in case you were wondering why its been taking me so long over the last year to get on with my own build, its because i’ve been elsewhere half the time, moonlighting on another job.
i’ve hand built a timber frame extension on and into an existing stone building, perched on the mountainside, with stunning views.
from some sketches, i came up with a working set of drawings, and set about building another frame. a lot was being asked of it, principally how it is going to be enclosed, so it needed to be good.
i started by making over a hundred oak pegs, that were split out of the round, then hewn and finally planed.
the foundations and plinth went in, whilst i made the frame.
this was a frame that was measured a million times, however it was also a frame that was gauged in every aspect, although the parameters of the frame were set by the parameters of the existing stone work, largely, the final dimensions and placement of members within the frame was a felt thing. you can scale off span tables for sizings of timbers to an extent, but having an eye for scale and proportion is important if it’s going to fit within it’s context, it’s frame of reference.
i cut the whole frame by hand, and it was raised by hand as well. the entire frame stood for several months with minimal pegging, and did not move at all, which tried the old adage that you should be able to remove all the pegs from a frame and it still stand, a testament to the framing joinery.
there were some entertaining moments with raising the frame as well as making it. there was a fair amount of awkwardness, no room to move or manoeuvre, and between minimal and no tolerance, everything just had to fit right.
the frame has had to have a great deal of precision and symmetry, as it’s all being glazed. and it had to achieve this with all the milling issues like unparalell timbers, large sections of wane, damaged, twisted, squiffed and out of line pieces. it’s meant using a lot of string lines and a fair amount of figuring.
the cill beams had mortices and tenons offset from the post holes for the stub tenons of the posts.
the girding beams (the supporting members than run at 1st floor height from one side to the other) had offset thru tenons into the posts (offset to avoid the joinery conflict of the girt tenons), and the girts (which ran in this instance from the back to the front posts) had tenons that ran only part way in. (again to avoid conflicting with girding beam tenons). there was even a schedule for the pegging of the girding beams and girts, as the girding beams had to be pegged before the girts went in as the girts hid the exit for the girding beam peg holes.
the braces were sized in relation to their location, as some had to be reduced to avoid conflict with later joinery of glazed sections. all the braces were hewn round.
there were a lot of studs to make and fit that relate to later glazing joinery, doors and fixed lights. and they all had an assembly schedule.
the floor joists were drop in step housed over a central girt that had to have stepped housings into the lateral girts, to avoid all the conflicts of joinery with the posts that intersected from above and below.
the roof plates, which sat on full width post top tenons, had to run back into the stone building as far as the valley beams ran, forming the outer limits of the new roof. there were no references to take the roof from since nothing of the existing roof corresponded with any other part of it, everything i did was subsequently gauged. the new roof had to be self referential, yet be able to tie into the exiting roof structures either side of it, which ran at different levels and pitches and from different heights, starting and exiting points. it meant finding common ground where none previously existed.
the tie beams had stepped housings as well as thru dovetails onto the plates. one of the tie beams, like one of the girding beams had a flaw in it typically at the area at which it intersected another member, which required cutting out the bad bits and scarfing in a new section.
the girding beam got a cogged insert, as the timber had dried out on the face to allow for a glued joint, whereas the tie beam had to have a scarfed and bolted insert as it was assembled during inclement weather.
the principal rafters with queen posts had to be raised and then lowered in order to set out and then cut the housings for the purlins before they could finally be housed, again this had to be carefully sequenced. (i don’t want to go on about how awkward it was raising the front pair just stood on the frame and minimal scaffold, but it was a long drop down the front). in order to house the second set of principal rafters with the purlins it needed everything to be tourniqued (sp) in.
housing the purlins into the valley beams was very entertaining given the complexity of the directions everything ran at. they were the start of a lot of compound mortices, all of the rafters running into the valley beams were compound mortices at one end and half lap tenons at the other. even setting out the valley beams, for cutting out their trough was complicated by the vagueness of the existing roof, where there was no defined roof line and it meant gauging it in the end, taking what appeared to be an average angle to set out the valley cut, in the end it ran right more on judgement than measurement.
the rafter material was all bowed, and had to be layed out to avoid crowning the rafters, which subsequently meant trying to pul the bow out laterally where they were fixed over the purlins without opening the half lapped joint at their intersection at the peak. all the common rafters which had seat cuts with oversailing tails, got double spiked into the plate, bolted into the purlin, and spiked thru the lap joint at the peak. the jack rafters all had step housings to run into at a compound angle where they ran into the valley beams at one end, and either met lapped at the peak or were housed into the stone back wall.
when i finished the frame, it was time too start roofing in