3D printers to make strong parts with decent finish

[rx8pilot]

I am beginning the research looking for a 3D printer to help make functional prototypes. Previously, I simply CNC machined everything because I had a shop. I no longer own that shop and want to work through iterations and ideas more freely and also make assembly fixtures.

I have had parts 3D printed and they usually have horribly surface finishes and they are modestly strong. Does anyone know where I should start looking for machines and materials to get reasonably strong and decently finished parts? This is not necessarily an entry level 'maker' machine but also not a $gazillion one either. Speed is probably not hugely critical either. Maybe something <$20k or so.

I have ZERO 3D printer experience so not really knowing what to ask for.

Thanks in advance.....

[Skimask]

My opinion means jack, but...
Got a Makerbot Replicator 2.  Totally a hobbyist machine, far from commercial, I get that.  I'm not printing out chairs, footstools, or car ramps, or anything like that.  But, when I need some stiffness in a project box, or circuit board holder, or whatever, I increase the infill % in the settings.  Gives more of a "web" inside the guts of whatever I'm printing.  The normal setting is somewhere around 5%-15%.  I've gone up to 75% printing out a shift knob for my lawn tractor.  It's still holding up.  The increase in print time is exponential...give or take a few.  Doubling the infill basically quadruples the print time.  And these are unquantified, untimed, unstatistical results.  Suffice to say, the more goop you print, the longer it takes.
As far as horrible finishes...  Sure, the finishes aren't perfect.  All sorts of lines and whatnot everywhere.  Then again, I don't use it for beauty.  If I don't like the lines and fuzziness of the end product, I'll either dip the whole thing in a vat of acetone for a few seconds, maybe a few times depending, or maybe brush on some acetone (see some of Ben Heck's videos for that).

Works for me.  But I've got low standards

[mairo]

If you have a chance go to a manufacturing expo and check each technology/manufacturer what they offer and more importantly check the build quality of the samples they have and make your mind based on your needs.

EnvisionTec Perfactory 4 Standard is the best 'low' cost system I have seen. All the rest  (especially bellow 10K) is rubbish (IMHO), you would be much better of with your current CNC.

But as I said, go and have a look for your self, these low cost, under 10k machines, you may find useful.
 

[Fred27]

If you're used to CNC machined stuff, then I think you might struggle to find a 3D printer up to the standards you're used to. I've got a 3D printer, a laser cutter and a CNC mill. The 3D printer gets used the least. Maybe a small bench-top CNC mill might meet your requirements at a reasonable cost? I'm happy with my CNC6040 - with rubbish electronics replaced of course.

Or a maker level 3D printer might be fine for checking fit and design if it's only you that sees the poor finish. When it's ready for a prototype to go to clients, you could get somewhere like Shapeways to produce it on a high end printer.

[thathardwareguy]

It likely depends on the size of what you need to print.. If it's small, an SLA printer like a Form2 provides very high quality surface finish, and properly post-cured they're actually surprisingly strong, but the build area is lacking for larger projects. You could also look at an Ultimaker 2+, it's definitely not lacking in build volume, but it's still an FDM printer that will have the standard horizontal lines in the surface, though it's probably one of the better FDM's on the market, with a 0.2mm nozzle the lines almost disappear, but your print time could quickly be measured in days.

With ABS, you can go with acetone vapor smoothing and get mirror like finishes, though it might distort some of the final dimensions, it'll take experimenting. Oh and FWIW, I've found more of a correlation between the number of perimeter shells in a print and strength than with infill, the same stipulations about print time/material use apply though.

[kaz911]

IMHO - Zortax M200 - mostly a plug and play printer that prints well. Company behind it is Polish and has an attitude issue towards customers. Lots of broken promises and features.

But printer still prints better ABS than most others do PLA. They have different kinds of materials dependent on what your aim is. Strong or great surface. Most printers like Ultimaker 2+ is not super for ABS - and ABS is strong - where PLA is not as strong and starts melting/warping at around 50 deg c.

[uncle_bob]

Hi

A simple 3D printer runs with a small handful of materials. ABS is the strongest that I have used and it's ... well... a thermoplastic known as ABS. It's no stronger than any other ABS and may be significantly weaker. Unless you CNC foam blocks, any material you are used to is stronger than ABS.

Simple 3D printers put down a line of melted plastic. The stuff piles up on top of previous layers and you get an object as a result. Precision is ok, but not anything to set the world on fire. If you look closely, you can see the boundaries between the layers. Dunking the part in solvent or other similar processes will melt the surface and reduce the impact of the layers in the surface finish. All of the techniques I've tried or seen used do nothing to improve the accuracy of the part. Sharp square corners and precise edges ... not with the finishing processes I've seen.

If I want a *good* part, I fab it up on the 3D printer and get it so it works. All of the nonsense with fitting this to that and "does it work" gets figured out quick and easy. The files then get shipped out to somebody with a *much* better system and for not a lot of money I get a good part. It's made of strong / hard stuff (think epoxy) and the surface finish is very good. Square corners / sharp edges / deep accurate holes ... no problem.

So simple answer:

Buy a cheap printer and keep money in the bank to do the fancy stuff outside. At $100 a print run ... much cheaper than the > $20K printer that does it really well.

Lots of fun !!

Bob

[exmadscientist]

At work we've got an Objet 30 Pro and the MEs seem to quite like it. It's not a cheap machine, either to purchase (~$40k about three years ago, probably much cheaper now) or to run (expensive custom resins), but it sure does give better results than those crappy extrusion printers. Surface finish has two settings, "matte" and "glossy". Glossy is excellent but only available on the top surface of a printed part. Matte is so-so on an absolute scale but pretty good by 3D printer standards.

One thing I've been continually disappointed by is what the thing can and can't do. It does fine on larger parts but for small stuff the material is just not strong enough to get the job done. (Specifically I was interested in replacing some parts in a stupidly designed 1225 coin cell holder on a prototype and was told it would probably be a waste of time as the printed parts would just break.)

To be honest, given the acquisition and running costs, if I were in your shoes, I'd probably try to find a local fab shop I liked and just subcontract to them. These things are not really cost effective for smaller shops to own yet, at least not for printers that produce usable output. (I will note that my colleagues disagree with me here, but not by much -- we all agree that it's fairly closely balanced right now.)

If the flexibility of having the printer in-house is worth it though, I think the Objet line is worth a look as it's a pretty good match to your requirements.

[rx8pilot]

Thank you everyone for the opinions and tips.

If you're used to CNC machined stuff, then I think you might struggle to find a 3D printer up to the standards you're used to.

To be clear, I have no expectations that a 3D printer will get anywhere near the quality of a CNC process  - for a couple of decades maybe. There are two main goals - shake down ideas and concepts during the design phase before they go to a CNC shop for prototyping. The models made at this stage can also be hand finished for trade shows and pre-release photos.  The second goal is to make various assembly related fixtures to hold, organize, etc.

I will take a look at the printers mentioned and try to better understand the materials options and the cost of printing. There was one that I saw that prints a continuous filament for strength  The Markforged Mark Two. It is a low-ish cost printer for sure but seems to focus on strength of the finished parts.

If I have to spend spend $25k+ to get the parts I am after - a service company would be a better way to start. In-house is always very convenient, but I have to figure out a balance between cost and convenience.

Maybe I should start collecting some sample parts. I am guessing that manufacturers will freely give out samples that I can use to compare. Hopefully the samples have a build time and materials cost that I can use as a reference.

[Spikee]

I have a Ultimaker 2 and it is surprisingly good printing quality.
I usually print with low - mid settings but at high is is (in my opinion) pretty much the best you will get with a extrusion type 3d printer.

If you really want a smooth product with a extrusion type 3d printer you have to use plaster or something like that to get it as smoot as you want. I even think this is the case with SLA and things like that. Yes shapeways is better ... but for a 2k machine this is quite good.

It all depends on your end application. I would advice you to look at a few in person.

[rx8pilot]

I have a Ultimaker 2 and it is surprisingly good printing quality.

How is the software? Is it a fairly simple 'send to printer' thing? Does it accurately estimate the build time? Can it estimate material volume (for costing)?

[uncle_bob]

Thank you everyone for the opinions and tips.

If you're used to CNC machined stuff, then I think you might struggle to find a 3D printer up to the standards you're used to.

To be clear, I have no expectations that a 3D printer will get anywhere near the quality of a CNC process  - for a couple of decades maybe. There are two main goals - shake down ideas and concepts during the design phase before they go to a CNC shop for prototyping. The models made at this stage can also be hand finished for trade shows and pre-release photos.  The second goal is to make various assembly related fixtures to hold, organize, etc.

I will take a look at the printers mentioned and try to better understand the materials options and the cost of printing. There was one that I saw that prints a continuous filament for strength  The Markforged Mark Two. It is a low-ish cost printer for sure but seems to focus on strength of the finished parts.

If I have to spend spend $25k+ to get the parts I am after - a service company would be a better way to start. In-house is always very convenient, but I have to figure out a balance between cost and convenience.

Maybe I should start collecting some sample parts. I am guessing that manufacturers will freely give out samples that I can use to compare. Hopefully the samples have a build time and materials cost that I can use as a reference.

Hi

All of the simple 3D printers feed in the raw material as a "filament" off of a spool. Depending on which one you get, the diameter of material on the spool needs to match up with the head on your machine. Since the material is fully liquified in the head on the printer, there is no real strength relation between the filament going into the head and the goo coming out of it. There are people who fiddle a bit with materials to vastly improve the strength for only a small increase in cost. That generally translates to 5% more strength for 5X the cost

Lots of Fun

Bob

[digsys]

Why not use those on-line services, and at the very least, compare results. That way you get to compare $10k - $250K printer outputs with little cost.
We have stuff done on local titanium printer, and after a little bit of polishing, up to 95% machine strength. In fact, metal printers are starting to appear.
Maybe an option?

[rx8pilot]

https://markforged.com/mark-two/

The website is a little short on details, but it appears that it has two nozzles. One is for molten material and the other is for embedding a continuous (non-melting) carbon or kevlar filament for strength. Seems pretty cool at first glance.

[digsys]

... it appears that it has two nozzles. One is for molten material and the other is for embedding a continuous (non-melting) carbon or kevlar filament for strength. Seems pretty cool at first glance. 

Also the price seems pretty cheap? Too cheap???

[rx8pilot]

Also the price seems pretty cheap? Too cheap???

I think it was a Kickstarter which are generally over the top ambitious and overstate capabilities. My biggest concern is the website that essentially says "It just works, trust us" It's a few steps above dirt cheap, but still in the family of low-cost starter units.

I will try to find something about it from real owners - maybe dig through a 3D printing forum somewhere.

[Spikee]

I have a Ultimaker 2 and it is surprisingly good printing quality.

How is the software? Is it a fairly simple 'send to printer' thing? Does it accurately estimate the build time? Can it estimate material volume (for costing)?

I have it now for 3 years or so and have only adjusted the bed level twice, no other maintenance.
Only bad filament can cause it to jam as in not putting any plastic out.
This has happend 3 times in 600 hours of printing with this thing. The plastic is quite strong, do not underestimate that.

The ultimaker 2 has a socket for a second extruder but as far as I know of the company has never implemented it (probably because of good reason).

In my opinion a second extruder is a gimmick and you will never use it. The software is cura. It is accurate , simple, works always.

The build quality is top notch. And the print quality is as best as it gets with the extruded type printers. All these printers can use the so called glass fiber filaments etc... But in my opinion that is not needed.
https://www.3dhubs.com/talk/thread/thoughts-about-new-carbon-fiber-filament-colorfabb

I have seen prints of makerbots etc but I was not impressed. Why spend double / triple the money for a equal or worse machine.
Ultimaker has been around for quite some time and makes great products (here in the Netherlands).
 
I always use the stock medium settings on my machine and to print you have to save it to a SD card and put the sd car in the printer.
You can download cura and have a play-around with it.


In my opinion it is the best machine you can get in the extruder category. All the cheap machines just suck, all the expensive machines do the same (or worse) for more money.

If you want really strong than just print your part and use it as a mold for harder material. Heck even use it as a mould to pour aluminium in it

[ConKbot]

It all depends on the expectations. Designing a part to be manufactured is key. 3D FDM/FFF printing introduces a grain to the plastic, so orientation during printing is important. But obviously that can vary with the part, surface finish requirements, machine dimensions, etc. If the stresses being put on the part is compatible with the results from a FDM/FFF printer, PET(G) and polycarbonate filaments can make for very tough prints.
SLA machines are high detail, not as strong since its unfilled resin. SLS opens up the possibility of using glass filled nylon, which makes for strong parts(fairly equal in all directions), ok detail (you cant do tiny parts in high resolution like you can SLA) and hollow parts have to have some way to drain out the unfused powder. 

Given the amount you're considering spending, I'd definitely get some prints made so you can evaluate the process vs your needs.