Attic staircase

The Idea and the Plan

After last year’s project insulating the roof of our house, we were left with a stack of about 20 wood pieces of varying lengths. Tossing them away would be a pity, so I started thinking about a possible way to do something useful with them:

Most of the pieces were about 60 cm in length, so a plan started to form to use them to make steps for a staircase. We were still using a ladder to climb onto the attic, but the longer-term plan was to turn the attic into a more homely space. When we insulated the roof, we made sure to install two roof windows, so that the attic area would be well-lit with natural light. That part of the plan was already taken care of. Another part of the plan was to have a fixed staircase for easy access. Staircases are pretty expensive to buy, so if I could reuse leftover wood to make a low-cost but reasonably nice staircase, we would be catching two flies in one swat!

So I started to think about what we would need, and what we would need to do:

• For the steps: the wood pieces would have to be cut to length, and planed to the right thickness. As it happens, my father has just the right tool for this job in his shed for cutting and planing. I have a routing tool and bit for rounding off the step nosings, so making the steps will be no problem.
• For the carriage: we need some additional wood to mount the steps. I have some lengths of 89 x 38 mm SLS construction wood in storage; it’s not the nicest type of wood, but for an attic staircase, it seems like two of those could do the job after thoroughly sanding them clean. Using the same routing tool mentioned above, we can also cut the grooves for mounting the steps. This part also seems to be taken care of.
• Cutting the grooves in carriage might prove not so easy. All steps have to be at the same inclination and distance. The best (and fastest) way to do that will be to use a mold. So we will have to make that before we can start working on the carriage. We have scrap wood lying around, so that shouldn’t present a problem.
• For safety, we should also have a handrail. That’s the weakest part of the plan right now :-)

Finally, some measurements and calculation needed to be done to determine the right number of steps and the desired inclination of the stairs. Looking around, it seems that regular staircases have step heights from 17 cm to about 20 cm. The height of the staircase is to be 264 cm, so I arrived at 14 steps, using 19 cm step height. (Actually the calculated step height is 18,86 cm, but for ease of working we round it off to 19 cm. The excess height is max. 2 cm; if that is a problem, we can take it from the bottom of the staircase.) All in all, the plan is as follows:

Making the steps

The steps are taken care of by planing about 3,5mm of thickness from the top and bottom combined, so that the final thickness is about 21,5mm. The front and rear are trimmed by almost 5mm, so that the width becomes about 12cm.

Since the wood is construction wood, it is not of very high grade: it has lots of knots, it is easily dented, and it is also pretty “rough”: when you plane it, quite often bigger chunks jump off that make it hard to get a nice, clean surface. Nevertheless, out of 20 pieces of wood, we should be able to get 14 pieces that are nice enough. You can see the result in the picture on the right:

Making the individual steps was an easy task, given the right machinery. Next up is the carriage beams, which requires an entirely new level of precision compared to what I’m used to. Routing a groove that is nice and straight requires at least a guide to slide the machine against, but in this case that will not be enough: all the steps have to be at the same angle, and have the same step height. Also, the left and right carriage beams have to be mirror images of each other. We will have to maximize our precision by using a mold; constructing the mold first will allow us to solve most of the practical challenges ahead of time, before starting to work on the actual wood. This is what the final mold looks like:

On the underside, two pieces of wood are placed so that the carriage beam fits nicely between them. On the top, two pieces of wood are placed at opposing angles, to use for guiding the routing machine.

An additional bonus is that the mold will allow us to work faster: we don’t have to work out angles and distances for each step, we just have to reposition the mold at the right distance of the previous step. As an example, the following picture shows the routing result using the mold:

The mold fits precisely over the carriage beam (or in this case, a piece of scrap wood of the same dimensions.)	The routing tool positioned on the mold.

Groove made by the routing machine.	The trial pieces look quite decent.

The mold needed some slight adjustment to fit the thickness of planed steps. The steps ended up at approx. 21.5mm thickness, whereas the routing bit is 12.7mm wide. This means we need two passes; the first one using the guides on the mold, and the second one using an additional 9mm thick strip of wood to offset the routing machine, so that the groove will be just wide enough to accept the steps. The intermediate result after the adjustments is this:

This looks quite OK, actually.

Two more things to sort out before actually making the carriage beams:

• All the wood needs to be sanded.
• We still have to figure out the best way to get precise step height. As is, the mold doesn’t help us yet with that.

Making the carriage beams

For the step height, my initial intention was to add a ridge on the mold that would fit in the “previous” step. This would work to some extent: all grooves on one carriage beam would be spaced perfectly well. However, the issue is that it would be very hard to have the exact same spacing on the other beam. And because measuring from the previous step each time is “relative”, any errors would compound when going further: an error of 0.5mm in the step height would, after 14 steps, accumulate to a difference of 7mm, which would be absolutely unacceptable.

In the end, I took the old-fashioned route of picking one point, and measuring all steps from there. Even if there would be a measuring error of say 0.5mm to 1mm for each individual step, the errors would not add up so they stay within an acceptable range.

(Living room working space. Only for non-dusty wood and non-sharp tools!)

After sanding the beams, it was time to actually start making the grooves for the steps. Using the mold and the markings on the beams that were prepared beforehand, this was a fairly straightforward task. Nevertheless it was time-consuming: every groove needed 10 passes: 2 times 5 passes, where the 5 passes each increased the depth by 3mm, for a total of 15mm, and then again to increase the groove width from 12.7cm to 21.5cm. All in all, it took about 2h30mins from start to finish, to make the 28 grooves:

Now that we had the two big items (beams and steps) in roughly final shape, it was time to test if it all fit together:

There is work left to be done: planing the “visible” side of the beams, since they are pretty rough; sanding everything until the finishing is fully done; and cutting the beams at the top and the bottom at the correct angles. But all in all it looks like it will work out fine.

The time for definitively putting together the staircase is coming closer. After more sanding of the steps, I could make a selection of which steps to use, in which order, and which orientation they would be. Criteria for the selection were: smallest number of knots, smallest number of scratches left after planing and sanding, and smallest number of chips that broke off during planing and sanding. The final selection was marked as shown on the right.

Another practical problem presents itself: how to accurately determine the horizontal position of the steps relative to the carriage beams? Again, the best and fastest way to solve this, is to use a mold. I decided to use the pieces I made to test the big mold, and put a piece of MDF on them at the position where the steps would be. This would allow me to clamp the “positioner mold” to the beam, and the slide the step in until it cannot go any further. This way, again, we can put all steps at exactly the same position without having to measure every time.

Putting it all together

The final step is to put everything together and mount it in place. This proved to be easier said than done. I wanted to use glue to put the staircase together definitively, so it had to be done before the glue would dry. My initial plan was to use G-clamps to pull the beams together and press them onto the steps, but this turned out to be too difficult: the force required to press the steps into the beams on both sides was more that my clamps could pull, resulting in steps that were pressed in only halfway on one side, and serious blisters on my hands from turning the clamps. Probably the main cause of the difficulty was a tiny difference in inclination between the grooves in the left and right beams, so that each step had to be slightly twisted to fit, thereby requiring much more force to press it into the second groove.

Another difficulty was that at the same time, I had to align to steps so that they would all protrude the same amount. For steps that were already pressed in somewhat, this also required quite a bit of force. Using the small molds with a clamp made this somewhat easier, but it was still not easy to it at the same time as pressing in the steps.

In the end, I took the staircase outside, and used a hammer and brute force to make short work of it. This worked better, but the result is less nice than I would have liked; not all steps were pressed all the way in. Possibly also the steps were also not all exactly the same length.

In hindsight, it would have worked better if the fit of the steps would have been slightly more loose, and if I had used the hammer route right away. Also, wider clamps would have been nice; I had to use two clamps hooked into each other to be able to reach across the staircase, which caused lots of frustration, not to mention the fact that they introduced physical risks; had they slipped off of each other when turned tight, they could have done serious damage or injury due to the forces involved. Not to be repeated!

The picture on the right shows the staircase in its (more or less) final position, together with the ladder that it will replace. It still has to be fixed in place, but that will be easy compared to putting it together…

Fixing it in place

Fixing the staircase in place was not much more work. At the top, I attached a plate with a ridge to the “wall”, on which the top step could rest. For fixation to the floor, I made two small blocks that I could attach both to the staircase beams and to the floor, to prevent the staircase from slipping away.