Modelling a house at 1:50

Why? I wanted to explore the construction of a house via a scale model. I thought this would help me put materials and shapes into perspective and develop my understanding of objects in three dimensions.

Timescale: roughly 3 months, on and off (during the COVID-19 lockdown).

The plan:

1. Measure the dimensions of the house, manually with a tape measure.
2. Produce sketches using the data divided by 50, estimating wall thickness by calculation.
3. Decide on suitable materials
4. Begin modelling – starting with the ground floor
5. Continue until the house forms as a scale model

---

PROJECT PROGRESS & POSTS

14/4/2020

The project begins! The first step is measuring – I had to make my way methodically through each room of the house, tape measure in one hand, pencil and paper in the other, and record the dimensions of each with a ‘plan’ type sketch – ie from above. Unfortunately for me, rooms are often less square than you imagine at first glance, with protruding pillars or irregular walls. But I did a little bit each day and after a few days, I had a rough idea of the shapes and sizes I would be trying to construct.

• 

• 

26/4/2020 – choosing materials

From my dimensional data and the sketches that it produced, I calculated that there are two different wall thicknesses: one, 25cm, makes up the outer walls (plus those that used to be outer walls before the house was extended on), the other, 16cm, makes up the inner walls. The purpose of this difference is likely to be structural and for insulation.

I was looking for a material that was structurally stable, easy to cut with a scalpel, available in the required thicknesses of 5 and 3mm, and preferably neat and smooth looking. After research and testing combined with previous material experience, I decided on two materials: – 5mm foam board, which is light and comes in white sheets, and 1.5mm (of which I would use two layers) mount board, denser but also available in a bright white.

28/4/2020 – construction begins

I began with the ground floor. With the help of my scale sketches and dimensional data, I made the outer and inner walls from the relevant board. I had to get my tape measure out again to record some of the vertical measurements, such as window height, but mostly my sketches had all the info I needed. I used a clear liquid adhesive to glue some walled sections together – but not all of them, and not to the base. This was because the tolerance of my ruler means occasionally, sections may need trimming by a millimetre or so just so they slot together easily; also, the flexibility of still being able to move things around is an advantage at this stage.

• 

• 

I obtained crisp edges by lining exposed cuts with thin white or black card, depending on the colour of the window or door frame at that location.

12/5/20 – stairs (1) construction

The stairs were definitely a challenge – particularly because precision is KEY. If, say, I’m half a millimetre out with the stair height, times by fifteen stairs and suddenly the top of the stairs are almost a centimetre from where they should be! So I recorded stair height and width to a higher level of accuracy than I had with the other rooms, obtaining values of 18 and 24cm respectively, corresponding to 3.6 and 4.8mm on the model.

The most obvious way to produce a stair profile is via a basic concertina card fold. It was important (again for reasons of accuracy on such a small scale) to ensure every angle was exactly 90 degrees if the stairs are to line up where they should on each floor, so I held the scored and folded card against an accurately drawn template as a check.

A backing of thick card holds the stairs rigid and forms the ‘roof’ of the little loo under the stairs.

19/5/20 – stairs (2) construction

The other set of stairs in this house leads from the first floor to the converted loft. Although I’d had some stair-making practice by this point, unfortunately for me, these ones are windy and turn a corner – complicating the situation.

The situation is this:

1. The first four steps are normal enough and rectangular.
2. The next five curve around a central, cuboid-shaped newel, making their horizontal platforms triangular or kite shaped. The vertical panels are, as you’d expect, still rectangles.
3. The final four are again rectangular.

Obviously, there would be more to this net than a rectangle with even, scored lines (as with the other stairs). I tried to think of the sections separately, and, with some careful planning, I created the necessary shapes to form the scaled, curved stairs.

I held their shape by attaching to the corresponding 1st floor wall and the newel.

• 

• 

• 

• 

4/6/20 – the roof

How to obtain an accurate, scaled roof shape? Basically: Pythagoras and basic SOH CAH TOA trigonometry, loads of times. The roof consists of a the main large pyramid, a smaller section at the same angle, and a protruding mini-roof over the front room. I did loads of sketches and angle/length analysis to come to a net which I thought would work.

From past experience, often the first attempt at something complicated and 3D will have some slight issue or element you forgot to think about when you put it together. So, not wanting to waste any precious mount board, I practised with cheap white card to get a feel for the dimensions and shape I was trying to replicate. Surpisingly, what I did come up with wasn’t too far off the mark – just one section needed lengthening slightly.

I assembled the prototype roof with low-tack washi tape to check it was correct, then disassembled and drew around what now became my templates for the real thing from mount board.

• 

• 

• 

13/6/20 – finalising the roof section

To complete the loft floor, I still had to cut skylights (roof windows) and construct the walls that exist up there.

For the skylights, I paid attention to their vertical distance from the floor in order to place them correctly on the roof shape. I’d deliberately left some joints of the roof unglued, so I could still press each panel flat in order to cut the relevant skylight out. (Looking back, I probably could have done the positioning of the windows on the card prototype before even beginning to cut out the final roof from mount board. It would have been easier to deconstruct and saved the annoyance of cutting into a partially assembled, 3D shape. But I didn’t think of that at the time, so I lived with it.)

• 

• 

the ‘walls’ of the loft

This was a tricky section, partly because it’s quite difficult to visualise what the end result will actually look like. When you stand in the loft, there are of course ‘walls’ all around, yet some are actually just the roof already, since they are sloping. I measured the vertical part of the walls to have a height of 80cm; at this point they should meet the roof, which carries on at a diagonal.

The 80cm-height strip wraps around the perimeter of the room. To make the joint with the roof as smooth as possible, its meeting edge needed to be diagonal, so I used sandpaper to try to achieve the required shape.

• 

• 

I arranged the outer walls as required. There’s also a little loo that needs some attention to construct correctly. The door is a normal height, meaning it runs along the central spine of the room, where the ceiling is tallest. The other two walls follow diagonally along the roof in trapezium shapes, which I measured out and scaled. After a bit of thinking I produced the little cubicle to complete the ‘net’ of the loft room.

Then the moment of truth: putting the actual roof on top to see if I’d done the angles right so it slotted together neatly…. somehow, success!

• 

• 

20/6/20 – doors

I now had a correct ‘net’ of the walls of each floor. At the moment, the doors are represented by rectangular gaps in the walls. I think that’s acceptable representation, but I sort of wanted to add some actual doors to increase engagement with the model (and add to the cuteness of course).

In cutting out the door apertures, I’ve kept the little rectangles that were removed in a little pot, because naturally, they are the right size to form doors now because they slot right in. There are two door sizes in this house, the smaller being used for toilets. I kept the sizes apart, making my work now much easier.

What’s the best way to attach a door to its frame in a way that makes it openable, yet firm enough to be vaguely robust? I don’t want to buy any expensive new materials or adhesive, and also want to stick to those which are biodegradable/recyclable. I thought about it for a long time, but in the end, I stayed simple:

The joints would be made of paper. A thin strip, scored length ways and folded, one edge stuck to the door and the other to the frame. This allows movement of the door to open and close.

I made a load of door + paper strip combinations, let them dry, then began attaching to the frames in the walls.

• 

•