Jump to content
John Page

The Birth of a Workshop

Recommended Posts

My first impression is "Boy that's a lot of beam cutting and joining to do." That is, assuming each of those pairs of lines making that 9=pointed star represents a beam line. If that's true, each of those beam lines is 5 separate pieces. Each piece has two end cuts and there are 9 beams...…..that's 90 beam end cuts. May the Gods shine on you when you fire up the saw.

At least you can use field fabricated steel connectors and getting the cut perfectly fit may not be so critical because you will cover the joint with steel plate. The only trouble with that is those really acute angles. Typically this type of joinery uses bolts through the plate/beam/plate with a nut on the end. Those tight angles make a very tight space to get a torque wrench into. That's still a massive number of fabricated beam connectors.

I'm still not clear on how the roof sits over these beams. In your section drawings, it looks like the roof structure is the radial design and these beams are simply the ceiling/collar ties that hold the beam tails from spreading. If this is true, then the beams are not supporting anything other than self-weight and maybe a ceiling? Your section has a collar tie up high in the framing. You might be able to do your 9-pointed star in this location with much shorter/smaller lumber and get the visual effect you are shooting for.

InkedSection cut_LI.jpg

I had another idea for the wood shop layout.  Those small corners could easily be made useful by building a shelving set in them.

InkedWood shop_LI.jpg

One last question: Where's the bathroom?

Share this post


Link to post
Share on other sites

Josh,

The ceiling would be radial, and all the beams at the top of the walls are there to take some of the spreading. The original idea was to have them be either be 9 single, continuous beams with joinery wherever they intersect, or at most two beams joined in the middle (making each one ~17-20ft depending on the scale). The line on the top would be similar in shape to the beams on the tops of the walls, but in addition to helping with spreading, they would also be the support for the air-moving apparatus to help with circulation. It is, though, a frightening amount of timber/joinery to deal with. Introducing either metal or fabricated beams makes it less concerning as far as size to weight to strength ratio is concerned, but that will have to be made by someone more informed and certified than myself.

Using the acute corners for storage would probably be what they end up as, but visualizing it from an other-than-top-down aspect makes more sense, especially with a primary compliment of hand tools around the edges of the space.

The bathrooms though! Having the long edge being the adjoining wall makes it so much easier to figure out, because it doesn't sacrifice any of the transit space down the length of the room. Otherwise, I would probably have to put it all the way at the end away from the hot shop, and the question of the grinding room is still up in the air. For the long wall adjoining, it would be on the far right side and also house the compressor/washtub/water tank that have no other place to live.

The engineering side of the project is finally starting, and so in the next week or two I'll have a more solidified idea of what is possible, and from there narrowing down the options accordingly. Once again, thanks for the input, it's greatly appreciated!

 

John

Share this post


Link to post
Share on other sites

Been talking to some structural engineers, and the possibility of a clearspan for the first layer spreading timbers is looking less and less likely, although not fully out of the question depending on overall scale of the building. During the discussions, the topic of 3D models came up and the necessity for clarity in communications, so I learned a little bit and devised a functional drawing. I'm glad I went through the process, as it made a few key points more obvious in their need for closer examination. While it isn't meant to be taken for dimensioning in any way, the way pieces align and the problems with intersections is much clearer.

First, the clean room was lowered by another foot to accommodate the main roof beam of the woodshop intersecting with the metal shop roof.

Second, and although I knew it would be somewhat of a problem, the way the column supports interact with the top/2nd level horizontal bracing beams is a bit off. Because the two overlapping triangles formed by the two positions of the vertical beams are concentric but not equal in size, there is no excribed circle which can join all 6 points at the same time. While in theory that isn't a problem, the later geometry is. Long story short, the extension from those beams to support the 2nd level beams is not perfect, and do not align with the corners perfectly. Maybe it's a problem, maybe not, but I'll leave that one to the engineers.

Third, the grinding room/restroom segregations within the clean room look like they are still viable if I use the front bit of the clean room where it adjoins with the metal shop, annexing almost the full width of the adjoining wall save a piece wide enough (5ft?) for a hallway leading to the back. With an overall length of the annex extended to 35ft, that leaves plenty of space in the back side for equipment and workspaces. I'll make another model for that eventually, but work has been just short of actual insanity lately.

Finally, the addition of the forge chimney. The way all the roof timbers intersect would make it near perfect for a near-circular chimney in the centre, but it is also looking like it will be scary tall. More to follow on that line of thought when I can do some more drafting. The alternative is still to have it on one of the north/eastern walls, there being enough space between main roof timbers to support a rectangular chimney there.

And now, the real reason for the update, some screenshots of the 3D model so far. No walls in place for the sake of being able to see what's going on, but all the major structural elements are there.

Cheers!

John

Edited by John Page

Share this post


Link to post
Share on other sites

Had some weird problems uploading images because of my connection, but should be working now--

Wood Shop end.PNG

Top Down.PNG

Side View.PNG

Isometric Front.PNG

Share this post


Link to post
Share on other sites

The last couple of weeks have been havoc, but a small amount of progress has been made in the design world. After a few discussions to the structural loading of the building, it was necessary to clad it in walls and the roof. I also added the interior divisions of the clean room/grinding area, but am yet to come to a conclusion on where precisely to put the bathroom/air compressor. Anyway, thought I'd share the progress, slow as it may be!

 

Ceiling View.PNG

Walls and Ceiling.PNG

  • Like 1

Share this post


Link to post
Share on other sites

I've always been fond of the octagonal shapes and building design.  The old Norse building type.  I've always been fascinated by it.  Even when I'm playing in a survival game I actually build things like that.  I'd really be interesting to see the finished outcome.  Are you planning on going with a glass top on a portion of your octagon? 

Share this post


Link to post
Share on other sites

For portions, possibly, although any windows on the main shop roof will likely be vertical rather than in plane with the pitch. The reasoning there being it is better forging light to be able to have a bit of darkness, and shuttering windows that high up would be a challenge! Most of it will be shingled, and depending on the feasibility of adding windows near the peak, it might not happen. In the wood shop annex, however, I plan on having plenty of skylights for natural lighting and the far end wall being almost entirely glass. I saw a beautiful shop built by the woodworker Mike Walker that has a similar vibe going for it and I can't get over the aesthetic!

Share this post


Link to post
Share on other sites

Hello everyone! Long time without much update, but I've been slowly rebuilding the entire model using what I learned the first time around as a basis for enough precision to build the joinery into the beams. Through all the conversation on the structural engineering, I've tweaked a few things here and there, changing beam size from generic visual beams to sized timbers. Because of all the changes and the workflow differences, I'm on version 4 of the model now, and there will probably be more before the end.

Adding in all the joinery is a bit of a trial by fire. I had to abandon layers due to all the weird interactions between lines and planes, so it's a lot more cumbersome moving around the model not being able to turn on or off various sections, but it's all for the better, because I've come across a lot of small but significant issues in the process. I may adjust the ceiling heights more before the end, but those have already changed some, and the roof intersection between the two sides of the shop are the current largest concern.

Anyway, thought I'd upload some pictures of the recent model. I'm up to somewhere around 18,000 segments, and inside all the intersections all the joints are there, hidden under the main mesh, with exception of the spreading load-bearing beams on the top of the main walls. I've run into some numerical precision issues with whatever algorithm Sketchup uses, and after so many modifications, things become approximate, which the model recognizes but cannot avoid, so not all of the planes are perfect. Quite a few ~90 degree angles, which after thousands of different pieces coming together makes enough error to be appreciable...

Isometric v4.PNG

First up is the overall isometric view. Still in progress is the roof of the adjoined building. I've learned quite a bit more than I expected about different sorts of structural solutions to problems unique to timber framed joinery, and because of that the shape is constantly evolving. Still up for debate is how to treat the walls of the lower half of that adjoined building. I'd still like to have a place where I can open the top floor to ground level, and if that means having buried/stone/non-wooden walls for part of it, so be it, but where and how that interacts with the rest of the building frame and foundation is yet to be determined.

woodshop floor.PNG

Next up is a closeup of the wood shop floor. Another unique set of constraints here, both with the joinery, assembly, and load bearing properties of the floor. It will ultimately be holding a collection of small, heavy machines, and clear spans become problematic. By introducing all of the joinery inside those beams, it gives me a much better idea of how they are weakened through the removal of material due to the joinery. If I ever figure out how, I might take a few video screen captures showing what's going on inside the beams, but without motion, the scale and context is totally lost.

1st Layer Ring_stacked expanded.PNG

Last up for now is an expanded view of the layered timbers for the horizontal timber ring on top of the main walls. Due to material constraints, having whole beams is either impractical on a handling perspective or prohibitive on an acquisition perspective. So, to remedy that issue I've come up with a series of staggered, interwoven lap jointed timbers. The six beams are repeated together a total of nine times, and supported by the vertical posts at critical points and bracing where held in the centre, as loosely shown in the first picture. Again, without motion, the context of the joinery is muddied, but the general idea is still there. Each of the lap joints is keyed so lateral motion is prevented, slippage is unlikely, and the two halves are wedged in the centre to prevent movement without making them impossible to handle due to either fragility or encumbrance. This type of joint, at least to splice the beams together, is one of the strongest for the type of load they bear. Also, by having the orientation of joints between each of the nine clusters figured out now, I was able to make each of them the same, so I only have to make one set of cuts nine times instead of nine different sets of cuts one time. Once they are all cut and stacked, I'll join them in the stack with forged bands that are tightened with steel wedges, again allowing for disassembly/repositioning/repair without compromising the entire stack of beams. Also it'll add an interesting aesthetic ;)

Reading up on historic timber framing and the progression of joinery as time, culture, and technology spread, it is more common to find repeated joints for obvious reasons, and although the repeated themes of how the timbers intersect bear similar joinery, there is a wide variety of how they all come together and the sorts of loads expected of the timbers.

Anyway, that's it for now!

 

John

  • Like 1

Share this post


Link to post
Share on other sites

This is outstanding. I have been watching the timber framing videos on the Northmen YouTube channel. If I were a single man, I would sign up for the course!

Share this post


Link to post
Share on other sites

Right! I've thought about trying to get over there for one of the courses. It's the perfect storm of location, approach to the construction, and style of building.

Over the last few days, I've been doing some more research into the possibility of using hammer beam trusses to support the roof rather than tying up the floor space with vertical supports. While there's no question that they are stronger, being able to take the entire vertical load straight into the foundation, I'm concerned with how much it constrains being able to move things around inside i.e. power hammer, mill, lathe, etc. The spacing is fine, but if I can get away with removing them, I'd rather have the entire footprint open.

Also, I realized I made a critical mistake in the foundation. When adjusting a few dimensions for vertical spacing, I realized that I did not leave space on the foundation for the space needed for the stone to sit proud of the ground. Key to preventing rot and moisture intrusion, I added some additional height to the stonework, but had to adjust almost all of the vertical beams in the process.

Happy holidays everyone!

 

John

Share this post


Link to post
Share on other sites

Changes a-brewing over here, as alluded to earlier. I couldn't shake the idea of removing the vertical supports and making the entire thing a clearspan, and with all the learning that's gone on with the software, it was less intimidating to go after.

Been working away at all the internal joinery and the detail is revealing all sorts of interesting things as I consider the order of assembly. When doing some research into foundations and stonework, as aforementioned I realized that I didn't leave any space aboveground on the foundation to have the timbers sitting proud of ground level. In the process of changing all sorts of vertical dimensions, I came back to the question of whether or not I can get away with the vertical supports in the middle of the main room and took the time to start (another) new revision experimenting with various alternatives.

 

Knowing little about the relative shear strength of non-reinforced and stacked/rubble masonry, I took a gander at some interesting yet oddly specific research about the displacement of wall segments under diagonal compression and there is a surprisingly consistent curve resembling a stress v. strain diagram ultimately leading to  fracture. Anyway, the necessary thickness of the wall obviously compensates for the transverse loading, but the question is where in the distribution of force is localized loading greatest? The weight of the roof comes down seemingly as a vertical load, but determining if the corner posts are preferred for load bearing depends on if the spreading distributed evenly across the entire vertical post against the wall, or if it is more torsional and therefore concentrated most at the bottom end of that post.

 

Trying to apply similar concepts to a wall that's about 2/3 wood and the bottom 1/3 stone, if I were to take a hammer beam and lower the brace so it's closer to ground level (stone foundation rising about a foot above ground) it’s probably safer to rely on the combined strength of the foundation and bottom most wall beam to absorb the spreading force of the trusses. By the info I've been able to gather, it looks like the loading at the top of the walls is mostly vertical, in which case the need for a tension ring (or amount of spreading absorbed by it) is reduced if most of the spreading can be deferred downward.

 

 Beam Centre 1.PNG

Early draft of a modified hammer beam centred on the wall, later discarded

 

Digging into the statics of hammer beam trusses, I happened across some designs that had 55+ foot clearspans, which is enough to convince me that I’m not crazy enough to try. I came up with two configurations, one with the hammers on the corners of the wall, another with them at the centre. The biggest concern I have is whether or not it is even viable, although other designs suggest it is, and the next is what to do about the spreading and weird torsion going on in the vertical wall beams. With some tweaking, I was able to fit the hammer beams on the middle between the intersection of the horizontal ring thing, keeping it there as a means of absorbing some of the spreading load experienced on the top of the walls which are less resilient to that sort of loading. Seems like there is a mixed opinion on whether it is mostly a spreading load or torsional that hits the tops of the vertical/outermost beams of the truss,  but in either case I think the frame needs some sort of transverse stability.

 

Because of all the radial symmetry, I'd have a set of 9 hammer trusses on either the corners or centre of the walls, filling the space between with a single long and fairly steep brace from near the bottom of the wall to somewhere midway up the rafter. Early frames use cruck (naturally bent timber cut down the middle to form an A-Frame sort of brace) and the same principle makes sense here for nothing more than conservation of space. I thought about doing some sort of scissor truss contraption where it is essentially two braces at different angles to the rafter, now used in the roof of the wood shop on either end, but it is inherently not suited for odd numbers of trusses.

 

The main concern I have/design constraint of shifting around the truss system to accommodate is what to do about the tension ring and how the introduction of windows into the walls would impact the overall strength of the wall. To leave room for a secondary rafter segment at an angle parallel to the main outer roof rafter, I'd either have to shorten the size of the hammer beam and bring it all closer to the walls, or else find a way to intersect the ring of timbers connecting the tops of the walls. It's visually a lot cleaner to shift how long the timbers span (connecting the middle of every other wall rather than every third wall) but in doing that, the angle of the timbers is significantly reduced, thereby decreasing the amount of spreading load absorbed by the timbers in the first place.

 

For the window situation, I figured it would be the easiest to add an additional beam, like in the hammer beams fitted to stone walls, that sits on the inside of the wall and provides support against the spreading of the truss and allows the braces to possibly extend farther down than the foot of the wooden part of the wall. Again, not knowing too much about the distribution of force and what it's doing at the top and middle of the wall interaction, it's mostly guessing.

 

So, over the last few weeks I’ve done a lot more fiddling around with models and some more research into stress analysis. Wishing I had brought along the old college computer that had some of that sort of software running on it from the aeronautics side of things... Probably would have answered a lot of simple but very specific questions along the way if I had, but alas, it’s back to spotty research and guessing. As it is, I found an analysis around that compares the stresses in hammer beams v. modified beams that have a second diagonal member from the main lower brace up towards the horizontal one. There’s a surprising reduction in localized stresses, and I think just the ticket for what I'm trying to do.

 

 Hammer Beam Roof Frame.PNG

 

Hammer Beam Roof Frame 2.PNG

 

After trying just about every possible configuration of tension bearing assemblies on the top of the walls coupled with quasi hammer beam trusses, this is what I've come up with. I originally put the hammers on the middle of the walls for better absorption of spreading in the tension beams, and also by changing how far they spanned. But, dealing with doors and things I couldn't find a good solution that didn't involve a lot of awkward intersections. Also, it increased the angle of the tension beam terminations by almost half, thereby reducing their effectiveness. So I reverted back to the original layout for the tension ring beams and rotated the hammers to the corners of the walls. This seems better for the long term settling of the frame as well, since it is not relying on the clean stacking of wall timbers as a stress-absorbing plane. Using the compressive strength and doubling the purpose of the hammer brace as a support for the long span of the tension beams, it (looks like it) might work out.

 

 Full Frame interior.PNG

 

Based on a few other structures I've found, the odd number of trusses meeting in a single post in the centre of the peak shouldn't be a problem, and it's looking like it will be around a 18" diameter  post up there. For the other beam sizes, I went with an 8x14, although it's easy to adjust at this point. Fits well with the other faces around and gives a suspected weight:rigidity ratio that's favorable under the loads and spans (unqualified opinion/complete guess formed by loose interpretations of other structures)

 

The current configuration gives enough overhead room for everything I need it to do, and the bottoms of the hammer brace are set at 5'5" above the ground where they intersect the wall corner post. That's only 6" above the sill timbers (12" square) and with the rigidity there I'm hoping it will be plenty to keep the spreading/torsion in check.

 

Anywho, that's all for now, hope to have some of the final designs (barring dramatic changes required by actual structural analysis) done by the end of the month.

 

John

Full Frame.PNG

Share this post


Link to post
Share on other sites

I just want to say that I am jazzed enough about this project to ask what your time frame is and where you plan on building this Taj Mahal of shop buildings. I may be able to convince the corporate office (wife) to let me come help build it with you when the time comes.....

Share this post


Link to post
Share on other sites

First off, sorry for the delay, turbulent times over here! The initial time frame is roughly:

-Imminent future: Finish the CAD model, get it vetted through the P.E. I've been working with and get the ball rolling in finishing stages of the design. Right now, it's about an 80% solution of which the level of confidence in that %% is much better than the first several iterations. Learned lots, fixed details in the right places, and ultimately realized that some things were silly and others necessary, and through now 6 (?) from-scratch rebuildings of the model, think that the transition from model to certified drawings is going to be faster, less painful, and (money maker here) more directly similar to expectations than when I first started.

-Spring 2019: Find land. At this time, looking in the southern-central portion of New Hampshire, in an approximate 30-45min circle around the Concord and Manchester areas where those two distances (preferably) intersect. Primarily it's looking like north/north west of Concord, although the hunt is still on. My original assessment of Vermont was with the underlying idea that the proximity-to-civilization:land-availability ratio was more favourable, which was not, strictly speaking, the case. Besides, NH has a bit more going for it as far as small businesses are concerned, and a moderate albeit secondary slew of other reasons. And so, if/when able to make the trek out to the grand old Northeast, it's over the river and through the woods I go!

-Spring 2019, the elder: Close on a piece of land, after a thorough conversation with people who know more about whichever-county's zoning laws, ground assessments for potential building sites, etc. At this time, the final site-specific adjustments can be made to the design, such as required foundation depth/thickness/etc. and the potential of the back end of the clean room being buried in a hill to allow the wood shop to have a drive-up door (highly desired but not deal breaking). At this time as well, I will be making the determination of how much material can be sourced from the lot. The sizes and locations I've been looking at will ideally provide enough of the raw resources (timber and stone) to begin the process internally. I expect that the time needed for curing timbers cut from the property (~1-2 years depending on size) will align with the later stage members I would like to use them for. If there is a suitable excess of timber, which is possible but unlikely given the enormity of the building, then I will try and take a trip out specifically to fell a decent number of trees, stack, and leave until I can return for them later.

-Fall 2019: At this point, I will (if the wheel already set in motion gathers no moss) be at the tail end of a career transition. Meaning, I'll have about 2 months to solely devote to getting the foundation dug, stone split, and foundation walls/floor masoned. Herein winter will make it difficult to do any more work of that sort, so I would like to have it completely finished before the frosty winds pick up.

-Winter 2019-2020: If able and resources located in the prior spring/summer, begin roughing timbers. This hinges mostly on if there is a barn on the property already, as I do not intend on building one, however makeshift, unless the situation absolutely demands it. In the future, I may build an outbuilding dedicated to processing timber, but that isn't in the cards just yet. So, provided the alternative does not present itself, this time will be a bit of a break from the project, doing adult things like being employed and surviving the winter...

-Spring 2020: Here, timber processing will begin in force. If the CAD model is good enough and the correlates to the drawings with enough precision (or another model build modeling the drawings) I will have an entire cut list already done, and the assembly will be very easy, from an order-of-operations perspective, to visualize. Because of that, I can do the batch work and assembly at the same time, building as I go, and creating a stockpile of ready timbers when conditions do not allow me to keep erecting the frame (i.e. being unable to lift timbers by myself/weather/unforeseen circumstances). I expect to have a decent number of timbers ready, mostly the main structural members, by the time the weather is dry enough to start raising it.

-Summer 2020 and beyond: Building season will be in full swing and a great deal of work will hopefully happen by the time winter comes around again. Enough that, if possible, I can use the 'clean room' (covered by the floor of the woodshop) to continue working through the winter. A lot of beam-hewing goodness to be done, repetitive cuts, riving, finishing, etc. to be done which does not need the full compliment of square-footage needed for those 30+ foot timbers. All said and done, I'd like to have the thing finished in 3 years. Of that, a lot will be gradual progress intermixed with moments of sudden excitement (roof raising!). For those, the won't be likely to happen out of the blue, and I'll do my best to put word out to anyone who might want to partake in the festivities. The old-timey fair style events of communities coming together for a frame raising are exactly the sort of community involvement, interaction, and merriment that makes me want to build a shop so large in the first place. Everything I know, after all, is a product of the very community of people who I one day hope to pass on the favour back to.

 

Share this post


Link to post
Share on other sites

And now, for something a little different!

Now that the main trusses are seemingly figured out, the question turned back to how these two roofs interact. Originally, I'm not sure how I did it, but it was simple. Probably because it wouldn't have actually worked that way. Regardless, I tried a few things to get it to mesh coherently-

Straight Extended.PNG

First attempt was to just push the lines all the way out of the walls and see were they intersected. Lots of weird voids and awkward spaces to fill with extra bits of wall, strange overhangs, and unsupported corners. No good here. Joints that do not intersect approximately parallel or perpendicular to the grains of the timbers are asking for trouble, especially when carrying the load of thousands of pounds of snow.

Extended clipped corner.PNG

The obvious solution was to just clip that bit of roof and add some timbers to support the new geometry. While vaguely deliberate looking, there are still a handful of unnecessary transitions on the insides of the walls/ceiling, which is more noticeable because you see it when you're in the building instead of floating around in space.

This was several hours into the problem, and about all I could reasonably come up with at the time. Thinking on it for almost two weeks, I was convinced there was a better, simpler way that did not involve skew intersections, floating beams, and unsavory corners.

Extended clipped corner height adjustment.PNG

I'm not entirely sure how I missed it the first time, but when I thought of the obvious (sneaky as it is) solution, the hardest part was getting the alignment right without existing reference surfaces. It turns out that, because I long ago made the innocuous-at-the-time-but-critical decision to centre the overhanging roof on the middle of the next tangent wall, I was able to tweak the pitch of the adjoining roof to match the extruded angle of the common rafter out along the axis of the woodshop. Because I had the fire and forget mentality on how the pitch of the roof needed to shed snow and not much else, it never occurred to me to backtrack and see if a different pitch made more sense. It does, specifically the one single pitch that happens to match the intersection of the other roof.

It wasn't quite so simple as moving some angles, because of how lines in sketchup interact, but a few hours later it was moved around, the roof bents replaced with steeper ones, the frame lifted higher to match the point of intersection, and some other minor tweaks fixed (the corner of the roof break shifted to the centre of a post vice the edge, for instance).

Extended clipped corner height angular dimension.PNG

The other key here is that the break in the roof does not extend all the way to the roof intersection. If it did, the entire purpose of moving things around would not work. That shallower angle I thought necessary to preserve overhead clearance with the stairs leading up to the woodshop but there should be sufficient height with the alternative (speculation yet unproven). Due to the nature of the scissor truss, it is little more work to extend the outer edge of both timbers as it is to keep only one. Hard to explain, but the next picture shows it better. So, if I extruded that steeper beam down to the same breadth as the crossing beams, it provides a constant angle (adjusted to match the centre of the common rafter) the entire length of the roof plane intersection. If you've ever clad a roof, you can appreciate that simpler is in fact better. Taking out the break at the intersection removes about 4 different weird corners, twice as many beams, and an infinite amount of head scratching.

Increased pitch woodshop alignment.PNG

Now, there is a bit of extra vertical wall that has to go in between the two arms of the scissor, but it's a small and easy price to pay for the return. I haven't gotten to cladding the entire roof yet (and might not, no reason to at this point) but the plane of that intersecting bit is a single surface that also covers the top half of the scissor trussed roof. A nice, ~50 foot piece of unbroken roofing real estate.

You may notice that the roof is now significantly higher, which it is. About 5 feet higher at the peak, although only ~6" at the wall head (lifted to accommodate the line of intersecting planes). That gives a total ceiling height of just over 16' at the peak, which is lofty but considering the other half of the building is not ridiculous. Especially because now the ridge beam of the woodshop is in line straight to the peak of the main roof and can nest nicely in the mesh of other trusses coming together there.

Looking back at it, it seems fairly simple, and it was probably just looking at a problem too long that made it take so long to figure out. In any event, the worst of that is now over with, and it's on to other details. Next up, supporting joinery, bracing, and seeing if there is too much of a spreading load in the woodshop roof to keep it free of other internal timbery.

Cheers for now!

Share this post


Link to post
Share on other sites

The fact that you made it work should not be taken as proof that you are not insane. You are insane, but I love this project.

A couple things I see to think about in that last photo. The intersection of the woodshop roof ridge and the point over the main shop will never come together at the same elevation, and you don't want them to. It's a nightmare to weather proof. Try to plan it so that the ridge of the wood shop comes in at least 12 inches lower than the point over the center of the main shop.

Questions: Why is that last post between the wood shop and the main shop supported on a horizontal beam? Why is there another horizontal beam below that one?

InkedLast cad dwg_LI.jpg

Share this post


Link to post
Share on other sites

Insane, indeed! :rolleyes:

Regarding the weird intersections, the peak of the two roofs will actually meet at the same elevation. The principle rafters of the main shop roof will come together to a kingspost like beam in the middle, but I haven't gotten around to flushing out those details yet. Geometrically, the ridge of the woodshop has to meet at the same place where the extruded posts all come together. Without that, the angle of the intersecting plane is off. I think I'm going to have the vertical kingspost beam extend a little above where the points all intersect to give it that extra height. Definitely don't want a leaky roof! Might make a sort of cap to put over the top of all the intersecting timbers too, which would lie underneath all the shingling to serve as a weatherproofing around all the complex joints and angles.

For that other horizontal beam, that's just leftover from some previous experiments, although it's in plane with the top of the floor for the woodshop. That one and all the ones on the same face but 2' higher are going away. I'm working through how to have a higher stone wall for most of the parameter of the clean room so the outermost edge of the woodshop can be at ground level, and for the time being it's looking like stone will comprise the entire wall beneath the floor of the woodshop up to the vestibule for the stairs and include the perpendicular wall there to the same height. The rest of the run and the outer parallel wall (where the beam is circled) will be the same height as the rest of the foundation, leaving a normal wall's worth of height as timber. That being said, the top of the wall will align with the shop wall height with an extra bit rising between the vertical support there at the last scissor truss.

Hopefully that makes sense! I'm working through the details in the model now to try and clean up the temporary timbers and things that won't be kept in the final design.

 

John

Share this post


Link to post
Share on other sites
13 hours ago, John Page said:

Geometrically, the ridge of the woodshop has to meet at the same place where the extruded posts all come together. Without that, the angle of the intersecting plane is off.

If I am understanding you correctly, I would disagree. It really doesn't matter what angle the plane of the wood shop roof is, or where it meets the main shop roof. The roof pitches (measured in inches of rise per foot run) don't have to match either. You can make any two roof planes meet through a technique called over-framing.

Edited by Joshua States

Share this post


Link to post
Share on other sites

To have them intersect, I see what you're saying and definitely agree. The previous iterations tried to accommodate the differences with adding timbers and extra supports, but became convoluted enough that I scrapped it. The only reason I constrained the height and pitch of the woodshop roof was to have it intersect the plane of the main roof at a location where there was an existing beam (happens to be in the middle of one of the faces) for sake of clarity in the frame structure.

I added the hammer post to the middle of all the rafters and I see now what you mean about the difference in height, which I'm still trying to work through exactly what's happening. I think that if I lower the ridge beam of the woodshop by the height required to make it at a level with the extruded edge of the main shop rafters, the corner where the two roof planes meet will shift towards the inside edge of the roof panel that supports it (on the inside edge of that beam rather than in the centre where it currently sits). The top of the woodshop ridge beam is in line with the tops of the main roof principle rafters, but only on the corner where the roof plane meets the vertical face of the timber. None of that is easy to explain with words, so I'll do a few more edits to the model and show you what I mean. I think it also simplifies some of the joinery, or at least the beam cutting, to have the planes intersect at the edge of the beam instead of the middle, but it'll be easier to model than to do the math :rolleyes:

Share this post


Link to post
Share on other sites

Right-o, took a little longer to bring the screenshots over than planned, but here's what's going on with the frame and joinery. I shifted the adjoining roof down a bit so the intersection of the planes is now at the inner edge of the common rafter vice the middle. That allowed for the ridge beam to drop a few more inches but it's now at the limit of where I can move it to. Looking at the extruded face and how it interacts with the kingspost, the lines of the rafters come to a point somewhere inside the kingspost but higher than the peak of the ridge beam. However in practice the ridge beam is slightly proud of the joint between rafters and kinspost. Still working through some alternatives, but this is as close as I can get it without adding extra/unnecessary beams.

Ridge Beam intersecting joists.PNG

Inside the frame, here's what the kingspost joint looks like. Took a bit of figuring on the size and interaction between the hammer beam ends, but this is what I've settled on for a few reasons. First, having the tennon ends meet less-than-full-beam-width, it allows some of the stress to be distributed and load taken partially off of the tennons. Second, it allows for a metal (forged, of course) reinforcement of the joint that cannot be seen from the underside due to the continuous ring formed by the trusses extending out to where the faces intersect (corner before the ends are chamfered).

Kings Post Detail side.PNG

Kings Post Detail top.PNG

Last up for now is a look at the adjoining roof of the entryway. This one also gave some trouble, but reading about certain types of interior corner beams, I figured out a fairly easy and math free way of determining how to pitch and cant the angles of the intersection. Still some work needed to clean it up, but the skeleton is there.

Entryway Roof Junction.PNG

In going back to add the stairs back in, I had to adjust the floor heights again, which is an annoying but critical piece of the puzzle. Turns out building code for stair dimensions is not conducive to having a wall height divisible by 6" increments. In the end, I shifted the woodshop floor down 6", the hotwork shop up 1.25", and the grinding/clean room floor down by some other amount that gives 3 stairs at 7.75" each and a ceiling height of 9'4" or so.

For the time being, I'm trying to refine the overhanging bit of roof that connects the two shops, and from there it's the finishing details of window dimensions and a few of the joints that I neglected to add earlier, and it's about done!

 

John

Share this post


Link to post
Share on other sites
On 2/11/2019 at 10:53 PM, John Page said:

I shifted the adjoining roof down a bit so the intersection of the planes is now at the inner edge of the common rafter vice the middle.

You lost me there.

On 2/11/2019 at 10:53 PM, John Page said:

Looking at the extruded face and how it interacts with the kingspost, the lines of the rafters come to a point somewhere inside the kingspost but higher than the peak of the ridge beam.

That didn't help. Did you mean to say that the hot shop rafters meet the king post higher than the ridge of the wood shop does? 

On 2/11/2019 at 10:53 PM, John Page said:

However in practice the ridge beam is slightly proud of the joint between rafters and kinspost.

That I understood, but I still fail to understand why this is necessary. There seems to be a lot of dead space in that wood shop roof. Unless there is a pitch limitation and those upper rafters cannot go below a certain pitch, I don't get it. Why not take those lower rafters, increase their pitch, and eliminate the upper portion? It looks to me like the current configuration doesn't transfer any load differently than a straight up common roof frame with collar ties. Unless the rafter span is a lot bigger than I think it is. Even if it is, you have those interior bearing posts breaking up the wood shop rafter spans, so you should be able to overspan those rafters a bit.

That valley rafter at the entryway shouldn't be difficult. It can be a true valley rafter that runs at 45 degrees off the ridge and the two bearing walls. It's pitch is equal to the pitch of the entryway common rafters, but it rises at a rate of x inches in 17 inches of run rather than 12 inches of run.

Common and jack rafter pitches are described in inches of rise per foot of run. Hip and valley rafter pitches are described in inches of rise per 17 inches of run so that everything in the roof sheathing planes flat.

So if the entryway common rafters are in a 5:12 pitch, that little valley rafter is cut at a 5:17 pitch.

Edited by Joshua States

Share this post


Link to post
Share on other sites

Another thought about that adjoining roof between the wood shop and the hot shop. Why not make the wood shop end wall a gable end and drop the ridge of the adjoining roof a foot or two?The adjoining roof rafter pitches would get slightly less, but that shouldn't matter. It's an aesthetic thing not a structural one. (and that is dead space anyway) You could probably even make the pitches match by relocating the bottom bearing points closer together. Just slide them along the hot shop wall plate.

Edited by Joshua States

Share this post


Link to post
Share on other sites
7 hours ago, Joshua States said:

Another thought about that adjoining roof between the wood shop and the hot shop. Why not make the wood shop end wall a gable end and drop the ridge of the adjoining roof a foot or two?The adjoining roof rafter pitches would get slightly less, but that shouldn't matter. It's an aesthetic thing not a structural one. (and that is dead space anyway) You could probably even make the pitches match by relocating the bottom bearing points closer together. Just slide them along the hot shop wall plate.

This! After a few more models sent back and forth, this is almost exactly the solution I'm at now. Originally I wanted to have some sort of opening (separated by glass for containment purposes) between the woodshop and the hot shop, but after the headache of joining the two roof planes and how much wasted space there is underneath there, I cut the losses and put a gable end on each side of the wood shop and called it a day. Still a few things to work out with a cricket in the trough between the two building surfaces, but a way easier solution than trying to figure out that 20+ foot span of nothingness. Would have been nice to keep the aesthetic, but practically speaking it made a lot more sense not to. And it allowed me to also drop the wood shop ceiling by a bit. The entire thing also came down a nondescript distance because of the building code of stairs being awkward and indivisible by most combinations of 6" heights that the ceilings and whatnot are based off of, so it sits somewhat lower now as well.

At this point, the cad model is nearly done. Still working through minor details like the exact window position and some stuff with other frames and how they interact, but for the most part it's done. Just waiting on the final word from the engineering side before actual plans start getting drafted. Lots of work there, but fortunately I think my hand in it is almost done (6? months later!).

Later in March, I'll be heading out to NH for some site surveys on potential properties. I have a small handful lined up with one being more appealing than the others, but I need some closure on flood levels because it sits near a lot of tributaries and the last thing I want is a sort of 'castle sinking into the swamp' situation a la Monty Python.

Related but not yet in the model, I've also been doing some research into foundations and how to go about using a combination of the natural granite so common to NH and modern materials to build a sort of composite foundation. The idea being having reinforced concrete where it matters, and thicken the walls with the masonry for thermal efficiency and aesthetics. For anyone interested, I came across a wonderful little book called 'The Art of Stone Splitting- Early Rock Quarrying Methods in Pre-industrial New England 1630-1825' and it's almost exactly what you'd think it is. Combined with some other framing books, I'm starting to piece together a picture of how this all comes together and what needs to be different from normal construction methods due to the dramatically different load distributions.

More to come soon!

 

John

Share this post


Link to post
Share on other sites

One last thing- forgot to mention that I redesigned the hot shop rafters (again) to lighten the frame a bit. After some sensible and firm recommendations, I removed the horizontal ring that intermeshed with the hammer beams. It required reconfiguring the hammer beams a little to make them handle the loads differently, but it turns out that the spreading absorbed by the horizontal ring was considerably less than I thought, especially with the whole point of the hammer beams being to turn spreading loads at the tops of walls into torsional loads in the bottoms of the walls (there being thick masonry that can withstand the forces). I'll get some screenshots up in the next few days as well as of the gable end situation.

Share this post


Link to post
Share on other sites
14 hours ago, John Page said:

Still a few things to work out with a cricket in the trough between the two building surfaces

Piece of cake. I can draw that detail in 10 minutes for you.

14 hours ago, John Page said:

the building code of stairs being awkward and indivisible by most combinations of 6" heights

I forgot to mention this last time. What building code are you designing for? IBC section 1011.5.2 says maximum riser height is 7 inches, minimum is 4 inches. You should be able to make a set of stairs work between any two floor heights. (As long as you maintain 80 inches overhead clearance above a line connecting the tread nosings.)

On a side note, have you looked into ICF systems for your walls? One product that comes to mind (although there are dozens of similar systems) is the Hercuwall system. Poured concrete into insulated forms. High R-values and you can stick masonry veneer on the outside and plaster or GWB on the inside. Some of these are tilt-up panelized walls, others are laid just like regular masonry and then grouted.

Share this post


Link to post
Share on other sites

The stairs things is more a problem with a combination of dimensions rather than any single specific measurement. The angle of the roof for the overhead clearance, the width, and the height all combined in the space I'm trying to stick it in along with the minimum landing requirements (had to have a 90 degree bend in it) mean something has to give. Instead of having a stair stick out into the room (because of the 10" tread depth) I adjusted the step height and floor depth to make it so the last stair is more or less flush with the wall.

Looked around a bit with some wall systems and it's impressive what they've come up with since the days of simple roll infills. I'll definitely be using some sort of composite wall on the outside of the stacked timbers, then with another wood cladding over the outside. I'm at a standstill engineering wise until a building site is located. Hopefully if the cards fall right, I'll be out in NH the end of next week looking at various places. Site surveys will be a bit challenging though with all the wintery weather happening over there recently.

Modified Hammer Beam.PNG

Here's a screenshot of the earlier progress at removing the horizontal stacked ring. Since this was taken, I've done some more cleaning up the remaining line segments around the common rafters and have started a few configurations for windows, being horizontal and vertical rectangles, as well as reconfiguring the common rafter segments to reduce the need for secondary rafterlike purlins. The direction I'm inclined to pursue is having two on the bottom half of each face, plus one in the middle for the top part. It'd mean having to add a second band around the circumference of the roof, but not a huge deal. Getting the proportions right might be tricky along with the safe free spanning loading, but if that's the hardest part of this I'd be surprised!

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×