Hacker News Reader

If anyone is looking for a good introduction to using (FDM) 3d printing for production, this is a great video of how super-car manufacturer Koenigsegg approaches them:

https://www.youtube.com/watch?v=Dw6cs7opvzA

The key to making FDM parts work for your use-case is generally not buying the most expensive printer - it comes down to material selection, print orientation (others have mentioned anisotropicity which is incredibly important to understand), and designing for manufacturability.

I've spent the last three years running a 3D printing / rapid prototyping lab and love to teach people about the subject, so if anyone is interested in getting into the subject please feel free to get in touch.

Good video.

I'm curious about the laser scanner that he showed. What kind of resolution does it provide? And how much does it cost?

Really, I don't care so much about the one he showed, but I want to know if it is possible to get one with high resolution (fraction of a mm) for a reasonable price.

You can get to a mm and even sub MM resolution even without a laser scanner but with a good enough camera and photogrammetry.

https://hackaday.com/2019/03/13/3d-scanning-via-diy-photogra...

The real problem with printing these materials is that the part isotropy is poor, so you get the expected strength on the x- and y- axes but the z- axis (where all of the layers are stacking up) is quite poor. Turns out that printing things in a layer-by-layer fashion has its own cost.

I'm quite involved in this area, happy to share more if anyone is interested.

Layer separation can be addressed trough proper material selection and print settings.

Nylon and resin based filaments have excellent layer adhesion, also keeping your print area sterile and at a correct temperature is very important dust is your worst enemy.

Also too many people think that printing thin layers improves strength it doesn’t it just compounds the problem.

I’m printing large functional and structural parts in nylon with a 1.2mm nozzle and 0.6-0.8mm layer height and with a glass fiber reinforced filament I’ll have a material failure before layer separation.

I'm not talking about printing nylons, I'm talking about printing high-performance polymers.

Define high performance polymers as Nylon is also counted as such.

If you are talking about exotic stuff like PEEK then from what I know it has even better layer adhesion.

The most exotic thing I got to print with was PEI which also didn’t had issues with layers.

Reading more about this drone I’m not even sure if they used high end polymers for it.

The majority of their focus was around the continuous iteration from design to production workflow not material science.

If the blades they printed for the video are the ones used in flight then they printed them with ABS not some alien stuff.

Drones like this don’t actually experience that much force.

No friction, very little torque, little to no shock.

People print multi rotor drones all the time they work perfectly fine, this one is a bit bigger but nothing that would require a ground breaking advancement in the actual printing technology.

> Define high performance polymers as Nylon is also counted as such.

By high-performance polymers, I mean polymers suitable for structural parts and the replacement of metals like aluminum.

> The most exotic thing I got to print with was PEI which also didn’t had issues with layers.

I'd be really interested to hear about your experience with PEI. From the manufacturers and users I've spoken with, printed Ultem (PEI) has diminished strength along the z-axis because of its amorphous morphology. Do you know what brand of filament you used? Did you post-cure or perform any pre-print steps? How did you test isotropy?

Happy to take this to PM if you prefer.

Well without understanding what polymers exactly it’s hard to have a discussion about it.

My experience with Ultem PEI was pretty solid I did do post processing mainly annealing: https://www.thomasnet.com/knowledge/annealing-of-ultem-pei-/

In a toaster over I hacked for temp control.

I also sand my parts and cover them with a protective epoxy (UV blocker).

For failure I’ve tested sheer and compression failures including in parts that where physically compromised e.g. by drilling screw holes or pressing in threaded inserts.

Unlike say ABS the failure was never across the layer lines and there was no layer separation or tearing if and when the parts broke they broke fairly cleanly.

Now this is for my specific application and use cases there could be stresses that could result in layer separation.

However I didn’t observed a significantly diminished z strength even when holes were drilled across the layer plain rather than just through it.

If you want to replace metal in general sintering or a lithographic process is likely going to be your best bet but FDM can work for many applications and you can always solve your material weaknesses in your design by for example altering the layer alignment across your design to ensure that the maximum strength is always where it’s need the most.

I beams are also much stronger in a specific direction but it’s not seen as a weakness but rather strength.

Commodity resins are used to replace metals, and in structural parts all the time. Much of it comes down to part design, filler selection, etc.

For example, glass-filled polypropylenes are used all the time as metal replacements in vehicles (injection molded), even in some structural applications, depending on the exact demands.

Nylon is a pretty good structural material, especially glass-filled or talc-filled depending on abrasion resistance and rigidity required.

I don't have experience with the very expensive stuff (PEEK, etc) though, as usually that's much more expensive than the metal and doesn't confer many benefits except in pretty high end applications.

https://www.solvay.com/en/press-release/solvay-announces-win... 100% Z-strength on a DIY FDM printer with polyphenylsulfone (PPSU)

Thanks, this is very interesting! Wish there were more details on how they achieved this...

Here is the German forum thread of the person who did this: https://reprap.org/forum/read.php?336,840514,page=5

Thanks a bunch!

> share more if anyone is interested.

What would it take to print 45ft yacht hull?

And not just the mould, but the hull itself?

A very large printer..

http://www.thermwood.com/lsam_home.htm

Technically they printed it for a mold but that's because fiberglass is better and while you can can print with glass reinforced materials the fibers have to be quite short which reduces their tensile strength compared to composite sheets.

That said if you really wanted too you could print a functional hull, but it's not a good application for it.

Not every manufacturing method is appropriate we can CNC mill a car from a single block of aluminum too but it's not exactly a good manufacturing process for this task.

EDIT: HanseYachts AG apparently prints a 10m (~33ft) hull for one of their Yachts using a cellulose composite filament.

Cars are mass produced, yachts are much less so.

You could do that, but it wouldn't be a good idea. Fiberglass works very well for that application and is much cheaper, faster, and higher performance.

It seems very large portion of yacht cost is interior installation. Printing might lead to all sort of safety, maintenance, performance and comfort advantages that you can’t achieve with normal manual labour.

I’m guessing some sort of hybrid approach could win.

How far away are we from being able to replace CNC machining with 3d printing?

I don't think you want to do that? They're different tools with different uses

Based on the size of most UAVs and the fact that it's NASA, they're almost certainly using a large format printer from Stratsys or 3D Systems in the $250K+ range.

I would think so. One of the Stratasys engineers at their skunkworks division in Minnesota walked me through their process for printing something as large as a full-size table (upside down to minimize support material). The technology is really impressive but the uses seem few and far between.

I have quite a few friends that work at Stratsys you’ll be surprised just how unimpressive their print quality actually is compared to many hobbyist printers.

They are very consistent however especially at their volume and the material quality is also excellent which leads to very good production results.

From the video at least they print the prop blades on an Ultimaker and an old one to boot this is something you can easily do at home.

The body of the UAV looks like it is a skinned composite so it wasn’t fully 3D printed I’m guessing they printed a skeleton and then put composite sheets over it.

This doesn’t look like anything you can’t print at home if you know what you are doing and have either a high end hobby printer or a custom built one.

With not very much free time it took me a few months if trial and error getting clean prints from my cr10s.

I also learned not to upgrade slicer software and firmware at the same time.

I’m gonna start with a boat before anything that flies. Any tips on plastic and water?

The vast majority of FDM filaments will not be water tight. I'd recommend sealing with an epoxy coat after printing and sanding.

A quick acetone vapor dip of ABS parts does a good job, at least for the vases and such that I have printed.

[edit] I found a page with some pictures for examples:

https://rigid.ink/blogs/news/acetone-vapor-smoothing

ABS is a pain to print without an enclosure, and ABS for the most part doesn't absorbs water also it's not as much of an issue with the final parts (aside from Nylon that can expand easily) but rather that moisture caught in the filament boiling during the extrusion which results in your print looking like a sponge.

It prints fine on my non-enclosed delta. My printer isn't anything special, it's just, like you said, water absorption by the filament is an issue if it's left out in the open.

The key is simply to keep the filament in a drybox before use. It doesn't need to be anything fancy- I use a 5-gallon Home Depot bucket, a small jar of Damp Rid (Calcium Chloride), and a seedling heater mat to keep things warm. It works great.

I also have a CR10(s) pretty decent printer if you are willing to work out its quirks.

For water applications PET or PLA would be the best don’t use nylon and ABS is a bitch to print with without enclosures.

I would recommend you getting some protective epoxy for any prints you want to last for anything that is going to be in water just sand the part and put a marine epoxy layer on top of it the additional weight isn’t an issue.

I did a few upgrades, dampers, all metal hot end, other bed, Bowden tube, and have to install the additional support still.

Dampers were, mind blowing difference, I kept postponing it because I did not expect such big difference. Pity not all motors can be dampened

I work in a lab that has a large variety of printers from hobby grade to Stratasys FDM and resin SLA, polyjet.

You're right that the stratasys machines are unimpressive in their quality and precision; I can get a better end part in many cases with a $300 Prusa clone than a $25k F370.

However you're also right that their consistency is their main strength, and in my opinion this is why they are worth the money (if you need high throughput and low failure rate). The Stratasys printers we have are easily the most reliably machines we run, and have a close to 100% success rate while being completely hands off. The calibration routines, heated build chamber, and disposable build plates are key here.

The other area where Stratasys leads is with soluble support. I haven't found any generic material that comes close to SR-30. PVA is an absolute nightmare for reliability. If anyone has found something that they are happy with, please let me know.

You can probably find the same soluble support Stratsys uses the price is just going to be bonkers like all their materials.

I just print with normal support structures but E3D soluble support is supposedly very good (but very expensive like $280 for 5 lbs.).

So I’m not going to spend that much money I rather print something that would print with minimal supports and just cut them off.

Yeah Stratsys doesn’t care about the final quality of your benchy but it will work every time and their integration along the entire toolchain is excellent including the design phase which means your CAD is going to be aware of the dimensionality impact of the printing as well as other constraints like support material placement.