This Simple Change Made My 3D Prints 30% Stronger!

How would you 3D print this part here? This  is a vacuum hose adapter for my belt sander  

that will be under constant strain during  use. Would you lay it flat on the bed,  

just as it loads into the slicer or would you  place it onto one of its ends but risking it  

breaking during use? I actually went  with neither and chose something in  

between. But does that really help? In this video, we’ll take a deep look  

at the crucial topic of print orientation — how  it affects appearance, support material, and,  

most importantly, part strength, which I’ve  tested with a bunch of tensile samples. So  

let’s find out more. Guten Tag everybody,  I’m Stefan and welcome to CNC Kitchen.

This video is sponsored by  KiwiCo. More on them later.

We often overlook the importance of choosing  the correct orientation when printing a part.  

I admit I often do the same and rely on the  auto orientation feature in today's slicer,  

which optimizes the print for sufficient  contact with the print bed and minimal support  

use. This improves printability but doesn't  necessarily find the orientation for the best  

strength! While many understand that layer  adhesion is a common issue in 3D prints,  

few realize how much the performance of a part  can vary depending on its orientation. Sometimes,  

the best orientation isn't  the most obvious choice.

Let's begin with a simple bracket.

First of all, there’s an indefinite number of  orientations you could print this part in. But  

the two orientations most of us would probably  choose are on its side or on one of the larger  

flanges. Printing it on the big face is  great for bed adhesion and reliability,  

but we face the problem that,  at the point of highest stress,  

we have layer boundaries — right where the  part will eventually break when it’s loaded.

One of the biggest challenges with  filament-based 3D prints is their  

anisotropic strength. Anisotropic means that the  

properties of a material aren’t the same  in every direction — and for FDM prints,  

that means they tend to break much easier if  you pull the layers apart instead of along them. 

And the difference in strength can be significant.  I printed a ton of samples for this video,  

especially to look at the strength of different  print orientations, but let’s start with the  

basics and with part printed at 0.2 mm layers  on the Prusa CORE One in Azurefilm Green PLA.

One set of samples was printed flat, whereas the  other one was rotated 90° and printed upright.  

I then loaded them one after the other into my  tensile tester and pulled them until they failed. 

The horizontal samples had an  average strength of 63 MPa,  

whereas the ones printed standing failed at  only half the load, 31 MPa. And this is very  

typical for 3D prints. Practically no filament  print is equally strong in all directions.  

The newly deposited layer will only  partially melt onto the one below,  

and the polymer chains are primarily oriented in  the printing plane, causing additional weakness.

And just like in 3D printing, there’s  something special about learning by  

doing — and today’s video sponsor brings that  spirit right into your home so your kids can  

become the innovators of tomorrow. This video is sponsored by KiwiCo,  

who make amazing crates filled with hands-on  projects that bring science, engineering,  

and creativity to life for kids of all  ages — and even for adult kids like me.

We’ve actually been KiwiCo subscribers  long before they became a sponsor,  

and we still have tons of their materials  because our kids love them — they’re such  

high quality and definitely not throw-away items.

KiwiCo sends these crates right to your doorstep.  The Panda Crate is for babies & toddlers,  

comes every other month and helps lay the  groundwork for learning and exploration. 

This time, we got a beautiful wooden barn for my  one-year-old, and it’s always incredible watching  

her have fun while developing new skills. KiwiCo always includes extra material for  

parents to help guide play and learning,  which makes this way more than a toy.

And with my five-year-old, we built a  treasure chest from the Koala Crate,  

which couldn’t have been more perfect because  she’s currently super into treasure maps. 

Seeing her actually build  something herself, use tools,  

and learn new techniques — that was such a  great family activity and a proud parent moment.

KiwiCo has something for every age  group and interest. The project I  

personally enjoyed the most was  this basketball trash can — it  

was inspiring to see how several simple  manufacturing techniques came together to  

make a really cool mechanical trash  can for some fun garbage disposal.

If you’re looking for something meaningful  this holiday season — for your own kids,  

nieces, nephews, or grandkids — KiwiCo crates  are the kind of gift that inspire curiosity,  

creativity, and hands-on discovery. 

You can go to kiwico.com/CNC or use the code  CNC to get 50% off your first monthly crate. 

And if you’re getting a Panda Crate,  that code also gives you 20% off.

So check out KiwiCo — their  crates are so much more than  

just toys. They inspire kids to  become the makers and innovators  

of the future — and they’re a fantastic  way to spend time together as a family.

Now back to the 3D prints. Depending on the  material, printer, and settings you use,  

layer adhesion is typically around  50% of the in-plane strength,  

as we’ve just seen here. It can go much  higher with some materials — but also drop  

significantly if you, for example,  print ABS with a lot of cooling.

So, to increase the strength of our bracket  without redesigning it, we simply place it  

on its side so that the stresses when loaded  are more aligned with the print lines — where  

the part is strongest. But this can also lead to  other problems with the bracket. Especially with  

warp-prone materials, printing on the side can  cause the ends to lift and lead to print issues.

And if we replace the standard through-holes with  countersunk holes, we face another problem: when  

installing a regular wood screw into one of these  holes, the tightening force will try to split the  

part. The layer lines again go directly through  that hole, creating a weak point — which wouldn’t  

have been the case on at least one side of the  bracket that was printed flat. So you see, even  

with such a simple part, the clearly strongest  printing orientation has its own set of problems.

When we go one step further  and look at a corner bracket,  

we quickly realize that there’s no ideal  printing orientation anymore. Regardless of  

which side you place it on, there will always be  one side pointing upward where the part is prone  

to premature failure due to weaker layers. In such cases, we either have to compromise  

strength in one direction or  come up with another plan.

So, if we have a part where, in  a normal printing orientation,  

one side is very strong while the other  is very weak, is there a way to improve  

the behavior or at least make the strength  more balanced? I mean, if the strength of a  

3D print is best when the load is along its  layers and weakest perpendicular to them,  

maybe we can just angle it to get something  in between. But does this really help?

Since I had the same question, I didn’t  only print tensile samples horizontally  

and vertically — I also printed them at 30°,  45°, and 60° angles and put them to the test.

The results were seriously interesting. The  horizontal samples from before were at 63 MPa,  

and the vertical ones at 31 MPa. The three coupons  printed at a 45° angle had a strength of 40 MPa.  

That’s still far from ideal, but angling the  part improved strength by 27% compared to the  

vertical reference — which is significant. The 30° samples were even a bit stronger,  

failing at 44 MPa, whereas the 60° parts  were weaker, at only 36 MPa on average.

This aligns quite well with the analytical  solution for orthotropic materials. The  

interesting thing here is that the strength  doesn’t change linearly between the best and  

worst angles but follows an S-shaped curve. The reason for this behavior would be too  

deep for this video, but the takeaway is  that small angle changes from the ideal  

orientation don’t do much — but past a certain  point, the strength starts dropping fast. 

On the other hand, that means if  you have a layer adhesion problem,  

angling the part only slightly won’t help much  — you should aim for at least a 45° angle,  

which is a good compromise between  printability and strength gain.

So, coming back to our corner bracket,  

printing it on its tip will improve  the strength of the weakest member,  

while slightly reducing the strength of the  others — but that’s still a good compromise.

Printing a part at an angle isn’t  always straightforward, though,  

because it often leads to minimal bed contact —  in the case of the corner bracket, just a single  

point, which obviously won’t work. If I didn’t design the part myself,  

I often use a simple slicer trick: rotate  the part into the desired orientation,  

then use the cut tool to remove a  small section from a corner or edge,  

just enough to create stable bed contact.  Add a brim, and with good adhesion,  

that often prints fine. But you can also add  some manual supports for extra stability.

But you need to be careful here. I thought I  was smart and did exactly that for my angled  

tensile samples — but all the prints with supports  failed, while the ones without printed flawlessly. 

What happened was that even when printing PLA,  parts always try to warp and curl up. Initially,  

the supports prevent this, but since part and  support are only loosely connected — you want  

them to separate easily later — tension  accumulates in the part until the support  

releases. Then the part springs up slightly, and  the nozzle crashes into it, ruining the print.

You can get around this by designing  custom supports that are only rigidly  

connected at a few points — I’ll link  to a video about that from the Slant3D  

channel below. The “dirty” way is to just  decrease the Support/Object XY distance in  

the slicer and live with supports that are  harder to remove, but hold the part firmly.

Another often-overlooked benefit of  printing a part at a different angle  

than initially think is surface appearance. If you print a cube, the top, bottom, and side  

surfaces all look different. But printing the same  part at an angle means all faces are printed with  

perimeters — and basically look and feel the same. I mean, sometimes that’s not what you want — many  

products are designed so that the bottom  layer remains visible because it looks so  

clean — but especially for organic  prints, this can be a game-changer.

Just look at this relief print of Mount Fuji. One  version is printed as it loads into the slicer,  

the way also the author printed it, but that  orientation leads to significant stair-stepping,  

especially on shallow slopes. If we tilt the part 90° on its side,  

it still prints great and produces a buttery  smooth surface, even at 0.2 mm layer height. 

This has limits, of course: as long as the slope  angles stay below roughly 45°, everything’s fine;  

anything much lower and the perimeters start  printing into thin air or you’ll see exposed  

infill. But you’d be amazed at what a 3D printer  can do with good cooling and good materials.

And this is not the only reason why the surface  quality can benefit from angled printing. I’m  

sure you’ve seen notches on outer surfaces —  the most famous being the 3DBenchy hull line,  

where a notch forms at the deck level. This is called a shrink line and happens  

when the thin hull is suddenly pulled together  when the full deck starts being printed and  

the plastic shrinks when it cools down. A simpler example is this arch: at first,  

the two thin sides print independently until  the roof connects them. The plastic contracts,  

the sides are pulled inward, and the next layer  is printed offset — creating the visible notch. 

Printing this part at a slight angle not  only improves the overhanging surface and  

almost eliminates the notch because  only a small portion of the roof is  

printed in each layer, reducing deformation. So, if you print the 3DBenchy at an angle, the bow  

line basically disappears because the deck isn’t  printed in one layer, but gradually over many.

That doesn’t mean you should tilt your  Benchy — but it’s important to understand  

why this happens and how angling can  reduce shrink lines in FDM prints.

With everything we’ve learned, let’s come back  to the vacuum hose adapter for my belt sander. 

Especially for parts like this, thinking  about print orientation and its effects  

on strength, surface quality, and  necessary supports is essential.

Printing it flat is probably the worst choice  — not only does it require the most supports,  

ruining dimensions and surface  quality, but even though it’s now  

strong in one direction, it could split 90° to it. Printing it on one of its ends seems ideal because  

it doesn’t need supports, but now the layer  boundaries are in the worst possible position —  

perpendicular to the load and right at the stress  concentration from the sander and vacuum hose.

So I printed it at a 45° angle. Yes, that  requires supports and a longer print time,  

but at this angle I’m not only getting roughly 30%  more layer strength — the layers themselves are no  

longer aligned with the stress concentration  points on the part. And printed like this,  

teh adapter did a great job through  the whole studio renovation last year.

Thinking about printing orientation can transform  a design from merely “nice to look at” into  

something truly functional. And sometimes the  most obvious orientation might not be the best! 

If you download a model from someone else,  think about what you want to achieve and align  

your part accordingly — because the default  position in the slicer isn’t always ideal.

If you design a part from scratch, start  thinking about orientation as soon as you  

draw your first sketch — because it can literally  make or break your part. How is the part loaded?  

Can you adapt the geometry to avoid supports?  Which surfaces need to be precise or clean? 

And please - once you share your model, export  the STL again in the proper orientation so others  

can print it as intended - because that’s not  always obvious and can lead to frustration!

But I also want to be completely transparent  here. Sometimes, changing the orientation can  

make things worse, and this already starts  with the quality of the corner bracket,  

where at this angle, the overhangs can curl up a  little. You can also shift problems. Initially,  

my vacuum hose adapter always broke during  use, but when I first printed it at an angle,  

it split at 45° during installation because  it was too tight. After re-printing it  

with the right dimensions, it became super  durable! Similarly, my test hook printed at an  

angle did not hold more load than the one printed  standing, and the point of failure just changed,  

because the loading itself is complex.  The interaction with infill, perimeters,  

and top and bottom layers can cause  issues. Angling can help, but it’s  

not always the magic bullet. So sometimes,  you need to think a bit outside the box,  

and this is where I need to highlight some parts  of the Multiboard organization system. This part  

clips into a baseboard to hold a bin, and at the  obvious printing orientation, the hooks are weak.  

Printing it at 45 or 90° would likely improve  strength but would require supports. Instead,  

they split the part in the middle and  connect the halves with a thin bridge.  

When you finish printing, you fold it, resulting  in a super-strong and efficiently printed part.  

And there are plenty of parts where they do this  and it’s really worth looking into. Kudos on that!

3D prints can be complex, and for  most parts, there’s no single best  

printing orientation — only compromises  between strength, quality, accuracy,  

and supports. And these factors often compete  with each other. But if you keep just some of  

these findings in mind, your prints will not  only look better but also perform better!

But what are your thoughts on this  topic? How often do you change the  

orientation of a part to make it print  better? And what’s your best example  

of making a stronger or cleaner part just by  reorienting it? Let me know in the comments!