Sometimes when creating a 3D print we need to make it more durable as it will be a critical component or piece that needs to do work instead of just be admired. The traditional way to achieve this has been to modify the infill settings. For our Printrbots we typically default to 15%-20% infill and increase the percentage as durability requires. This typically does the job but comes at an increased cost for each print.
Infill works by inserting a mesh pattern within a hollow area inside a 3D model as a means of providing stability and support. Essentially “filling in” areas inside the model that are hollow. Not every model available for print has geometry defined for its insides, and this is usually left as a hollow area. Without this infill, in most cases, the model would also not be printable since extruder heads would have no filament to build atop. Infill provides the lattice that subsequent layers in a print need to build upon. The Slic3r manual has a deeper dive into infill, as well as an image showing some of the various patterns that have been used. Well worth reading for more information about this.
The increased print cost comes from more filament being needed for greater percentages of infill. As each model has fairly unique topology the cost of increased infill is as varied.
As an example we have five estimates from Cura to print a model with the only setting difference being an increase in infill. The model we’ll be printing is the Thinking Trooper from Thingiverse, a model I’ve printed with great success without having to modify so if you’d like to download it and follow along you can verify what I’m about to show.
|Percent Infill||Weight (grams)||Print Time||Image|
|100%||1226 g||13 hours 37 minutes|
|80%||1207 g||12 hours|
|60%||805 g||9 hours 40 minutes|
|40%||581 g||7 hours 17 minutes|
|20%||358 g||4 hours 53 minutes|
As you can see, prints get rather costly the greater the infill.
But, what if we don’t need the stability of the grid? What if we just need enough internal structure for the print to succeed? If we shift the requirements this completely throws out a lot of previous assumptions and opens the door to alternative infill solutions.
- Boolean channels in infill grids
Since the main objective is opening up the infill to allow materials to flow within the model the simplest solution is, of course, just make holes.
- Extend support material calculations to infill
There are some great applications out there which build support structures using shapes that require less filament than others. The calculations done for this give just enough support for the print not to fail. If we now no longer need as rigid a support matrix on the inside of a print then these very same calculations can be repurposed.
The alternative methods I would like to try implementing would take the same model and lower levels of infill, but weave channels through the infill so that another material can flow freely inside the 3D print.
Method 1: Offset and modified lattices
This method would be geared toward accommodating fluids and is a modification of how infill patterns are generated in existing applications. The infill patterns currently lay lines and patterns that are meant to provide stability on their own. Grids may be great for this but are lousy for our purposes since they create segmented columns.
The simplest implementation of this is ensuring that the lattice has portions removed which facilitates liquids flowing into the model. If we cut out diamond patterns in each layer we can maintain the stability of the infill while allowing channels for a fluid (or alternate infill material) to flow through our model.
This would give us infill that went from the more traditional full rows of rectangular objects that form a grid:
To something that will still give us the grid we are looking for but contains open areas that could resemble something like this:
To give a bit of detail to this here is what each row of infill looks like in the above example:
With the diamond pattern cut from each row we have an ample channel for fluids or other materials to flow within the 3D print to add additional stability that is not possible just using extruded materials within non-commercial 3D printers. The opportunity to pour in supplementary materials also reduces the print times as we no longer are printing full rows for support, which can speed up each print by an as yet determined amount of time. We will have to do test prints to know this for certain.
Method 2: The Spiral
Let’s assume that we have a regular geometric structure.
With this it would be possible to calculate a singular spiral that travels from what will be the base of the print through to the other end of the model. We would, of course, end the spiral a few centimeters before the other end of the model so we don’t risk breaking through.
This approach gives a few variables that can be set:
- diameter of the channel – By setting the diameter of the channel we allow for ball bearings or other fillers to be poured with less concern for clogging
- radius of the spiral – By setting the radius we control how close our spiral comes to the model’s sides
- number of revolutions – By setting revolutions we increase the density of the spiral and give us more area to fill within the model
- material diameter – By setting a material diameter we can estimate the total number of materials that can enter the model
Additionally, we can use these values to create branching side channels to flow more material into the print. These channels could let more material flow to get an even greater fill.
I’d imagine that this would be close to something like this:
Method 3: Foam channels
Another way of implementing this could be to use an expanding foam that will expand into crevices and open areas within the model to provide some rigidity while also maintaining some flexibility in the final product. Similar to the spiral method this would allow access to an open cavity within the model but would not be a regular tube. Ideally we would have the support material code that runs for external supports running internally as well so that we only have the bare minimum support material within the print. This would leave a very large open cavity within the model, and plenty of room for an expanding foam to be able to grow if injected.
An expanding foam is another good way to add stability into a print without needing to increase the amount of support material. These foams tend to be quite durable and are used by some companies as a packing material for sensitive items, but there are also cans of this available in most hardware stores. Generally it is used in construction as a gap filler and that’s exactly what we have in our model. The central cavity is just a large area that needs filling.
In the coming weeks we’ll be trying out these methods and possibly some variations to see how they perform. Are they time savers? What is the actual cost in time and materials when compared with denser infill? How durable can we make something that comes from the printer?
We will try answering these questions as we continue to investigate infill in future posts.