Blog

I recently had the opportunity to build a bed from scratch, which was fun and I learnt a lot – this is roughly how it went.

The Requirements

Build a bed tall enough to store deep lidded really useful boxes under it in three months ( minus the day job, visits to Fife and other important commitments ).
So a nice broad brief and only a teensy bit of time pressure.

The Design

I did initially consider building the frame out of scaffolding for that rugged industrial look but decided that I’d have more fun building it from wood. Also aesthetically the room really isn’t a scaffolding sort of room. With that complex decision out-of-the-way I promptly decided to build the frame out of oak using new untreated railway sleeper for the bed posts. It was at this stage that I realised that I could actually make the bed king sized and thus actually long enough for me, this was a great revelation. The rest of the “formal” design process involved creating two very rough 3D models in Daz3d. This was partly to work out what timber I actually needed, compare two options and to get feedback on the design. They were not very complex models (though all of the timbers are to scale, just not well put together or touching the floor).

Really the only difference was which way round the bedposts went, but given the size of the bedposts that would make a significant difference to the width of the bed, because of this I sadly couldn’t go for the wing back design if I wanted a king size bed. With that problem solved a few other design choices were quickly made:

1. There would be niches in the head-board posts to hold a mug/glass or a mobile phone, with routing for charger cables.
2. There would be a USB hub in the back of the head-board for the routed cables to plug-in.
3. If possible it would have built-in reading lamps
4. It was going to be built from solid oak
5. I wasn’t going to use varnish, screws or nails.

At this point I hadn’t actually paid that much attention to the time frame I had to build it in (82 days).

Having worked out a design and some criteria I thought I should find out the actual dimensions of a a king size mattress and start thinking about what timber I needed. Whilst looking at timber costs I realised that it would be useful to know how much the frame was likely to sag so that I could make some sort of informed decision. Thankfully the internet came to my rescue here with the amazingly useful Sagulator. As I was building a king sized double I decided to use a design load of 200kg just to be on the safe side, though that was for the uniform load I also assumed a point load of 100kg. Whilst this might be a slight case of over engineering I wanted this thing to last. After playing about with various options for the wood based on meeting my load criteria, cost and if I had central supports of not (central supports rapidly become a certainty) I decided to allow myself three exceptions to the all oak rule:

1. The headboard could be ply as it was getting covered in damask anyway
2. The dowels for the joints would be a variety of timbers to make them more of a feature
3. No way was I making the slats they were coming from the home of flat pack ikea

With that all settled and actual notes taken (sort of), it was time to order the timber (79 days). Finding suitable dowels was fun but G&S specialist timber came to the rescue as well as providing wood oil and wax, the rest of the timber was from Alsford Timber as they’re local and I’ve used them before.

The build

The materials fortunately arrived fairly quickly and was soon stacked along with one of the sets of slats ready to start construction ( 69 days ). At this point I realised that I was definitely going to need a circular saw and a belt sander as hand tools weren’t going to suffice.

It was also around this point that I realised that railway sleepers are not very straight and have a lot of cracks in them – something I hadn’t previously given much thought to. Also they’re really quite heavy – which made me quite grateful that I had a pair of safety boots I decided I should probably actually wear them for a change. This after all could be dangerous unlike using a chainsaw to cut up logs. After some experimentation involving placing bits of timber between chairs and sitting on them gently, and then repeating that with the made up set of slats I had a plan. Not a written down plan obviously but enough of a mental plan to get started, I also now had a circular saw. It turned out that the circular saw wasn’t actually large enough to cut through a railway sleeper in one go, which would make a few things more interesting later.

The first step was to cut the sleeps to size as everything was going to be built around them, as sleepers aren’t exactly straight they had to be measured from the floor to the same point above the floor rather than just running a tape along the wood. this is pretty much how all of the cuts had to be measured, so that the various timbers were level even if not jointed at the same point on each sleeper. Once the bedposts were cut I did a dry lay out using clamps to just check the plan and lengths of timbers, as the notional cut lengths and actual cut lengths aren’t the same.

I repeated this process through out the build after each timber was cut and fitted to make sure that things stayed straight and to try to spot problems early on. As I wasn’t trying for a very sophisticated finish I just used half-lap and dado joints through out, with everything held in place by 6mm dowels encouraged through 5mm holes. I used a circular saw to do the heavy work for cutting out the joints, then hand finished with a chisel as I don’t have a router, which worked pretty well and didn’t take as long as I’d feared. My other concession to using all oak (apart from the bed slats ) was the head-board which I used a sheet of ply for, as I was going to cover it anyway. Thankfully I only mucked up one joint cutting the grove on the wrong side of the head post, but the initial fit was promising.

That was the easy part done, cutting out the niches for in the head posts was quite a bit more work, not having a router ( at least not the right sort ), I again used the circular saw to cut out the bulk of what was needed, then chiselled out the rest. Finishing off with a chisel as a scraper to smooth out the back. Not the smoothest of finishes, but in keeping with the rustic look and splinter free. A hole was drilled in the back corner using a brace and bit, to fun USB cables through later.

Having got all of the joints cut, and niches carved it was time to start sanding. There was an awful lot of sanding to do. Raw railway sleepers do not have the best finish, so took an awful lot of sanding. On the bright side this was a perfect excuse to get a small belt sander, then it was just a matter of working up through the grits until the finish was acceptable. During the sanding I discovered that my workmate knock off wasn’t actually strong enough to have 50 plus kilos of timber dumped in the middle of it, repairs were needed. But after sanding the bed posts really did look quite a bit better.

To deal with the crack in the sleepers I mixed some blue glow in the dark powder with clear casting resin and poured it ( mainly ) into the cracks. Meanwhile I attached the vertical part of the side supports to the base using PVA and wooden dowels. I excused myself from doing this with a brace and bit as there were an awful lot of holes to drill. I used a pencil sharpener to slightly taper the dowels to make them easier to fit, and clamped them pieces together to stop them slipping whilst the dowels were put in. Once the glue was set on the sides, they could be sanded to make the head of the dowels flush. When the resin had cured on the bed posts it was time for more sanding to make the finish on the resin smooth. Another final test assembly and time to start treating the wood, as I had decided against varnish I went with an initial treatment of hard oil followed by wax. Given how well it brought out the grain I was really rather happy with this decision.

Once the head-board had been covered and test fitted, with subsequent tweaking of the grooves it went into.

the last of the timber could be oiled an waxed ready to be transported to their final destination. As the bed was to be held together by dowels it wasn’t possible to do more than dry fit testing until it was in situ, when the interesting problem of how to manage the final assembly could be addressed. All of the parts did arrive safely and looking a lot more convincing than when they had arrived as raw materials.

The main problem with the final assembly being that 5mm auger bits aren’t very robust and have an annoying habit of bending, so once more we had to revert to use of an electric drill.

Fortunately it did all go together with only the expected amount of encouragement from ratchet straps and tapping sticks and so far it’s not creaking too much and hasn’t fallen apart. I suspect if it does ever need to be moved then that will be quite a challenge.

I have however had to change the lights as those ones glowed too much. The glowing resin works really well but it a subtle glow to I haven’t been able to actually photograph it yet. There are a few things I’d do differently next time, mainly use deeper timbers for the side runners to give a nicer finish, and if the room has space make the wing back version so that the niches can point in towards the bed. Also obviously some of the joints could be neater and tighter, but the project came in pretty much on spec and on schedule, fortunately there was no defined budget.