Thanks in advance,

//Joe

 

[mostlyatomicbob 34]

Joe,

As a fellow modelmaker/propmaker who has been working in the Uk film industry for 15 years, I can say yes!

If you buy a UM2 you will do a lot less tinkering than the UM original (I have both). You will spend time tinkering and tuning but not that much, of course it's a learning curve and don't expect miracles to start with but these machines are bloody good.

Printing at 0.02 is really not necessary. 0.06-0.08mm is fine enough.

PLA is not as good as Hips to sand, it gets a bit 'chewy' went it get hot, but you can sand it and fill it. I usually prime it with plastic primer (dusted on to begin with) but then I use car body filler. and I have got some really nice results.

You can stick PLA with Dichloromethane (dichlo) and it bonds as well as styrene.

As for holes in models, Cura is pretty good a t fixing stuff, It is a free download so give it a whirl.

Bob

 

[valcrow 145]

Hey Joe,

I come from a 3D artist background so I had many similar concerns that you have. Having not much experience building physical things, I worried that a printer would require too much electronics/machining/tinker know how to be of practical use. I'd have to say there's less tinkering than I expected to get good results, but there were also some limitations that aren't immediately apparent.

I purchased the Um1 Kit about half a year ago. And it seems they've come a long way to make the UM2 require less tinkering. But I was up and running printing stuff at 40-100 microns within a few days with no modifications after I got the initial problems resolved.

I've never gone down to 20 microns as it would just seem impractically slow. And while it sounds great, you're not really gaining that much. While your Y axis is 20microns, your X and Z is still 400 microns (your nozzle is .40mm) Which makes it kind of challenging for really small parts.

I think the other big issue you'll run into building those props is the freedom offered by commercial printers. You don't need to care about parts floating in the air or overhangs and such. With the UM and all other FDM printers, anything that's floating won't print properly.

You DO need closed objects, intersecting faces is a no-no in my experience. While it 'may' print. your ideal 3D models for print needs to be watertight, no duplicated surfaces, single mesh, no intersections. It's not really a big deal as long as you're ok at modeling since you can just boolean everything together when you're ready to print.

The biggest thing would be cost saving and convenience. You'll save a ton of money especially on the big objects that you're going to sand down and finish anyways. And you can test parts really quickly unlike commercial prints.

Not sure about HIPS, but PLA works great. $40/KG, doesn't shrink (much), you can sand it and it's pretty durable.

You might still need printing services for small detail parts, but for the medium sized stuff the UM would be perfect.

Big wall of text if you want further reading: http://umforum.ultimaker.com/index.php?/topic/2128-new-ultimaker-kit-review-first-samples/

 

[Daid 306]

 

• Will an Ultimaker2 print 20 microns (i.e. golden yoda and devil head quality) with what comes in the box (aside from perhaps upgrading software) or will I need to experiment and mod hardware like crazy to get there?

• If the UM can do HIPS (as I have seen indicated here and there), why are so few using it? I mean... it's fantastic material to work with if you're used to gluing models together... and it sands very well.

• When modeling for one of these printers, does the mesh have to be CLOSED if it is orientable for it to print? Often when I export my models (from a high-end prog) Netfabb says they are not closed even though I made triple-sure that they are before exporting. It usually isn't a problem when I send out for prints.

• Can I have intersecting geometry (i.e. one closed rectangle stuck halfway into a closed sphere to create ONE solid model?) I ask because I sometimes do very complex models made up of dozens of smaller pieces that intersect. The "big" print vendors can handle these types of models.

 

 

I'm from Ultimaker, so I cannot give you a proper answer to "should I get one" as my opinion is not objective enough.

I can however, answer your questions.

The UM2 can do 20microns out of the box, but, I recommend starting off with 100 microns, that is usually enough quality for most goals. This was done at 150microns: http://daid.eu/~daid/IMG_20130429_142634.small.jpg

It's best to have a closed single mesh. However, the software (which is my responsibility) tries to do an best-effort when you have holes or intersecting meshes. You can already download Cura (it's free) and check out the features and results. It shows how it will print something in the "layer view". It also has an X-Ray view that shows holes in red, which can be quite useful to spot big holes. (Tiny holes usually do not matter at all)

Forcing requirement of water-tight messes is really annoying, so that's why our software does an best effort and does not bitch when your model isn't perfect. It usually turns out fine as long as it does not have any strange internal faces.

Most likely you will need to check the "Fix horrible, combine everything - type A" in the expert settings when you have multiple intersecting geometries. (I'm actually considering making this the default) I'll skip the details for why this is needed for now :wink:

Commercial fab houses have people checking and fixing your models, which is one of the reasons why production is quite expensive at them.

We have printed HIPS on the UM2. However, our material supplier warned us to print it in a well ventilated room, as there are some gasses released during printing. We're not yet sure if we're going to sell HIPS ourselves. There are some environmental questions unanswered right now on HIPS.

 

[joatrash 1]

Thanks for the very informative answers guys. Looks like I'm definitely leaning towards an UM2 some time early in the new year.

I'll do some experimenting with Cura. The holes that I sometimes get in my meshes seem to be the result of the exporter generating two identical vertices in the same place but not welding them. So, the models are actually watertight. The intersecting meshes however, might pose a problem. Boolean operations to fuse several meshes are usually messy affairs so I tend to avoid them.

I guess my next move is to get a printed sample of pla and ABS to see how they behave when glued and worked.

 

[joatrash 1]

Hi again all.

You know, the more I read, the more unsure I get it seems. Or rather, the more sure I get that the sub-$5k printer industry just hasn't come far enough for what I need... yet.

Being honest with myself, if I'm not getting these results:

http://www.hive76.org/insane-3d-printing-resolution-ultimaker-under-the-micro

...in 8/10 prints within my first week or two of owning the thing I'd probably get pretty frustrated and likely give up. When I need to print, it will be very specific jobs and I'm not at all into printing for the sake of experimenting or tinkering. Of all I have seen, only the UM2 and Form 1 come close to what I need (and the latter seems to be so full of startup gremlins it's a miracle when someone gets a print... if they're lucky to even have taken delivery of one, lol!). But warpage, sagging, banding and all the other things worry me.

My other option is to keep using my two go-to vendors for printing. (I use one for the ulta detail stuff and another for pl2000 prints). The only thing is that I have one very big job planned that will cost 1/3 of the cost of an UM2 to do in pl2000 that is mostly flat surfaces... my idea was to sibk hat money into a printer and do it myself, but I don't have the time for too much tinkering.

Maybe there's a way I could get one of my models tested on an um2 somewhere as a proof of concept. (I know there are a couple Swedes here faily local to where I live.) Any ideas?

Thanks again for all the friendly respoonses I've gotten so far, and Happy Holidays!

//Joe

[gr5 2,235]

SLS is far superior to SLA or FDM - you don't have to worry about overhangs and support and shrinkage is an issue but not nearly so much. Like injection molding, with FDM you need to design for the technology - while you are designing the part you often have to worry - "what's going to happen when this shrinks? What will warp?" and "how do I support this?" and so on. SLA has a much much finer resolution but still has support issues.

And with injection molding there are also tons of issues - you might have to design a curve tighter or more loose in the mold so that when it cools it changes shape to the desired shape. You can't have some sections too skinny or the plastic can't get in there and so on.

SLS is great but you really need a clean room, bunny suits, and so on and it's only economical if you print 100 parts at a time. And it's still covered by patents I believe.

So I don't think we will have the "perfect" 3d printer in the next 5 years.

Having said all that I love my UM printers. I love being able to design something and print it up and use it right away. I like the ability to customize and so on. It's just fun. But I care about functionality - not art. So we probably care about different things.

The UM2 can print beautiful things but... well... it's complicated.

 

[joatrash 1]

The big players don't really seem interested in the prosumer space, besides the careful dabbling that's going on. I don't think things will progress at the pace of current hype either. People compare it to the evolution of the 2d printer but the main difference is that the average person doesn't have any real use for a. 3d printer. I see them evolving at the same pace as maybe Wacom tablets.

I could deal with the design issues of SLA if the general tech were more evolved and if there was a stable/reliable source but buying something like a F1 feels like it comes with the risk of being a very expensive paperweight within a short amount of time.

The UM2 seems like a fine machine, but I might need more predictability before I get one... we'll see.

It's a shame 3D Systems haven't done a prosumer MJM machine... that'd be something!

[joatrash 1]

Also, you just added shrinkage to my list of worries, lol! I design lots of precision addons for props. Shrinkage in sls is barely noticeable, but if you print too thick parts they can warp...

[gr5 2,235]

Shrinkage in sls is barely noticeable

 

I think they account for that as the entire part shrinks the same amount at the same time after printing is done. So they just print it 1% (or whatever the value is) larger. But with FDM things shrink as you are printing. So it is much more complicated.

 

[joatrash 1]

How much does ABS vs PLA shrink?

[gr5 2,235]

Not sure but ABS shrinks much much more. You need a heated bed or even a heated chamber to print anything more than one inch across.

 

[roger-fujii 0]

Bare in mind that your headaches are going to come from non-uniform shrinkage (ie curling) (though if you had uniform shrinkage, it will pop off the bed). There's some ABS with additives that try to minimize (note the word is not "eliminate") the problem. As your results will depend on variables that isn't in the printer's control ( the filament properties, workspace environment ), 3d printing /will/ require some tweaking. If you are familiar with baking ( bread, cookies....), it's like that.

 

[jdaiglejr 0]

There is a new SLA machine, Pegasus Touch, on kickstarter, at about 60% cost of Form 1(and 7"x7" build area, but it's "order now, wait 'til summer"):

http://www.kickstarter.com/projects/fsl/pegasus-touch-laser-sla-3d-printer-low-cost-high-q

It's manufactured by laser specialists/laser cutter manufacturers.

 

[LePaul 331]

Yeah I saw that Kickstarter mentioned on CNET

I researched, agonized, debated about getting an Ultimaker for my R2-D2 project. I wanted to be able to prototype R2's varies gadgets and also print various brackets/mounting pieces in plastic for the many electronics my R2 has inside.

I bought an Ultimaker kit last September and just missed the UM2 release by 10 days (boy was I pissed...) To save money, I went with the kit but bought the Ulticontroller and the second extruder upgrade (yet to install).

Short answers for you...

* The build, in hindsight, wasn't too bad. I learned a lot about how these things work

* This forum provides a wealth of support and you'll get answers quickly.

* If you are hoping this will be plug and play, we are years away from that. You will have to be prepared to dedicate some hours to learning, calibrating, tweaking and so on. Once you get the print built, you have the bed leveling, which can be challenging. Then you learn the best temps to print at, how to use Cura and what the Advanced Settings will do. In short, you won't really be able to unbox your assembled printer and instantly print out your finest works. We aren't at that level yet. But, if you wish to learn this technology along with us, the results can be very rewarding

* If you want to draw and see your creation in 3D minus the headaches of learning the printer, you might want to consider using a service like Shapeways.

I'm still very much a newbie but I've learned a lot from reading in here and my conversations with GR5 and others.