Making Tactile Models

A model that is an accurate representation of a fundamentally 3D shape can avoid the type of problem shown in Figures 12-1 and 12-2. In these figures, we printed an actual cube and a classic perspective drawing of a cube.

Making a raised-line drawing of existing book figures is a common way of making a textbook accessible. As you can see in Figure 12-2, though, for a geometrical object such as a cube, a 3D representation is far more useful. This is the same pair of objects as in Figure 12-1, but as you can see, the perspective print is pretty counterintuitive viewed from the wrong angle. A blind student (who does not routinely see the world as a 2D projection with perspective the way a sighted person does) may find it misleading at best. As a blind friend of ours noted, though, it might be useful if you are specifically teaching a blind person about perspective.

A lot of subtle issues are involved in developing good tactile models. Do you want Braille labels, and if so how do you make it clear that something is a label and not a feature? What is a comfortable scale for a model, and how much detail should you keep? What is the minimum feature size someone can feel, and how does that relate to the best your particular printer can produce? If you are collecting a database of validated models, what are good criteria for selection? Can models really stand alone, or do they need extensive background materials to be “classroom worthy”?

We and others are involved in trying to create some guidelines. Standards get unwieldy if they are too detailed, but can leave too much to the imagination if they are not detailed enough. Various groups around the world are attacking these challenging issues.

One thing we have discovered in the course of thinking about blind student users is that models that are useful for them tend to be useful for students generally. For example, learning about the relative volumes of different simple geometrical shapes is challenging for blind students, but really, that is true for everyone. In architecture, this philosophy is known as universal design, and the general principles apply here.

Case Study: Constant-Volume Models

Rich designed a set of hollow geometrical shape models (Figure 12-3) that all have the same internal volume. The intent is that you can pour water from one to another and prove to yourself that these geometries indeed hold the same amount of water. He put them out on Youmagine as an open source, freely available set of models (available at www.youmagine.com/designs/fixed-volume-objects ). The little balls on top of the taller models, added at the request of a teacher, make the models less pointy for blind children’s use. This modification also has the side effect of making them easier to print, because a point is difficult to print reliably. (This is a good example of how easy it is in 3D printing to modify designs on the fly.)

We have been surprised by how popular these models have been. We have found that almost no adults remember that a cone has the same volume as a cylinder one-third as tall with the same-diameter base; this also applies pyramids and prisms with any number of sides, so long as the base area remains constant. The models are very good for impromptu geometry refreshers—and for letting kids who just learned about that to feel very superior to Mom or Dad. Our takeaways from this experience included the following:

• Simple is good.

• We decided Braille labels would be fussy and distracting and did not use any.

• Attention to detail and use cases are important. (As mentioned, students may hurt themselves on a pointy model, especially blind students, who cannot see where the point is before they touch it and may reach for it too quickly).

• The “storytelling” of the model is critical and may evolve as you use it in teaching. For example, it works a lot better if you have three of the short models to show that the cone is three times the height of the cylinder.

• Models should be conducive to “experiments”—in this case, stacking the models, comparing their bases (the base areas are the same), filling them with water or sand to test out the volume relationships, and so on.

• Models need to be technically correct. In this case, the internal volume had to be the same, which required some care and knowledge of how the 3D printer was going to lay up material. CAD programs design outer surfaces more easily than internal ones. In other words, it helps to have intimate knowledge of how a 3D printer will produce the model.

• Models should print easily on the poorly tuned and somewhat abused 3D printers found in schools, and not require too much teacher fiddling with settings.

3D Printed Braille

One final challenge is deciding whether and how to add Braille to a print. Braille has a standard size and dot shape. Unfortunately, these dots are small enough to be dicey to print. Figure 12-4 shows Braille printed in three orientations.

One might naturally try to print Braille on the top surface of a print, but as you can see in Figure 12-4, the print layering will result in rough dots that are uncomfortable to read, and that likely have stringing between them. Also, as we discuss in Chapter 9, prints are stronger within a layer than between them. Braille printed on a top surface of a print is prone to being rubbed off. Our conclusion after some experimentation is that it is best to print Braille on the vertical sides of a print, to get maximum resolution and strength.

TVI Challenges

In the end, the biggest challenge for extensive use of 3D printing for teaching these students has been that teachers of the visually impaired (TVIs) are often itinerant, travelling from school to school to meet with their students. They have very little time to learn to use a printer, and are hampered in printing things themselves because they are often at different schools on subsequent days.

Designing models is harder than it looks, too; there is more on this in Chapter 6. We created a Google Group, called 3D Printed Education Models where TVIs can post requests for models they would like to have, at https://groups.google.com/forum/#!forum/3dp_edu_models . The idea is that teachers at schools with makerspaces who could not think of projects could fulfill these requests and put the results up as open source, as we did with the constant-volume models. As of this writing, we have many more requests than attempts to fulfill them, but we hope this will be a future inspiration for students looking for a good project.

There have been various other attempts at similar matchups, including a hackathon by Benetech and community calls by others, but progress remains slow. We hope that our books of 3D printed science and math projects are making a dent in this void. We would like to acknowledge TVIs Lori Schindler, Yue-Ting Siu, Michael Cheverie, Jim Allan, and Sue O’Brien for technical suggestions and inspiration, and Lindsay Yazzolino for model design feedback.