DIY Staircases

[N. Thornton]

DIY Staircases

--------------

Straight staircases are really not that hard to make yourself, for under £100 using good materials.

Clear pine (redwood) is a good quality low cost wood, though hardwoods are better. Spruce (whitewood) is best avoided for this, as there is little to restrain warping in this design, and spruce warps badly. Do allow woods to dry out indoors first to avoid warping after assembly.

It helps a lot if you dispense with a lot of moulded fiddly details, turned spindles etc, and just go for flat square heavy solid looking woodwork.

A description of one follows:

2x 45 degree main beams are planed 2x5

triangles of wood attached to those to make flat tops to support the treads 1.25x(?8" or so)

Note when fixing the triangles to support the treads, it is wise to employ redudancy, ie enough fixings so that if a fixing fails the staircase remains safe.

banisters are planed 2x2 with 2mm planed off the corners at 45 degs. Spacing 3 and 3/4" so that a 4" sphere can not fit through.

Despite being very simple, it also looks very good. And of course if its your own house, an ideal opportunity to add something special in there. (Perhaps some fancy latticework?)

NT

re: house layout vs. return on investment ??

[Mike]

Hate to be a harbinger of doom but if your BCO asks for strength calculations it becomes a serious headache. Ours said he would so we decided to buy one from a company that will do calcs if needed. Of course then he said he didn't need them from them.

[Ian Stirling]

(in general) is it acceptible to test in place of calculation?

For example, "Staircase passes 600Kg load test in center of span to 100 cycles"...

[Rick Dipper]

Nightmare, so its back to cheep junk, for the BCO, and then rip it out, for doing a good looking qualtiy DIY job when he leaves. Most of us can not afford a hand made hardwood staircase, unless we make it ourselves ...........

Rick

[Mike]

Not even cheap. And certainly nowhere near as strong as I imagine most of us would build it.

[N. Thornton]

Why is this a headache? Sagulator makes calculating this easy, does it not?

a single piece of 2x5 pine, 110" long, 450kg will produce 0.5" deflection. The stair uses 2 pieces of 2x5, so real world deflcetion for 450kg closer to 0.25".

Apply a 4 ton load to the stair, evenly distributed, and total deflection now 2.25". At 100kg a piece, 4 tons would represents 40 people on the stair. And thats 40x 220lb people! I reckon that should do. (They would have to be evenly distributed to even think of getting that many on there.)

For the individual treads, lets say 8" x 36" x 1.25", with 2 people standing on a tread: deflection 0.13". All these deflections are well within breaking point of softwood. Am I missing something?

NT

[Rick Dipper]

Cool, thanks.

[Andy Burns]

[Mike]

However I think you're underestimating the deviousness of most BCOs. They aren't interested in the calculations - in fact they don't understand them if presented with them - but want an easy life by stopping all these pesky DIYers doing anything.

The myths surrounding Part P, which after all is no more onerous than Parts A through whatever, are probably spread by them.

[John Rumm]

Hmm, not wanting to piss on that particular firework, but having just done a side by side calculation against Superbeam, the web site seems to get the wrong answer.

I did a calculation on a "shelf" (i.e. a joist) made from "Pine, Spruce", with a distributed total load of 240kg over its 4m length (i.e.

0.6kN/m) and a size of 200x50mm. The web site calculates just under 7mm deflection. Superbeam gets approaching double that.

(all the sums done on the web site seem to be embedded in a few pages of (obfuscated) javascript. If one CBA, one could extract the logic and see what it is doing).

[N. Thornton]

I'm going to pull a microsoft on you...

1. Since IRL the deflections we're using are small anyway, it doesnt really matter if it the answer is out by a factor of 2, all will still work ok.
2. If people dont know it gives the wrong answer, who cares anyway :)

NT