In this post, I write about wall profiles in 3D printing, and I will explain why they are essential to creating stable parts with minimal filament use. Also, I will show common mistakes that result in ugly prints and discuss this topic by designing a simple 3D-printed box.
- A new Box is Required
- Fixing the Issues with the Thickness and Sharp Corners
- Bulging and Bending
- Wall Profiles Increase the Stability
- Stability Improvements
- Why do Profiles Increase the Stiffness?
- Remaining Problems
- More Posts
I am no 3D printing expert, and the topics I discuss here are nothing new. Wall profiles play an important role in stabilizing sheet metal parts, vacuum-formed plastic foil packages, and even creating strong elements from paper-like materials.
They are fundamental for any design engineer and an absolute must when dealing with thin or soft materials. Yet, for many hobbyists designing their own 3D printed parts, the simplicity and usefulness of wall profiles are not immediately apparent.
In my examples, I use Fusion360, free for private use but with some limitations. All shown principles work with other parametric CAD software as well. For free software with no hidden catches, I recommend FreeCAD, a full-featured parametric CAD solution (that may be more difficult to get used to).
I like to give everyone a good understanding, so I start with the absolute base knowledge. Feel free to skip these first sections and read on the more advanced topics.
A new Box is Required
Let us assume we require ten new custom storage boxes that fit into a space of 65mm × 125mm and have a height of 60mm. We like to print them fast with minimal PETG or PLA filament on our 3D printer.
An absolute beginner would construct the box like this:
Our box looks fine but has two obvious problems when printed with the fused filament fabrication:
- We chose a wall thickness of 1.9mm for no reason, a rule-of-thumb estimate.
- The box has sharp 90º corners.
If you want to experience the problems in real life, download the box model here and print it using a shiny PLA filament.
The Thickness of Profiles in the XY and Z-Axis
In the FFF process, objects are printed layer by layer. So, there is an extreme discrepancy in the resolution between the XY-axes and the Z-axis.
To choose a good thickness for the bottom of the box, which depends on the resolution of the Z-Axis, it should align with the printed layer height. Our bottom, with its 1.9mm, does not align with the intended 0.2mm layer height. A slicer can round the value down to 1.8mm or up to 2.0mm – which is an uncertainty we do like to avoid.
Even worse is the 1.9mm thickness for the XY-axis. While we have great precision in the movement of the printer head, the size of the nozzle opening greatly influences the print quality. If the printed geometries are thicker than the inner and outer perimeters and therefore are printed with infill, the exact width of a profile does not matter. Yet, if you print thin profiles, where inner and outer perimeter lines solely define their thickness, the width has to be carefully chosen.
The 1.9mm we have chosen is a little bit too small for five filament lines but too large to print four equal lines. Therefore, in the best case (modern slicer algorithms), the slicer increases the widths of the inner perimeter lines to fill the geometry. The wall will look okay and often have the right dimension, but usually, these lines do not bond very well.
In the worst case (old slicer algorithms), the slicer squeezes a fifth line into the profile, which usually overfills the profile. The dimension of these walls will be too large, and often this squeezing gets visible artefacts on the outside.
The Issues with 90º Corners
As a rule of thumb: There is a very limited number of cases where a sharp 90º corner on an object makes sense.
If you look around and inspect the objects you use daily, most have rounded edges or chamfers. This is not just a design statement but has practical reasons. For soft materials like wood, chamfers and fillets make the edges more robust and reduce the risk of fibres being ripped out of the material. Hard materials like metal, chamfers, and roundings reduce the risk of injuries and give the objects a nice feel if you touch them.
Many production processes require fillets, like injection moulding or casting plastic and metal. Without fillets, the objects would get stuck in the forms, or the edges would rip off when removed.
For the fused filament fabrication, sharp corners with angles of 90º or steeper force the printer head to a sudden stop and then gain speed in another direction. This not only slows down the print but also causes vibrations that are visible on the printed object.
Also, filling a corner precisely with the material is challenging. Often, because of the sudden speed changes, the filament cannot be extruded with the required precision, and the resulting over- or underfill is visible outside the object.
Sharp corners in the Z-axis are luckily no issue, as the object is printed in layers. Chamfers added on edges printed in the Z direction are not added to avoid shortcomings of the printing process; more on this later.
Fixing the Issues with the Thickness and Sharp Corners
Let us fix the two issues mentioned above.
Before designing anything, we write down all key measurements as parameters. This is an important step, as it simplifies further construction and forces you to consider how to structure your design document.
I add the parameter “wall_thickness” for the profile thickness on the XY-axes and “bottom_thickness” for the thickness on the Z-axis. I use 1.67mm for the wall thickness, a great value that prints in four even filament lines when using a 0.4mm nozzle and 0.2mm layer height. You find these ideal values in the settings of PrusaSlicer, in the section “Vertical Shell Thickness”. This value also works great with 0.6mm and 0.8mm nozzles, printing three or two balanced lines for the profile.
I chose 1.6mm for the bottom thickness, rounded to the layer height of 0.2mm.
For more complex designs, I often call these two parameters “horizontal thickness” and “vertical thickness”, as these apply to many parts in the XY-axes and the Z-axis.
After preparing all the key parameters, I construct the box using them. You can see all the steps I did to construct the new version of the box in the following slide show.
Now let us load the improved model in PrusaSlicer and see if it improved the situation.
The walls of the box are now printed with four balanced filament lines.
Extending the perimeter lines, the slicer can generate a perfect fill in the corners.
But wait! We fixed two major problems, but many remain. If you like to print or inspect the improved model, you can download the model file here.
An Alternative Fix with Fillets and its Risks
Instead of doing a complete redesign, sometimes sharp corners can be easily fixed using fillets.
Yet I often see that the created fillets are not concentric. To modify the sharp corners into ring segments of even width, the outer radius has to match the inner radius, exactly adding the thickness of the wall.
Bulging and Bending
Plastics like PLA and PETG are soft materials and bend easily.
The simulation above shows the displacement if the box is made of ABS (with a softness between PLA and PETG). In the image above, a light force of 5N (~500g) is pressing against the side wall. This happens easily if you grab the box on one side.
The next simulation shows what happens if you fill water into the box. While shown displacement in the images is exaggerated, around 1.2mm at its peak, the simulated forces are weak. Press with your index finger on a scale until you read 500g (17oz); this gives you a feeling of what a light force the wall bends and bulges.
So, how to stabilise the walls?
Increasing the thickness of the walls makes them stiffer, and the box gets more stable – yet, this is not an elegant solution. Even doubling the wall thickness makes the gained stability a bad compromise.
Wall Profiles Increase the Stability
Adding profiles to sheet metal is a common practice to increase the stability and make them stiffer along an axis. The image below shows trapezoidal steel sheets that are used to cover roofs.
A New Wall Profile
The same principles can be applied to the thin plastic walls of our box. So instead of starting in the XY plane, we will focus on the vertical profile of the box.
There are two new features on the wall of our box: Two identical recesses at the upper and lower side of the wall and a rounded chamfer that creates a transition from the wall to the bottom of the box.
The recess looks more complicated as it is because the roundings soften the angle transitions. It is a similar shape you see in the image of the trapezoidal steel sheets above.
The angle of the trapezoid is 35º, and to avoid sharp edges at the transitions, I added arcs with a radius of 4mm. All the recess dimensions are defined with parameters, so they can be easily replicated and manipulated to find the best values.
If you look at a depth of the recess, the dimension 2.33mm seems odd. Instead of defining the depth of the recess, I set a parameter that defines the total width of the profile. This is from the outer face to the closest inner face of the geometry.
This total width is the key factor controlling the stiffness of the resulting profile. Deeper recesses will increase the stiffness, but you also lose storage volume in the box. Therefore, I chose a total width of 4mm. If you subtract the thickness of the walls of 1.67mm, you end up with 2.33mm for the depth.
Both recesses on the top and the bottom are identical, except for their position relative to the edge of the box.
Why Offset Lines for the Z-Axis do not Work
If you wondered why I did not create a simple offset line from the outer profile definition, look at the following drawing:
As the 3D printer works in horizontal layers, a simple offset will produce the wrong widths at angles. So instead of the optimal width of 1.67mm at the 35º angle, the wall will be printed at 2.039mm. That will create unwanted printing artefacts, especially at the transitions.
The Rounded Chamfer
At the bottom, I added a rounded chamfer. The chamfer starts at a 45º angle and ends in an arc with the same radius I used for the recesses above. As the side walls are attached to the flat plate of the bottom, the chamfer has a positive effect on the stability of the box. On the contrary, its current form makes the bottom bulge more easily – but we will address this issue later.
The chamfer is there purely for the look. It hides the transition from the bottom to the side profiles.
Extruding the Box from the Profile
After drawing the 2D sketch with the wall profile, I extruded these and built the final box. I documented the whole process in the following slides:
If we simulate the same 5N force to the same spot as the previous version of the box, we see a reduced distortion from 1.168mm down to 0.6914mm. This is a 40% reduced deformation! The previous box version required 59.71g of filament, and our new box only uses 56.12g for the print.
The simulation of filling the box with water shows improvements too. The previous version had a maximum deformation of 0.0643mm, and the new version was only 0.03395mm, an improvement of almost 50%.
If you like to print or inspect the improved model, you can download the model file here.
Why do Profiles Increase the Stiffness?
The following explanation is a simplification to illustrate the topic. Besides the described simple physics, materials have more complex properties that influence the behaviour under stress. Therefore, the following description shall provide a basic understanding any unnecessary details.
In the illustration above, two walls bend and have the same deformation. The upper wall is thin, and the distance between the outer surface and inner surface is small, while the lower wall has a profile that adds a larger distance between the inner and the outer surface.
When the wall bends, one surface is compressed, and the other is expanded. The expansion and compression are balanced because they are made of the same material. So it does not matter how strong a material can be compressed or expanded. The important information is the difference between the compressed and expanded surface.
Under each wall, you can see a comparison of one compressed section with one expanded section of the wall. The violet line shows the difference between the two sections.
To bend the walls at the same rate, the caused difference is much larger for the lower wall with the profile. We can assume that the force required to compress and expand the material in both walls is the same. Therefore, it takes much more force to deform the lower wall to the same amount as the thinner upper wall.
So, the key factor in increasing the stiffness along an axis is to increase the distance between the inner and the outer surface of a wall. This can be done by making the wall thicker, which uses more material, and adding a profile that increases the stiffness with the same amount of material.
The X and Y-axes’ profiles increase the walls’ stiffness, but it stays soft and flexible along the Z-axis. This problem is apparent when we look at a shearing force to the box. There are no structures in place that provide resistance against these movements.
We did also ignore the bottom of the box entirely. For the previous simulations, the box was flat on a surface. If we repeat the water simulation, but make the box sit on its edges, you see how easily the bottom bulges under the pressure of the water.
There is more to say about profiles at a later time, but I had to stop at this point to keep this text at a manageable size. I hope you found these pieces of information useful and that they will make your models stronger but lighter.