An easy way to use subtraction in your modelling of complex shapes.
One of the outcomes we’re often after with 3D printing is housings or mounts for existing objects. These could be components, pieces of hardware, just about anything.
There are a whole host of ways to build mounts and models around your hardware, but one I have found particularly useful is using negatives.
This method involves modelling the thing you’re mounting or integrating with, just as much as it does the case or mount itself.
Sounds like extra work for nothing, right? In many instances, it will be worth it.
Say for instance you want to create a series of mounts for an Arduino UNO? Once you have a model of the UNO, you can very easily use it to create cavities and adaptors to suit a variety of applications.
Alternatively, if you have a Raspberry Pi model, you can create mounting points, cut-outs for ports and power, and ensure all clearances are where they need to be.
While you can clearly do this by trial and error, our experience is that using negatives yields a better result, faster.
NEGATIVES AND POSITIVES
Before digital cameras became standard, film came in two different types. Positives, and negatives. Negative film was the standard for consumers, whiles magazines and commercial studios often used transparency film. The difference being that negatives are an opposite to the desired output.
A colour negative has dark areas where the print will be light (such as highlights) and light areas where the print will be dark (such as shadows). Negative film was cheap to process and was fairly robust, making it great for consumers.
A colour transparency appears true to life to the human eye. The highlights are light, and the shadows are dark, and colours are true. It was temperature sensitive (kept in the fridge) and more costly to develop, however, it made selecting images for commercial publications easier without having prints made (as well as other benefits).
While the differences between positives and negatives with regards to film don’t really apply to modelling, they highlight how different approaches to solving the same problem coexist, and can suit various applications.
3D negatives and positives in the physical world are nothing new. Indeed, just about every casting method uses negatives to manufacture products. In industrial design, clay modelling, lost-wax casting, and other various forms have been used for many decades to create positive and negative moulds to suit various production processes.
In commercial manufacturing, these moulds are referred to as “tooling”. They’re costly to create, usually many thousands of dollars even for basic injection moulds.
Fortunately, we’re not creating physical moulds or tooling, we’re just using them in our 3D space. So they cost us nothing but a little time to create.
WHEN IS A NEGATIVE NOT WORTHWHILE?
Arguably, creating a negative is really only worthwhile when you have a second use for the negative down the line, or it provides a shortcut to the modelling in some way.
If the negative is going to take more time to build than it’s going to save you using another method, then the question of suitability should be asked, of course. However, if it’s a common fitting, a complex shape, or you need a snug fit for something, then modelling around a negative may well be the appropriate solution.
CONSIDERATIONS
It’s important to consider the limitations and tolerances of 3D printing when doing this sort of modelling. Though it perhaps should be said that, considering the capabilities of 3D printing is a critical step regardless of what or how you’re modelling.
All 3D printers will have a variable degree of tolerance and expansion. This will depend on the filament you’re using, the printer you’re using, and also the printer settings.
These factors need to be included when creating negatives to ensure the cavity being created is indeed large enough. It is generally just a matter of increasing your negative by a few percent in size to cater for any space required.
If you require absolute precision down the <1mm or so, you may require some trial and error to learn the tolerances of your machine. You can then build these into the overall design to accommodate accordingly, and you’ll end up with a nice clean fit.
EXAMPLE ONE - FLUSH MOUNTS
For a variety of projects in previous DIYODE issues, we made use of a 4.3” Touchscreen LCD from 4D Systems. It’s a great full-colour panel, however, mounting for maker applications can be a challenge due to its flush-mount design.
This led us to create recessed mounts to hold the screen firm, with a perfect finish the way we wanted it. However, modelling the required recess directly is somewhat a tough challenge if you’re not an experienced modeller.
The approach was therefore to create a model of the screen itself, which could then be used as a negative to create the recesses required. The recess depth, the notches in the back panel, it all was made simpler. This was a simple process of working over the screen itself with a vernier caliper or ruler, to get the basic shapes.
The only tricky parts were getting the rounded corners to have the right radius, and ensuring the extra notch at the bottom of the screen was the correct size.
You’ll notice on the models that irrelevant details are omitted. Some features such as what is screen and what is bezel are entirely irrelevant for this purpouse. So the model is very much an “outer shell” of the actual item. It may look a little plain, however, it will still perform the same function we need from it.
Once you’re satisfied that your model is accurate, you may want to do a quick print to verify it. This is not always necessary, however, if it’s being used in a much larger print, it could save you valuable reprinting time if you have modelled your object with an error. It’s easily done, so use your judgement on whether or not a test print is useful before using it as a negative.
Next, we’re going to use our model as a negative in our larger build. In this case, it’s the desktop case for the screen, as part of a function generator project.
For this project, we had already developed the rough shape for our case, for a nice desktop mounted display. You may notice a lot of cylinders and rectangles in some of the images. These are being used for screw holes and recesses, which we’ll address another time. However, notice how we place the screen model into the face-plate.
This process gives you a 100% visual representation of how the finished build will come together. You can recess the screen just enough so it’s flush with the panel, or embed it a little further into the panel if you like. If we invert the panel, you can see how the screen protrudes through the back of the panel. This is useful for checking clearances of other hardware you’re planning to install into your case, and is very difficult to “guess” any other way.
Once you are satisfied with the position of the screen, you can switch it to being a subtraction rather than an addition (this will vary based on your software). Once you do this, the picture of your final object will start becoming clearer.
If you’re using Tinkercad as we are in the examples, grouping objects and subtractions will merge them. As you can see once we do this, the final object becomes clearly apparent.
Now all that’s left is to lay everything flat for printing and getting it printed! Once complete, you'll have a very neat case!
This method is exceptionally practical for user-facing cut-outs such as this flush-mount screen. You can "estimate" a little more when it's hidden from the user, and if something's wrong, it doesn't quite matter to the same extent.
EXAMPLE TWO - SIMPLE NEGATIVES
In this example, we’ll show you how using negatives is useful for visualisation and planning, even when the objects being used as negatives are fairly simple.
You may recall the Acrylic Lamp project we published, also in Issue 08. It used 6mm acrylic panels which we edge-lit using LED strip.
In this case, we didn’t really have to model anything at all, since they’re just rectangular shapes. However, modelling them into negatives to complete our design still assisted in expediting the development overall. With the previously mentioned caveats regarding sizing, it allowed us to visualise and plan every aspect before a single gram of filament was used.
While it’s entirely possible to do some maths and work out the spacing, we did change the LED strip a few times, as well as the acrylic thickness and spacing changing along the way purely for aesthetics.
As you can see in the two renders shown here, using negatives allows us to visualise, and even print, various options that would be difficult to estimate any other way.
In these examples, we’ve spaced our acrylic panels at 20mm intervals, and 40mm intervals. You can quickly see the difference, and if you wanted to print both options for a physical review, it’s entirely feasible.
The use of the negatives automatically creates the slots in the support struts to position our acrylic perfectly, as desired. If we had created supports or notches manually, they would be more prone to error and miscalculation.
Likewise, if we decided that 10mm acrylic would be better than 6mm acrylic, we can adjust the design rapidly to visualise it, and print it if desired, in barely a few minutes.
Where you’ll find even more benefit, however, is when you’re building something around your objects. For this example, we also experimented with another edge-lit design which suspended the acrylic from just one corner. Having our negatives with appropriate spacing and sizing for our acrylic, meant that we could rapidly test new designs visually on-screen.
We tested and tried various different ideas before printing, which would have been tough to do without the use of negatives.
Ultimately it’s about flexibility, and creating designs that are thoroughly thought through before printing. The use of negatives helps give you a complete picture before you waste filament and print time, while providing you with everything you need on screen for a great model.
WHERE TO FROM HERE?
Now it’s time to get modelling! Try creating a negative of your own objects you want to recess or mount, which would otherwise be difficult to replicate.
Fortunately, with the open source community, there are already many models for complex shapes and common items such as Raspberry Pi and Arduino hardware. So get modelling and take your 3D printing to the next level!