Tuesday, December 10, 2013

Can't See the Forest for the Trees (III)

In the previous post I delved into the topic of log scaling to a certain extent, and now want to move on to look at the sawing process. Just as there are various ways to scale a log and compute the volume and/or board footage, there are various ways to saw up a log, and pros and cons for each. Again, I don't speak as a professional sawyer, so my knowledge is limited in that respect. Still, nearly every woodworking project involves sawing material at some step, and having some grasp about how a log is cut translates into better understanding and planning when assessing a board to be cut.

I'm going to assume the log we will deal with is a tapered, truncated cone, or frustum. The  model tree is 48" at the base and reaches 100' high. This is the sort of taper one might see with a cedar tree. Other trees are less tapered, but using a stronger taper shows the results of the cuts more clearly, I think, so that's the direction from which I'll proceed.

I drew a tapered cross section, made it into a cone, and then chopped the bottom 10' off:

The resulting cut log tapers uniformly 2.4" each side, over 10' length, working out to about 1.15˚ per side:

I don't think there is anything especially unusual about this section of log, save for its large size and geometric perfection.

There are various saws with which we might slice up such a log, and we could make the cuts in a horizontal orientation or a vertical one. There are bandsaws that cut each of those directions, or angles in between for that matter. Sans power, there is pit sawing, where the cuts are made using a long handsaw, either by one person alone, or two working in tandem, top dog and under dog fashion, or two working separately:

Or one could sit by the side of the log and rip it horizontally as well.

So, for purposes here,  I'll simply assume the cutting will occur horizontally, the sort of thing that could happen on a Woodmizer or chainsaw-based portable mill, bearing in mind that the various ways of cutting we'll look at could equally be done on a machine which saws vertically.

Speaking of the various ways of cutting up a log, there are many, and most methods developed revolve around maximizing yield. we're going to considerably simplify matters though, and ignore yield altogether, focusing in on cutting the same log in three different ways, the aim being to cut three different boards. Those boards are, specifically, flatsawn, riftsawn, and quartersawn.

In cutting method one, we take the log section we have, and rest it on the deck of the mill 'as is'. This is an exotic softwood, heh-heh, having alternating green and white growth rings.

The three cuts we looks are are labeled on the end of the log, as A (flatsawn), B (riftsawn) and C (quartersawn):

In cutting method one, no compensation is made for the taper of the log. The log lays on the mill cross members 'as is'. The whole log, relative to its pith, is tilted down 1.15˚.

In the above view, the narrow end of the log is facing us, as this is the end at which cutting is always originated, proceeding down the tree trunk to the butt, for obvious reasons.

For board A, we would make a preliminary slabbing cut then follow it up with a second slabbing cut 2" lower, which would produce a waney board like this:

The 'cathedral' pattern on the face of the board is probably what most people think of when they think of 'wood grain'.

That 2' slab is then edged, the cuts run in alignment to the pith of the log, and a 2"x12" board is produced. I'm choosing this size of board for illustrative purposes - a wide board shows the grain patterns that result from sawing more obviously than does a narrow board.

We can execute a similar set of cuts for the other two boards,  B and C, slabbing and edging, until we have produced the desired three sticks, A, B, and C:

Let's set those 2x12s aside for the moment and start all over again from the beginning with the log. This time we will make some adjustment for the log taper, such that the very center of the tree, the pith, is aligned to our cut direction. The pith is therefore horizontal. We can call this 'sawing to the inside' of the log. Some sawmills have built-in devices to raise or lower one end of a log, and if we were cutting a long log we would need to look at supporting the log at intermediate points as well so as to maintain a straight section throughout.

Again, we see the three cuts we are after, which I will label A', B', and C':

Note the presence of a shimming board under the narrow end, roughly a 2"x4". To keep the pith of the log level with the sawmill deck, the log is tilted up 1.15˚.

We take the three cuts as before, slabbing and edging, the edging being done in alignment to the pith. That leaves three boards, A', B', and C':

We'll set those aside and start again.

The last cutting alignment we'll look at also has an adjustment for taper. This time though, instead of shimming the log so that the pith of the log is parallel to the cutline, we shim up the narrow end a bit more so that the top of of the log is parallel to the cut line, i.e., horizontal. Our three boards are now A", B", and C":

As you can see, the shim used is much thicker than when we cut in alignment to the pith, twice as thick to be precise, a 4"x4". Now the lower surface of the log is tilted up 2.3˚ relative to the deck of the mill.

Previously, we cut to the 'inside' - now we are going to cut to the 'outside' of the log.

The three boards are produced once again, labeled A", B", and C":

Now we have produced boards which are flatsawn (A, A', and A"), riftsawn (B, B', and B"), and quartersawn (C, C', and C"), we can compare them to see how the run of the grain compares. First, we look at the A set, flatsawn:

As you can see, the log sawn without regard to taper produces a flatsawn board with considerable grain run out, and numerous cathedrals on the tangential faces. Here's another view of the same three boards so you can see the run of the grain along the board edges, along with the pith-facing portion of the board:

Again, the log sawn without attending to taper produces the board with the most angling of the grain. The board sawn so along the pith line, A', exhibits less grain run out, while the board sawn to the outside of the trunk, A", exhibits virtually no grain run out. The alignment of the grain perfectly along the long axis of the stick makes the board sawn to the outside of the tree, A", the strongest of the three were they to be used as posts. If the above sticks were to be used as beams, again A", with the grain running straight down the stick, should bend the most evenly and predictably of the three I would think.

The other two methods of sawing, A' and A, produce some amount of grain slope on the faces of the sticks. While the end grain view of the three sticks is very similar, you can clearly see the degree to which the grain runs at a slope along the sticks by comparing the narrow edges of the boards.

The B set of sticks were all sawn in a rift grain orientation when viewing the end grain. Let's see how they compare:

Again, the sticks sawn with accommodation to taper, B' (to the 'inside') and B" (sawn parallel to the 'outside' of the log), show the least amount of grain slope, with B" showing the least amount of grain slope of the two.

Here's another view of the B set:

Finally we can compare the sticks which were quartersawn, C set:

Here, given that the grain runs 90˚ radially, the effect of sawing for taper (or not) is minimized, however it is still clear the sawing to the outside of the trunk, as in C", produces the least grain slope of the three cuts.

Another view:

We'll look more at 'sawing to the outside' in a subsequent post.  There are various reasons why we might want straight grained timber as a result of sawing practice - better strength, which is as critical in a dining table as it is in a building column, is one important reason, and unless sloping grain is desired as an aesthetic, conveys the line of a frame more clearly than does sloping grain, which, if extreme enough, 'fights' the lines of the piece.

Thanks for coming by the Carpentry Way.


  1. ooh, these posts are going to be interesting!

    i never spent much thought about the intricacies of slicing up a log and how they affect grain orientation in the resulting boards. i take it that the A″/B″/C″ versions are the most rarely encountered ones, judging from how many board stacks one has to flip through to find one or two pieces without grain run-out.

    here is another use for straight-grained boards: wooden bows. while bowyers traditionally work from split staves, using unviolated growth rings as the bow’s back (the side pointing away from the archer, the one put under tension), bows can also be made from straight-grained boards, no matter whether they are flat-, rift- or quartersawn.

    sadly, grain orientation is of so little concern in most places that we are either left to hunt through board stack after board stack or buy from a specialty provider. (at a premium, of course!) or cut, split and dry our own staves, of course, which presents its very own set of challenges.

    many thanks for the interesting posts!


  2. The power of Sketchup is strong with this one...

    1. Hah! Well, i will say that Sketchup allows for the possiblility of clearly seeing how cuts are manifest in the boards, but it was a tedious process. I tried playing around with the section plane tool for a while, but ended up simple using the 'intersect faces' tool, over and over, and re-sticking the boards back together after deleting the rest of the log. It took hours to complete the drawings, and I am sure there must be an easier way, so SketchUp function, or special plug-in that I have yet to learn.


  3. Hi
    I like the way you start and then conclude your thoughts. Thanks for this nice information regarding regarding I really appreciate your work, keep it up.

    1. Dave,

      most kind of you to say, and rewarding to read. Thanks for your comment!


  4. Hi Chris,
    I am enjoying this informative series and your sense for sharp analysis. It always shows things from a clear perspective.
    However some questions rose up while reading in regards to the difference between the theoretical and practical side of things. I can see why using a cone and small section of this cone-shaped trunk is a good choice to explain the theoretical side of resawing. As a carpenter building timberframes we often use long and large sections of lumber and since trees don't grow perfectly straight resawing them is not as straightforward as the illustrations might suggest. Even if the tree has very few defects, knots and is 'straight grained', the direction of the grain can still run 'all over the place'. this occurs sometimes in only a small section of the log or on one side but nonetheless.
    Dealing with logs like this becomes more challenging. I too find it very important to orient every single member in a building in it's most optimal direction and often find it difficult to decide how a piece with less then perfect grain should be resawn or used. I hope to read some more about how you deal with the practical side of resawing and which compromises you might make over others if necessary.

    Being located in Western Europe at the moment doesn't really help in that regard. Big trees have been cut long long time ago and regarding softwoods not much remains unless you are satisfied with knotty 2by4's. On top of that I can say that here in Belgium it has come to the point that building with large timbers is a big challenge and utterly impossible if you do not want to rely on import. In this place on the globe building with local materials means building a brick house.

    1. Mathieu,

      good to hear from you as always. Yes, the posts so far have been very theoretical and use a perfect section of trunk which probably has never been produced by nature. At least it is true that, for the most part these days, we are sawing up trees that have far from perfect shapes and these do provide added challenges. As this series proceeds I'll be able to explore that topic in more detail.


  5. Interesting post, I'll have to forward it to my brother that has a woodmizer. The only thing I question is when comparing the quartersawn boards-- I would think board C' would have the least grain runout, because you are taking it from the halfway down portion of the log. When you "saw to the inside" it seems you will have best grain alignment with a shim that sets the pith level to the mill bed.

    1. Pete,

      you know, thinking about it, I would agree that C' should show the least grain run out. I'm not sure why the 'pieces' i cut out for the three cutting patterns did not show that result, however it is also worth bearing in mind that a quartersawn orientation, with all three boards, remained the most consistent and least run out. I'll revisit my drawing though and check that i haven't mixed a couple of pieces up or had some other odd thing happen. Thanks for your comment!


    2. After my above reply, I later realized why the C' cut is not as straight grained as the C" cut: the edging cuts were different. In C', I followed the idea of the slabbing cuts for that board, which were made parallel to the pith, and made the edging cuts also parallel to the pith. If those edging cuts were parallel to the bark, the result would have been perfectly straight grain.

      In case C", I again follow the logic of the slabbing cut, which was made parallel to the outside of the log, and made both slabbing and edging cuts parallel to the outside.


    3. I think I follow what you are saying, but for sure C' would have the least grain run out as seen in the edge plane, which is what you are adjusting from C to C' to C'', at least for that midway point of the log quartersawn board you show. But yes how you do the edging cuts can adjust the amount of run out as seen in the face plane. Right? :) Thanks for the replies, and hopefully you don't take this as nitpicking.


    4. Pete,

      you're quite right how the edging cuts are done definitely affects the amount of run out apparent on the broad faces of the stick. Doing the edging cuts for C and C' in the manner of C" would have meant less visible run out, however this only applies in the case of the quartersawn stick.

      consider: In A-A'-A", the edging cuts are automatically parallel with the pith, while in the B-B'-B" set pf boards, cutting to correct the grain direction on the broad faces would entail a lot of waste on B, a moderate amount of waste on B', and almost no waste on B". Overall, slab cutting to the outside, and edging parallel to the pith produces the straightest grained boards, all things being equal.


  6. Chris,

    A great post, elevates your usual great post quality to a greater level of greatness! Semi-kidding aside, this post is very helpful for me since I do quite a bit of chainsaw milling so this issue of getting the best grain with least runout is one that I reflect on now and again.
    I was in bed last night doing mental gymnastics trying to visualize the same issue that Pete raised: why C’ did not have the best grain. Thank you for clarifying that.
    Is it not the case that with cut C having no height compensation, it is therefore parallel to the bottom of the log? And that cut C’’ has height compensation so that it is parallel to the top of the log. Since the log is symmetrical, are C and C’’ in fact identical?


    1. Hi Dan, and thanks for your comment and questions.

      "Is it not the case that with cut C having no height compensation, it is therefore parallel to the bottom of the log?"

      Well, the cut itself IS made parallel to the bottom of the log (parallel to the log deck), however the stick in cut C was located along to the top of the log. If cut C were made so as to take the stick along the bottom of the log instead of the top of the log, it would have produced a cut identical to C" (assuming the edging cut was made in the same manner as C").

      Cut C however was made along the top of a log sitting on the deck with no taper adjustment, and given that the log is a cone, the angle of the grain run above the heart center of the log is different than the run of the grain below the heart center (as it is parallel to the log deck and therefore the sawblade cut) in the log. The cutting remains constant, parallel to the log deck of course, but the results of the cut differ based on where the cut is taken relative to the pith line, and whether adjustment has been made for taper or not. It can be a little confusing to visualize, and that's why I went to the trouble of sketching a tapered log section with pseudo growth rings, to aid in clear explanation.

      To repeat, cut C could be made to be identical to cut C" if cut C had been made along the bottom of the log instead of the top, and if the edging cuts were made parallel to the outside of the log instead of the pith.

      Chainsaw milling, at least in terms of the initial cuts, has the advantage of those cuts being parallel to the outside of the tree rather than the pith, given that the the initial set up board which guides the mill sits on the surface of the log (atop lag bolt heads) in parallel to the log surface.



All comments are moderated:

Anonymous comments will be deleted

Spam will be deleted

Comments containing links unrelated to blog content will be deleted