Hi all,
I've found interesting pictures from TFG workshop . I have a lot of questions how to connect crooked and curved logs with different diameters. But the first question is: could anyone give me just a brief and simply explanation of whole process on the pictures? Especially I don't clearly understand how to transfer lines and angles to logs.

Thanks in advance!
Ferd

Brief? Convert it to square and that will shorten the process. Silly log builders.

I don't completely understand the process in the pictures. It appears to me the carpenter has made a template to use as a basis for transferring dimensions from one timber to another. I have never done this, so I could only speculate on how it works.

I don't like joining logs together in the round, with contrary grain (that is, perpendicular or at an angle as shown). This is because of the way the wood will work in a round log so joined. I always prefer to reduce the wood around the joint in some way, even if the rest of the log is left round. This can be done to be very attractive and in keeping with the nature of the wood.

It's kind of like Tim says, convert it to square, only here it is only squared at the point of the joint. Study Norwegian log building, and you will see this is how they do it. I'd recommend you get the Book, The Craft of Log Building, by Hermann Phlepps, if you can find it. That will teach you a lot about working with logs as a building material.

Like Tim, I like things to be square. Even log walls, I like the wood to be square timber!

Thank you for reply. I think converting whole log into square takes more time. I know it is easier but it is different technology. Here are another interesting pictures by Joe Bartok. Some logs are converted to square and others remain round. Especially this valley housing makes me wonder. So I would like just to know how to transfer these angles to right places on round irregular logs.
Is here anyone who can explain the procedure? Or personally attended this worshop and can give some advices?

The logs in the link, if you look closely, are reduced at the joints to regualar rectangular dimensions. This will be your fastest and most efficient method.

Transferring round dimensions from one log to another in complex assemblies is possible, but it is time-consuming and difficult. There is much opportunity for mistakes, and it is very hard to visualize such joinery. Also, cutting the round profile joints that result is no simple task. I never bothered to learn how to do it, because it never seemed worthwhile. But that is entirely a matter of personal preference.

I wish I could help you more specifically, with the aspects of scribing round to round,

I have a 3d model of one of Roberts joints. I'll try and post a picture later. I was working with my little one at the barn raising (also in those images) - so I didn't see the whole process. I'll try to sum it up, and I'll invite Robert to post here.

When it comes to what is fastest,

round building is fastest for raising a relatively crude skeleton frame. This is seen historically, where round post construction or round log building happened first.
The disadvantage is in enclosure. You can make an undulating enclosure (that is, wobbly) easily, but to make a flat surface is difficult

The next step is to square surfaces that are going to have anything attached to them. For an example, look at the floor joists in many old barns which are squared only at the joints and along the top face. Another example is found often in Norway, where the interior surface is made flat while the exterior is left round.

Taking the whole structure into consideration, the fastest approach really is to square everything. This takes some time to hew or mill the timber, but from there on everything is faster. It is easier to envision and lay out the joints. The joints themselves, being square, are easier to cut, and the resulting frame is easier to close up.

However, I think maybe you are after the look of the round log building. If that's the case I would still recommend trying a reduced joint approach, where the logs are made regular at the joint locations.

Round joinery is a difficult skill to master and is labor intensive. From what I have seen, you don't find it historically.

All this into consideration, you are free to do whatever you want. If you want to learn scribed log joinery, that's great. There are few who do it. I've let you know why few do it, but if you still want to do it go ahead! Despite what might seem like skepticism from me, I do think it's a nice look.

I must say they did a beautiful job with their joinery.

I don't have much personal experience with round logs. So I'm only saying how I would approach it.

My first goal would be to envision the timber in the log. To do this I would start with a log a bit longer than the original dimension. I would rough cut square both ends as close as I could by sight. Then I would string a line from one end to the other in the middle of the top. This is a reference for the centre of the log. It will be a set distance above the centre point of your stock. If you need to offset this line at one or both ends with a plywood disc or square that is fine. Depending on your stock you might want to also snap a line. At this point you can square down off this line and left and right and establish a square end cut in relation to the line.
When your ends are cut square you can establish any point on the log with relation to an offset string line. Depth and mortise location can be determined by plumbing or levelling from the line.

We join round timbers quite often and it usually involves scribing with one piece set over another, then working off snap lines to level the pieces. A bubble scriber or laser level is used for round to round pieces or round to flat. For a description of the process. see issue #78 of Timber Framing or see the Timber Framing Fundamentals for the three-part scribing series I wrote. Both publications can be bought through the Guild store.

The Burlington process shown in photos indeed used a template and ground plane drawing. This was a radically new process developed by Rob Chambers because it allows you to join timbers of the same diameter (there's a mortise and tenon inside that joint). Previously one timber had to be at least two inches smaller in diameter to the other in order for the joint to look fluid.

Thanks to all for you contributions. I still hope I will find someone with experience in this method. Thanks Will B for comment, I know scribing techniques with scribe (compass-like tool) and I've done several logs. I understand how to use the cribe though I am not scribing specialist.
The logs on the pictures above are not scribed with traditional scribing technique I think. The logs were not set one over another. They were not scribed with scribe. They were prepared separately. It is interesting how precisly is possible to transfer tenon and mortise size to irregular logs.

Ferd, although I did not attend the conference where the presentation was offered nor have I been instructed by Robert Chambers, I have used some of the ideas underpining this method. The regulating concepts are a level and lined log, tools that detect plumb, level and square and finally control of slicing planes at closely controlled angles. The innovation here is in the use of a template with the elliptical cutout that models the miter plane. So by straddling the log held precisely at the miter angle, the log can be marked by projecting the plane to the log thus marking the cut line. The real challenge here is to make the initial cuts into the receiving piece while retaining the material for the saddle. Then the saddle is developed and cutting is continued.
so in a way this is not to my way of thinking a scribe process but rather more of a mapping out process.

Roger, I understand it the same way. The first step is leveling the beam log, the second step: draw vertical and horizontal cross on both ends of log and connect them along the log with chalk line.
OK, the next questions. I understand how this big board with elliptical cutout can help to draw line around log. If I hold pencil on the template surface the pencil line will be flat even on curved log. But is this template general-purpose (for all logs) or is it done differently for each log diameter? Why are there two cutouts? The bigger one (elliptical) is probably for sloping lines and cuts. But why is there another circular cutout? Is it for right angled or perpendicular cuts? Or for different logs? What is Robert doing on the first picture? What is measuring on the pictures #2, #5 and #6? And why? Is there any other way than using this big board with cutouts? As I said-I have a lot of questions....
Another question: the beam and the rafter here in pictures have almost same diameter. How to proceed if diameters are different and rafter is smaller? Then the crossing of both logs will be above the beam middle line..

Here are my thoughts having seen bits of the presentation, from knowing what the joint looks like, and from looking at the photos:

centerlines are established on the logs

geometry developed based on the centerlines for a rafter to tie connection
(with logs of similar diameter, one could assume a near center line to center line match in how the logs plane out to one another)

the geometry of the joint in this case is developing 2 planes as primary to layout, and an interior rectilinear block of timber

the elliptical cut outs are simply a convenient way to trace the planes onto the surface of the logs - i would guess one could have several of these for various angles and log sizes - but if the cutout is close in size, flat to itself, with the angle held correctly, and the pencil used judiciously, it would seem you could make this work with a few templates - longer ellipses for shallow angles, shorter for steeper - and i'd imagine a scenario where having a bubble gauge or inclinometer mounted to the patterns would be useful

once the planes are developed, and the angles off of centerline known - the geometry can be added to both timbers

the rafter is easier to cut - in that the 2 planes can be cut completely through, and any interior joinery removed later

the tie is more difficult - as the width of the interior rectilinear shape needs to be projected onto the curved surface of the log before cutting so that one maintains the internal shape


i see timberwrestler in a few of those photos - i'll see if he can comment. and i'll try and get robert chambers to check in here. his website is here - there might be some useful information there:
http://www.logbuilding.org/TOC.html

and he makes these great scribers (i have no need for them - but have held them at a TFG conference):
http://www.chamberslogscribers.com

In this case, I think the template is specific to the logs diameter and the varied slopes of the planes requires two different elliptical cuts. I think that there is a close tolerance in the cuts to limit error in transfer. Also this looks like a center line layout, so the origin of the layout is on the horizontal center line. Therefore the template must extend past the critical dimension mark, so the template can be assured to register on the mark.

Since I did not witness the demo, I think that any specific comments on each photo could be wildly wrong. A number of things are not clear to me in the photos. I can guess but I'll pass for now.

Roger,

If the plane is held at the correct angle, and the pencil held flat to the plane, then it would matter little how close the ellipse follows the timber - so long as there was enough meat for the pencil to ride accurately. This assumes that there is a way to register the template to the proper location on the log without relying on the templates fit to the exact curve, as you'd need an awful lot of templates - unless you were working with very closely sized poles.

The trick I want to know is how the planes are held at their exact angle once their position is determined on the log.

But yes, it would be good to hear from someone that was at the demo.

My observation relates to the photos. It does seem to me that the template must register to the critical dimension marked on the horizontal reference so a reasonable closeness is needed.

With a level log, I would gauge the slope with a Bosch electronic bevel, it's in my kit. Not a clue to how Chambers goes about this.

Hello Team,

Sorry that I've missed the conversation. Busy.

At conference I had a full classroom session plus a demonstration, and that was not enough time for anything but a brief introduction. The method is new, the joinery is new, and the layout methods are new. At least, I've never seen this before in my 30 years of working with big wood.

First, a few comments on philosophy and my life. For those who think that all wood should be either square or round -- that's a rather small idea. The TFG members who I know, and who I work with, don't see it that way. I choose to work with logs because I love the look of naturally-shaped, smooth, tapered, hand-drawknifed trees. To 'neck' a beautiful, naked, natural log down to a square timber at the joinery is against my religion. I won't be discussing my religion in any further posts. No need to comment on the superiority of square since 2x4’s are square. And it’s quite easy to make strong buildings out of them. So, it is obvious to me that the 'square' idea is not the best idea to be found in timber framing. I think your best idea is more likely to be the 'big wood' idea, and the traditional joinery idea.

Sometimes I use a scriber to transfer the natural contours of one log onto another natural log. Log walls are built that way. But for structural assemblies, it’s difficult to make scribed joints strong. Not impossible, but difficult. But the second problem is that the initial super-tight fits of scribed joints and notches suffer unless we can use gravity and shrinkage to help us keep them tight. The corner notches of log walls get the benefits of gravity, compression, and shrinkage, and these keep our corner joints tight over time. But we don't get all those benefits in the joints of, say, a roof truss made of natural logs. So, using only scribed joinery for space-frames like trusses, or bents, is not a good idea.

My new joinery is not scribed, but it does appear to be a scribed joint because the surface contours of the two logs intersect cleanly, as scribed joints do. My new joint shrinks together, and stays tight over time. It also has big, flat, bearing and shear surfaces totally hidden inside . . . and the engineers tell me they like that. Ed Levin was my co-presenter at the Vermont conference, and Mack Magee at the log builder's conference— maybe Mack can answer engineering questions in this thread?

Each log gets 4 chalklines, and at both ends they get 'cross-hairs' connecting the lines. In a truss, the 2 chalklines that are in the plane of the truss are snapped so they divide log diameter in half at the joints. These sorts of chalklines will be familiar to some of you. But, the 2 chalklines that are in the horizontal plane on each log, are snapped so they go through the thickest part of the log at the joints. (If I used machined logs, then these 2 horizontal lines would be the same as the ‘thickest part,’ but since the logs are naturally-shaped we need to reference off the ‘thickest part.’) You can think of ‘thickest part’ as being the greatest diameter (but actually it’s the longest chord length). This horizontal chalkline is a new kind of chalkline.

Where log meets log, both logs should have close to the same diameter. The closer the two diameters match, the better the joint looks when it’s finished. Here’s why: When two cylinders of equal diameter intersect at their midpoint axes, the joint surfaces are made of two flat planes. (The edges where the two cylinders meet are not straight, they are hyperbolas.) Equal diameters (chord lengths) is the key to good looks. And of course, the interior, flat bearing surfaces are easy to cut (because they are not coped).

If two cylinders have unequal diameters where they are joined, then you get a complex curve in three dimensions -- and if you want the joint to look good, then it really has to be scribed to make the surfaces of the two cylinders meet happily (and that's how boilermakers weld these type of joints). Again, this is why I am not using logs of unequal diameters (at each joint). Of course, the logs are tapered, so their diameters vary at each truss joint. It is easy for log builders to find logs with equal diameters at each of the joints because we have a lot of logs in inventory. This lack of choice will be a hurdle to timber framers. If you try my method to join an 11" log to a 12 -1/2" log you will be very unhappy with the result.

The ‘vertical’ chalklines, in the finished and installed truss, are in a single plane, and this plane is plumb. The ‘horizontal’ chalklines, when viewed in elevation in the finished and installed truss, are perfect right triangles, and the pitched chalklines of the top chords are at the roof pitch. The horizontal chalklines meet at points on the surface of the logs (which is the 'hinge' of each joint, where the two internal flat surfaces meet).

We will be building one of my trusses at the ILBA conference on Vancouver Island, in a hands-on course March 26-27, 2014. www.LogAssociation.org I’ll be publishing a booklet on the method, and maybe an article in Timber Framing—Ken and I have talked about it. But, it's still new, and I want to use it more before I get it into print.

I do not regularly scan these forums, so I apologize to you in advance for my future slow responses.

Yup, I was at Robert's talk and demonstration. Very cool stuff.

Ferd, it's really a very specific joint for pretty specific situations. It's technically a mitered joint, but only for logs with the same diameter (as Robert pointed out). You can miter logs with different diameters, but you end up having to fair them out at the joint, because they will never meet perfectly. What Robert realized by looking at some obscure geometry is that they do perfectly meet when the diameters are the same. The result is a really clean joint, that also can include some internal joinery.

He presented a few methods of executing the joint. The standard way would be with lofting, but there are some disadvantages to that. The method he is showing is basically using a template. I'd have to review my notes to try to explain it, and it's not something that is very explainable in words alone. I don't think that he's measuring in those photos, but rather bringing a point up from the template.

I can try to answer more specific questions, but most of this sort of thing is hard to explain. I would definitely check out Will's articles that he mentioned. And the ILBA has a book on cutting with jigs that goes over a bunch of different approaches.

Each joint has two joinery planes. If we have 3 points on the surface of a natural log then those 3 points determine one joinery plane. Another layout method requires only 2 points plus the slope of the joinery plane (2 points and an angle determines a plane, as does 3 points). Which method is preferred depends on how fast you want to work, whether your logs are roundish in section, or are oval/lumpy/weird, and how many jigs you want to own.

Two of the points that we need for both layout methods are easy to establish: these 2 'easy' points will always be located on the horizontal chalklines. And, if you could connect these two points with a line through the log, then that line will be normal (90 degrees) to the plane described by the vertical chalklines. (You cannot actually connect these two 'easy' points until after the joint has been cut, of course, there's wood in the way!) I call these 2 easy points the 'hinge' points -- because the two joinery planes of each joint appear to hinge here. There are always 2, and only 2, hinge-points per joint.

For the 2-point-and-angle method, once we have the 2 hinge points, we need the angle. Each joinery plane bisects its involved angle. I love how simple this is; and the engineers like that the bearing surfaces are always exactly the Hankinson angle they want for grain slope. An example: two logs to be joined to each other at 45° (eg the top chord of a 12:12 truss joined to the bottom chord). The joinery planes for this joint will 22.5° (half of 45°) and 67.5° (half of 135°). These angles are measured from the horizontal chalklines of each log. A piece of MDF can be positioned like a horse collar over the log, and it is lined up with the 2 hinge points, and then is set at the desired angle. (The MDF in the photos is not a template: it has no markings on it, and the ellipse was roughly cut only so it would fit over the log. It is a reference plane, not a template.) Trace from the MDF onto the log’s surface. That’s your cut-line for this first joinery plane. Re-position the MDF through the hinge points and at the other angle you need for this joint. Then repeat on the other log that will join this one (one is male, and one female, of course). All measurements (span, height, length of pieces, etc) are made along the horizontal chalklines. That is, the length from one joint to the joint at the other end of a log is a measurement from hinge-point to hinge-point, and the required length is found by trig (or proportions, pythag, or etc). All the joinery can be laid out on all the logs before anything is cut, the logs are never positioned over each other, are never leveled lengthwise, there is no lofting, and no lines are snapped on the floor.

I should mention that snapping accurate chalklines on natural logs is not as quick and easy as it might appear. There are unique skills to putting chalklines onto logs. The line must be aimed, not just pulled back and let ‘er rip, but that’s another topic. Just a heads-up that you need skill in the basics to get good results here.

A much faster way to lay out the 2-points-and-angle can be viewed in my short video here:
http://youtu.be/hO6w3njpzqY
In the vid, ‘HP’ is hinge point—you can see that the jig really does hinge on these two points, and then is locked at the required angle. The 2-point-and-angle layout produces joinery planes that are always 90° to each other, so you set the jig once and mark both joinery planes. (It’s great the way the geometry works out. I wasn’t expecting this when I first started working on it. But when layout, engineering, & cutting all click like this, I knew I’d found something sweet.) If you’re keeping score-- from the example above, you’ll see that 22.5° plus 67.5° = 90°. For accuracy, I use large MDF triangles cut on a table saw to the angles I need, not that tiny digital protractor. And my jig, made of “8020” parts, has been slightly improved since I made the video last summer. The truss in the still photos was built by students in my 2013 Univ of Alaska course. They had no timber frame experience, and had been working with logs less than 3-weeks. This was the first-ever truss built using my method and joint, and I think they did a pretty good job with it.

Now, for the 3-point-layout, which is the method I demo’d for the TFG, we use the 2 hinge-points (identical to the HP’s used in the 2-point-and angle method) but we also need to determine where on the vertical chalkline the surfaces of the two logs will intersect after they have been joined—we need ‘surface points.’ If the logs are not very circular in section, or have substantial sweep or bumps, then the 3-point method can produce a more attractive joint than the 2-point-and-angle method. (Not a better fitting joint, a joint that has log surfaces that are more “fair”.) To find where the log surfaces will intersect on the vertical chalklines, you could position the two logs over each other and then use a scriber, or a line-projecting laser, or a plumb-bob. But accurately positioning 1-ton logs over each other is a time-consuming, hair-tearing-out hassle. And it has to be repeated for every pair of logs you will be connecting. So I came up with a portable story-board (foam core) to find the surface intersection points. No log has to be leveled lengthwise. The story board is pinned to the HP of one of the logs and then that log’s surface is projected onto it, and drawn. Next, pin the story board to the HP of the mating log, and project it’s surface and draw it. The log surface projections will cross each other on the story-board and these are the points we are looking for. They are easily transferred from the storyboard back onto the logs.

This gives us 3 known points marked on the log for each joinery plane: 2 hinge points plus 1 surface-point. I recommend you use the MDF horse collar, sling it over the log so it goes through the 3 points, and then trace it, score the line, and cut. Someone truly skilled with a flexy rule can join the 3 points without using the MDF horse collar, but I don’t think many timber framers have that skill yet—it takes practice to get accuracy.

The 3-point method and portable storyboard produces lofting benefits and lofting results . . . without lofting. The logs to be joined can be sitting in different corners of the yard, or even in different hemispheres, if you don’t mind mailing the portable story board. I decided to demonstrate this technique at Burlington because I know that some timber framers loft, and I figured that they might want to consider the portable storyboard idea for their own purposes. The portable storyboard solves lofting problems that come with timbers (and logs) that are heavy, long, awkward, unstable, numerous, or distant from each other. You just carry the information on the storyboard from timber to timber, without having to carry (or position) any timbers. It could even be used to 'loft' pieces that cannot be moved—say if one piece is already part of a structure.

Please note that 3-point layout produces joinery planes that are probably not at 90° to each other in a joint. This isn’t a problem: I’m just saying that only the 2-point-and-angle layout produce joints that are necessarily 90°. With the 3-point method I never even measure the angles of the joinery planes, because I don’t need that information. The joints are equally tight with both layout methods.

thats pretty close to the type of jig i envisioned as i was thinking through this last night. i was figuring you would be offsetting the plywood ellipse to the bottom of the jig for the thickness of the carpenters pencil....

very cool robert. and I appreciate you posting here.

Nice jig.

Very creative!

I can see such a jig being used on so so square timber.

Hello Robert and all others,

Thank you for your participation in this thread and your comments. I hope this topic will be helpful not just to me but to other people too. Especially I would like to thank Robert for his detail description and interesting video. I am surprised how precisely you are marking your logs with marking knife. It seems to be a Japanese furniture and not part of roof. :-)
Robert, I share your philosophy. I know there are probably a lot of other ways that could be cheaper or faster. But I like naturally-shaped logs as well.
I couldn't understand what you measured in pictures with bubble-level because I thought you need just measure angle between board with cut-outs and horizontal plane/side line on the log as you described in method 2 points and angle. It is much clearer now.

If I understand your 3 point method, you are looking for logs with same horizontal diameter. And if there are differences in vertical diameters you'll find intersection point of both logs by drawing their shape onto board and find their intersection. So the angle of joinery plane is not exactly 22.5° (=45°/2, the bigger part of joinery) but slightly differ depending on log vertical diameter differences. It is great idea!
And it is really nice how logs are connected without any gaps and overhanging. If there are not any bumps and knots the log surfaces are smoothly connected. If there are bumps the logs are smoothly connected at least in both HP and both points where vertical plane meets log surface (I mean points of crossing of upper chalk line on the beam log with upper and lower chalk line on rafter log).

You have a big advantage you can choose from a lot of different sizes logs, different shapes and diameters. Unfortunately I have limited supply. If I have problems to find the right logs for such joinery I hope there is another solution. It's not the same philosophy, but it could be helpful. I mean adjusting just rafter log shape to flat/square shape. And leave the beam log untouched circular. The same way is done this or this joinery. It is not so nice but it could solve my problem with different log sizes. And the way how to transfer lines and angles onto log is probably similar but more difficult.
If someone has any other ideas I'll appreciate them.

Hi Ferd,

My book, the Log Construction Manual, has a chapter with detailed instructions for how to lay out and cut these joints when the two logs are not the same diameter. It's for sale at my Log Building website and Amazon.