Let's talk about 3D printers...

So, I’m exploring 3D printers for my office. I’m certain there are many of you who either own one, have access to one, or know way more than I do about them.

I found this site that seems to be a pretty decent resource.

My preference would be something that generates fairly smooth parts that could be painted. Perhaps something that could even be cast, although I know that means the process would get much more expensive (right?).

We’ve got a budget of around $5k, so we’d be in the small business/prosumer arena, I believe.

So:
What do you use or what have you used?
What do you like/dislike about it?
What would you do differently if you could?
What are you using to generate files?

Appreciate it, everyone.

How big are the parts you want to build?

Ive worked with a few different hobby level 3D printers. I’ll give my feedback on a few of them.


Form 1+: Great for Visual Models ($3,300)
SLA

Pros:
Very Good resolution. We used it to print out watch cases at Fossil. Would pick up tiny details on crowns and dials.
Looks great in the office.
Pretty quiet to operate.
Software is great and super easy to use right off the bat.
Very accurate build time estimation

Alright:
Few Materials. Limited in color. Clear, Black, Grey, White, Wax, Flexible.
Build time is reasonable.
Parts have to be completely solid or allow drain hole for resin to escape.
If you have a super long print you can leave it over night. Nothing will happen if the print fails other than a resin tank mess.
Have to be there when the part is done. When the part is finished its suspended over the resin tank. If its there for a long time it has a chance to potentially fall back into the resin tank. If this happen the part willl melt back into the tank and make a huge mess than you have to clean.
Machine is a bit fragile. Cant get any dust/resin into the optics. So it cant be in a dirty shop.

Cons:
Very Messy in Post Process. Resin isn’t totally cured when parts done. Need to give them a bath in IPA. You’ll need a dirty room, or dedicated table for this. You also have to bake them under a UV lamp or place in the sun to fully cure the part.

Cost - Resin is $150 bucks a liter. Trays are also $60. Trays can only be used with one type of resin. So can’t use the same tray for black and white. Trays have to be replaced after 2-3 liters.

Support structure means one surface is always poor. The way the part has to be suspended always means youll have these little bumps where the support attached to the model.

Side Note: Fossil eventually moved on a super pro level printer. The Formlabs was just used to make super quick models if someone needed to see something and couldn’t wait for the turnaround time of the other machine.


Lulzbot Taz: Great for Functional Models ($2,200)
FDM

Pros:
Huge Build Area. About a 12" Cube
Wide variety of materials: ABS, PLA, Nylon, Flexible, Glow in the Dark, Brass Fill, Wood Fill, Carbon Fill, Every color under the rainbow etc.
Material is Cheap. $20-$30 a Kilogram.
Pre calibrated from Factory.
Open Source - Made in Colorado
Can produce strong structural parts.
Lots of online support for problems/issues.
Can set up a dual extruder if you want to print with dissolvable support.

Alright:
Fairly quick. Can be much faster if you modify/upgrade it a bit.
Large ABS parts require an enclosure to print without warping.
Bit finicky to set up. Once you get over the learning curve of an FDM style machine the repeatability is pretty great.
Bit louder than a SLA machince
Resolution is good, but not high fidelity. If your wanting to check out what a controller feels like. Or if a bracket will bolt on. Perfect.
Trying to evaluate a watch crown. No way.


Cons:
You’ll have to trouble shoot a lot of things at first.
Learning curve to really dial in settings on the machine.
Things like leveling a bed, un clogging a nozzle, replacing bed surface tape. All part of normal maintenance.

Side note: My personal printer at home for my side projects. Current job uses Makerbots for general size study and some quick prototypes before sending out for more high fidelity protos.

3D Systems CubePro Duo: ($2,800)
Stay Away. Do not recommend one bit.

Pros: Decent looking.
Large Print Area.
Dual Heads
Nice Hardware
Comes well calibrated
Wifi Enabled by default

Alright:
Decent build size 10" Cube
No heated bed means only PLA.
They say they support ABS, but without the heated bed your just asking for bad time on larger prints.

Cons:
Proprietary cartridges ($100 bucks for 0.5kilo) Vs $10 for 0.5 kilo elsewhere
Proprietary Software (terrible, terrible, terrible) (can get around, but at that point your better of just getting another machine if you hacking the firmware)
It’s expensive to run and have little control over certain things.
Enclosed build means if something goes wrong its very hard to access things to replace. (Filament broke in our feed tube. Never were able to get it out. So we were down to only 1 extruder.


PolyPrinter 229 ($2,400)
We had 3 of these at our local Makerspace. Made by a local company.
Awesome printer

Pros: Very Fast. Fasting FDM style printer I’ve ever used. Easily hit 250mm/s
Very Reliable. Dallas Makerspace had 3 of these and they ran about 12 hours a day nonstop and they help up fine. They are also mainly used at local libraries. So very easy to set up/use.
Heated Enclosure means large ABS parts not a problems
Also means it can be in a dirty shop and be ok.
Pre calibrated.


Alright:
Texas Based company who business model is based on supporting the printers locally. (hence the libraries being a main customer)
Not sure how well support is outside of Texas. Example Heads clogged, they’ll come out and fix it if your local.
Otherwise it’s like every other printer, you’ll have to learn to do maintenance yourself.

Cons:
Pretty Ugly Not a problem is if its in a shop, but for a design office might be different.

We’re not exactly sure, right now we’re looking at it as finding a use for the tool rather than really needing to have it for a particular project. Nothing huge, maybe an 8" cube. Or is that pushing it?

Thanks for that thorough review, Sain. Good stuff there.

Not to hijack, but I have some additional questions.

What is the annual overhead? Not only material costs, but labor costs for running and maintaining.

Where is the roi? If I spend $x at a dedicated print house, what is the number where I should bring it in house?

I may be the exception in that the largest 3d printer in Illinois and one of the largest in the country is located between my office and my home. I submit a file today, I pick up parts going to lunch or going home tomorrow. Accura white or clear SLAs are $20-$80. FDMs are the same. They can print an SLS nylon. Also the Tango clear/black/gray.

Even if most of my parts are in the accura, when do you bring it in house?

Good questions, iab, and certainly relevant to my questions. Anyone who can answer, please do.

@NURB: I’ve several machines in house and have worked with countless companies from a cross sections of industries and it nothing else that I’ve learned from all of this is… a company will find it hard to justify the cost benefit of owning/out sourcing for 3DP if they aren’t fully aware of their internal metrics. This may seem like such a simple thing to say but there is this intangible that we all know that getting something RP’ed is beneficial but just “how” much is generally not something companies know how to attach a $$ value to.

The reason why I start out here is that I ask, “Why is 5k the limit?” vs say spending 50k. Now some of this can be due to having the available upfront funds to shell out that much capitol at once. But here’s a quick example from a shoe company when it comes to metrics

  • Company sends design team to overseas manufacture for 2-3 weeks to visit factory at least 6 times a year.
  • Generally takes 3 months to get back the “correct” sample going back and forth with vendor. (i.e. fist samples are generally not the final)
  • Company buys 3D software and 3D Printer makes several samples in-house and then sends sample to overseas vendor.
  • Company sends design team 2-3 weeks but now only 3 times a year.

The key part here, when it comes to metrics, is that by purchasing the 3D software and 3D Printing in-house was paid for by not sending the design team the other 3 times. Think about that for a second… Not sending the team paid for the software and hardware. That’s an absolute metric that can be tracked, measured, and is quantifiable but only if the company is tracking the cost in the first place.

This may not be exactly what you may have been looking for but if this is not part of the discussion right now, then how can the best technology for your needs be truly figured out. There are definitely a wide range of technical factors to be considered (i.e. materials, post processing, resolution) and these cannot be discounted, but think that if it is to be truly leveraged then it can only be helpful if other factors as known as well.

I think that’s a valid example you’ve given.

Our company is pretty small in terms of the need for 3D printed elements. We’re mostly just trying to find a way to use the technology in our industry (I’m in the sign and environmental graphics industry, so most of what we’ll be doing are small accent pieces or architectural elements). We may use it for concepting, we may use it for finished parts. If we really find some good uses for it, we may invest in a 50K machine. As of now, we can afford to spend a bit to learn the ins and outs and see if it’s more than just a toy for me and the owner to play with…

If we can make things that our competitors can’t, that’s a win in our book.

Example for us at Fossil. We were able to check our CAD data and double check it on the wrist before sending to factory. We were also able to print out the CAD the factory gave us back and double check the fit of that sample before they committed to expensive samples. If we caught a change in size or form from that 3D print we could update the technical drawing and save ourselves a revision sample round. All this was possible before the end of the day.

Eventually it was enough to convince them to get a nicer machine in house.

That’s awesome. And to the previous example, there’s your metric to justify the cost of the machine.

We have an Eden 500V (polyjet). I think its about 250K and we got it about 10 years ago. I think it was one of the first ones in Southern California back then.
Pros:

  1. Great resolution
  2. Easy and not messy to clean.
  3. No need to sand parts
  4. Multiple materials (rigid, flexible, polypro like, transparent, etc)
  5. Parts within ours.

Cons

  1. High cost
  2. Material cartridges as expensive (about 1K)
  3. Need to plan accordingly. Can’t buy every materialand have it sit there unused until it expires
  4. Maintenance/Upgrades are expensive
  5. The headaches when the machine goes down. Somebody has to spend hours on the phone or troubleshooting.

I don’t have any numbers but over 10 years it has been paid off and made us some money. We use it internally for projects as well as provide RP services to other companies and design firms.
It’s extremely beneficial when you need to compare multiple iterations of a project or you need to test size and fit in order to speed up the design process. We do outsource some RP jobs when we need specific materials (SLS, FDM, SLA, etc) or when we are at capacity. With some of the RP companies we have to wait at least 2-5 days depending on the complexity of the parts. It’s getting faster because there are so many RP companies popping up so I can just image it will get cheaper and faster very soon with all the competition.

In your case I would track how much you spend in a year for RP parts. Once you have that number then you can make an informed purchase. Makerbots are cheap and fast but then you have to do a lot of finishing on the part. Another client bought a makerbot-like machine and I think he ended up suing the manufacturer in order to get his money back. Turns out that once the part got to a larger size, the resolution and build quality were horrible and inacurate. Another client with a makerbot still has us do some RP parts because they do cell phone cases and they need the high resolution quality.

Develop3D as a good breakdown of options covered in their 3D Printing for the Professional March 2014 Special Edition

You can download it here, by searching the special guides.
http://www.develop3d.com/downloads/index.php

How does that justify the cost of bring a machine in-house?

It certainly justifies the use of printed parts, but again, something the size of a watch would cost me $20-$40 in a very high-res print off of a $250K machine that they have down the street. I’d have to print a lot of those parts to justify the capital and running costs.

Thinking back on it, I suppose it doesn’t. I was looking at it from verifying what the factory is using, and being able to check that prior to a run being completed. Cuts out potential errors and re-works. But, yes, you can do the same thing sending to a rapid proto shop, I suppose…

Each machine will have its own ROI based on cost, maintenance, operation, etc. The $250K machine was an example. It works for us because of the volume we do and the size of the parts. If you’re doing small $40 parts, and the volume is low, and the RP vendor is close by then it doesn’t make sense to buy a machine. If you print parts weekly or monthly then you may need to look into it, specially with so many out there now.

I guess what I would do is get samples from the machines that are in your range. Yes, surface quality and accuracy is the biggest issue.
Maybe create a sample file and have each printer send you a sample so you can compare apples to apples.
Most of them also have reps selling them, maybe they can come to your office to give you a demo.If they don’t use support material, how do you create complex shapes? That was one of the issues we noticed with one of the vendors that came to our office. Basically once the part gets tall or wide, the machine would have trouble. Sorry, can’t remember the name.

After that, it’s a matter of cost, maintenance, upgrades, volume, etc. Only you can figure out if it makes financial sense to you.

I wasn’t on the logistic side of things when they made that call. So not sure if the cost save vs cost spent side.

But in Dallas high res prints were no wear near $40 bucks. Closer to $200-300 for a watch head locally (off an objet). More if you needed it faster.
Sure you could gather all the models and get a bulk rate if you were willing to wait for everyone else to send there models in at the same time and fill a bed… (But now your organizing across multiple brands dealing with multiple team budgets.)

Even shapeways in highres was close to $70 plus shipping time. Maybe in areas that have more prototyping needs RP houses are more prevalent. But these guys catered to Texas Instruments. In Dallas I could probably get a local guy to print out a watch on his makerbot for $20 bucks if I was lucky.

Also now thats its in house its not just one team. But the entire company that can utilize it. Watches (the 10+ brands), leathergoods, softgoods, bags, small goods, etc. Also from a business side its pretty cool to show licensors and investors when they come in.

Main thing though was time it saved by potentially eliminating a bad first sample.

Also high end machines have way more color options. Multiple colors to show part breaks without having to assemble different parts, dials in color. Clear Crystals sampled without having to cut/mill glass. Flexible bands printed to show new silicone straps. Those process in a RP house all cost extra and it starts to add up when you have 10+ brands doing the same thing.

For a large company I think it made sense to get a nice machine. For a smaller place, it for sure would have been overkill. But entry level machines have there place too.

Wow.

That’s staggering.

These are the guys I use. Get a quote, at least that is free. http://www.buildparts.com/

I mostly print in the Accura Clearvue. Great resolution and seems to be much more humidity resistant than the Accura 25 Although super glue will work with the 25 and you need to use epoxy for the clear.

Just had a tour from these guys http://www.tinkerine.com/

Produces some good quality PLA parts for appearance models. Easy to use software. $1899 USD for their pro series Ditto

Found this the other day: 3D Printer Comparison Chart

Might be helpful?