During my design process I started with sketches on graph paper and quickly refined those into 3D renderings using SketchUp, a free version of a great 3D modeling program. Once I got to the point where I liked the general look and proportions of the tiny house I then started on the physical 1/8 scale model. I learned more than I could have imagined. By taking my ideas from a 3D drawing and building them in reality I was able to see more clearly what did and did not work. Now, I am in the process of editing my 3D computer model with what I have learned building the physical scale model. I also foresee this back and forth between conceptual and physical as a helpful tool for solving many problems before the actual build.
Here are several strengths and weakness I have found with each design method:
With 3D modeling software the accuracy can be perfect, however with a scale model that depends on your level of skill. Measuring, cutting, sanding, gluing, and fastening are just some of the steps involved to create and accurate scale model. Also, a mistake of 1/16″ in an 1/8 scale model means a 1/2″ mistake at actual size. The smaller the scale the more precise you will need to be with each step.
The ability to quickly change dimensions & textures are big advantages of 3D modeling.
Scale models are slow to build due to all the details that must come together just right. 3D modeling, on the other hand, can be relatively fast. You don’t have to worry about drying glue or even how to fasten materials in order for them to stay put. Also, the ability to make quick and painless changes and as many alternate versions as you like, is a huge advantage to the this process.
Assuming most of you have a computer and access to the internet, the 3D modeling can be done for no additional cost with the use of a free or trial version software. The materials for physical scale modeling, although inexpensive, may be difficult to find or cut to the exact scale dimensions required. I found that many elements like cardboard, mat board, short sections of 2x4s and aluminum pie crust dishes (for metal flashing) can all serve as modeling resources. Keep in mind that even the thickness of your intended plywood sheathing should be kept to scale to keep your model as accurate as possible.
I did not realize the satisfaction that holding a physical representation of my tiny house would bring. To feel the weight of the model, the smell of the wood, looking into the open doorway and seeing the sleeping loft made the project feel that much more rooted in reality. I enjoy my print outs of the computer models with the color textures and rendered perspective, but for me it does not come close to replacing the experience of building something physical.
Scale model with ‘metal roof’ and siding.
Several times I discovered areas of my scale model that did not join up with enough strength and required reinforcement. When this was the case I built the reinforcement into the design, instead of just adding more glue to the model. My thinking was this, if my scale model needed additional support then a full size structure would benefit from the same reinforcements. This would most likely occur by feel or sometime by structural failure – usually when I was moving the model. This is something that would be difficult or impossible to test in a free version of a 3D modeling program.
Depending on your skills and knowledge on each method you could tailor your design process to meet your strengths. There is no reason to attempt 3D modeling if you do not enjoy designing on a computer. Or likewise, with building a scale model. Even though I am familiar with 3D modeling, I found that building a scale model for the first time was fairly intuitive.
All this said, I strongly urge anyone who has the time and patience to try out both methods of design. The strengths of one style help to offset the weaknesses of the other. And I assure you that the learning will be exponential as you switch between the different modeling practices applying what you learn each time to the other modeling method.