After an extended hiatus, I’m back to blogging.
I’m just settling into my postdoc in Digital History and Pedagogy with the Department of History at the University of Delaware. Over the summer, I corresponded briefly with one of my new colleagues, Barry Joyce, who inquired about the feasibility of building 3-D models from topographical maps so he could use them as a teaching aid in his lectures. A very brief google search led me to this Youtube tutorial by “Shapespear,” who also wrote a series of blog posts outlining the process. Shapespear certainly isn’t the only one offering free tutorials on the subject, another one can be found here in one single pdf. I won’t repeat the content of the tutorials, but the basic workflow is as follows:
- Dowload DEM (Digital Elevation Model) data – black and white images created from satellite images by the NASA and the US Geological Survey to model topography.
- Import the DEM to QGIS, crop to desired area and save as a .tiff file
- Use 3DEM to convert the .tiff file into a .dem file
- Use AccuTrans 3D to turn the .dem file into a 3-D model and save as a .stl file
- Use Blender to finalize the 3-D model into something a 3-D printer can spit out.
I ended up skipping steps 3 and 4 in the workflow because history department’s fab lab at Western is being relocated, so I did not have access to a 3-D printer. But I set out to build a model anyway because 3-D printers have a few limitations. Most importantly, a 3-D printer can only print an object as large as its printing tray. For most commercial printers, that probably means that a model will only cover an area of about one square foot. While they show an amazing amount of fine detail, Shapespear’s models only cover an area of a few square inches. If the object is to create a large model that can be used in a lecture theatre, I thought I would explore an option that can create models of a larger size. Moreover, not everyone has access to a 3-D printer so I thought I would explore a cheap alternative. I remembered Josh McFadyen’s 2013 digital history seminar mentioned the Autodesk 123D suite of apps that made templates for creating models out of cardboard, which could be scaled-up or -down, as desired. Once I decided this would be a cardboard model, it did not seem necessary to create a highly-refined model because most of that detail would be lost through the coarse medium of corrugated cardboard.
So I got to work.
Drawing back on my previous ruminations on topography and operational history, I figured that Vimy Ridge was as good as any spot to model. For teachers or students who do not have the luxury of travelling to visit battlefields, 3D models can be an inexpensive way to get a sense the ground.
I used http://earthexplorer.usgs.gov/ to acquire the DEM data.
Then I imported it as a raster layer into QGIS and used the extractor tool to crop out the section of the DEM data that covered the area of Vimy Ridge and exported it as an image file.
I imported the image file into Blender, an open-access 3-D modelling editor, to turn the 2-D image into a 3-D model. Blender has a function that reads DEM images to create 3-D models from black and white images by raising lighter areas and lowering darker areas.
The result definitely looked like a ridge. The model was a bit rough, because I skipped steps 3 and 4 of the workflow, but I figured the detail would be lost anyway because I was making this out of cardboard. I should point out that step 4 (using AccuTrans3D to incorporate finer details into the model) is necessary for any model that involves a coastline. Raw DEM data does not account for water features, so lakes, seas, oceans only appear as featureless depressions. I exported the model from Blendr as a .stl file and imported it into Autodesk’s 123D Make app.
123D Make can turn a 3-D model into 2-D shapes that can be traced onto corrugated cardboard (or any other material) and assembled into a 3-D object. The 3-D model of Vimy Ridge was reduced into a horizontal planes that could be cut out, stacked, and glued into place.
The resulting model looked a bit rough, but 123D Make allows users to switch from horizontal to vertical planes. Building a model out of vertical planes would require many more sections of cardboard, but produce a much smoother surface.
Once I set all of the other parameters, such as the desired length and width of the model, 123D Make produced a pdf of with the outline of about 70 shapes that I would need to trace and cut out of corrugated cardboard. This is where things got a little tedious.
I took the better part of an afternoon to trace and cut out each section of the model with an X-acto knife. Each shape is numbered so that they can be assembled in the correct order to produce an accurate model. The result produced more or less what I expected. It’s not much to look at, I suppose, but a layer of papier–mâché would make the top surface more presentable.
I kept the model quite small (about 8″ x 9″) because I was going to be doing all of the tracing and cutting myself and there are limits to my enthusiasm. 123D Make is designed to allow users to scale their models up or down, depending on their capabilities. The default setting assumes that models will be printed on 8.5″ x 11″ paper and built out of corrugated cardboard, but these parameters can be increased if, say, the model were to be printed onto longer pieces of paper, traced on 1/2″ plywood, and cut out with a jig saw to produce a much bigger model.
My model of Vimy Ridge is much less impressive than Shapespear’s smaller mode of Mount Everest (then again Vimy Ridge is a much less imposing physical feature than Mount Everest) but the detail lost by finding an alternative to the 3-D printer is offset by the ability to produce a model of a much larger size. A 2″ wide x 2″ long model of Mount Everest is quite impressive with its jagged peaks, but I’m not sure equally small model of Vimy Ridge would effectively convey the more subtle topography. If I were to ask students to produce their own models as part of an assignment, the question of using a 3-D printer or another material like cardboard would raise questions about the importance of size, scale, texture, and detail. Working through these choices forces students to confront questions about how we view, understand, and represent geographical spaces as we consider their impact on historical events.
It’s also worth considering the accessibility of this technology. Commercial 3-Printers are still relatively expensive and even while local maker spaces can offer cheaper access to 3-D printing, a large detailed model of a topographical map might not be affordable. With 123D Make, howerver, anyone with access to a printer, some cardboard, and an Exacto knife could assemble this model.
UDel is in the process of opening a 3D printing lab and I understand the lab will be equipped with a printer that has a sizable printing tray. I will follow up with another post if I am able to use the lab to make larger models.