Hip Roof Model 2, mailing 2: keta intersection

An on-going study exploring traditional carpentry layout techniques.
User avatar
Brian
Deshi
Posts: 1090
Joined: Sat Nov 29, 2014 11:46 am

Re: Hip Roof Model 2, mailing 2: keta intersection

Tue Jul 11, 2017 5:20 pm

Ah that makes a great deal of sense that it would fail that way, and I remember in your writing a mention of angling the sides of the pin to increase the resistance to that twisting effect.

In the case of the pin, is the mode of failure a sheared pin?

I've often wondered if blind or half blind joinery is slightly stronger due to a higher resistance to splitting, but maybe not.
User avatar
Chris Hall
Site Admin
Contact:
Location: Greenfield, Massachusetts
Posts: 2385
Joined: Mon Nov 22, 2010 12:46 pm

Re: Hip Roof Model 2, mailing 2: keta intersection

Tue Jul 11, 2017 6:58 pm

Brian wrote:
Tue Jul 11, 2017 5:20 pm

In the case of the pin, is the mode of failure a sheared pin?
It depends upon the details of the connection and the load path, along with variances in the wood itself as to what portion of the joint fails first and in what manner.

There is no doubt however that the pinning does not make for a really rigid joint, and again, depending upon things, that is usually a good quality.
John Whitley
Deshi
Location: Seattle, WA
Posts: 148
Joined: Sun May 04, 2014 8:56 pm

Re: Hip Roof Model 2, mailing 2: keta intersection

Tue Jul 11, 2017 7:57 pm

The last few posts in this thread are really fantastic, particularly the shachi-sen failure photos. I increasingly feel like discussion of failure modes is vital to fully understand the function of a joint, and joined structures as a whole, and is so often missing from even well-regarded resources beyond "hey, yeah, wood, it moves."
User avatar
Brian
Deshi
Posts: 1090
Joined: Sat Nov 29, 2014 11:46 am

Re: Hip Roof Model 2, mailing 2: keta intersection

Wed Jul 12, 2017 9:00 am

Thanks Chris! Appreciate the information as always!
User avatar
Chris Hall
Site Admin
Contact:
Location: Greenfield, Massachusetts
Posts: 2385
Joined: Mon Nov 22, 2010 12:46 pm

Re: Hip Roof Model 2, mailing 2: keta intersection

Wed Jul 12, 2017 1:33 pm

Brian wrote:
Tue Jul 11, 2017 5:20 pm
Ah that makes a great deal of sense that it would fail that way, and I remember in your writing a mention of angling the sides of the pin to increase the resistance to that twisting effect.


Actually, as regards to making shachi sen which are in cross section rectangles, versus parallelogram in section, I think this is an interesting area worth a look in detail.

Let's examine first a rectangular-section pin, placed within the layout boundary lines of 0.25" for depth and 1.0" for length:

rectangular pin.jpg
rectangular pin.jpg (116.15 KiB) Viewed 5927 times
Note that if you employ those boundary lines and cut to them, the pin itself has to be about 0.0075" thicker than the nominal 0.25" it might appear to need to be otherwise. I chose to meet the depth line at the intersection of the 1.0" line, and projected over 90˚ from there back to the edge line,, and you can see we end up therefore over-running the 1.0 boundary.

One could choose to meet the boundary of 1.0" and edge line with the 90˚ corner of the pin, like this:

rectangular pin II.jpg
rectangular pin II.jpg (135.3 KiB) Viewed 5924 times
In this case, the pin needs to be over-dimension for thickness and under-dimension for width to meet such a mortise. Using a stock 1.0"x 0.25" thick material for the pin in either of the above cases, and cutting to the marks results in mortise to tenon fits which are not as sweet as they could be. A cut out pitfall for those that inspect closely afterwards, perhaps.

When the rectangular pin is driven in, a portion of the load is transferred laterally outward into the receiving member, and we have seen what happens in those joints when stressed to failure and the rolling of the pins levers the side of the stick open until it splits.

The parallelogram-section pin can now be compared:

parallelogram pin.jpg
parallelogram pin.jpg (110.39 KiB) Viewed 5927 times
The main improvement realized by the parallelogram is the load transfer directly into a 90˚ end grain abutment in the receiving piece, in parallel alignment to the grain of that piece. That appears strong and all, however there is a slight tendency now for the pin load to be transferred to shear the wood abutment away from the receiving piece inwards - longitudinal shear in other words. Pins for this situation are actually made about 0.0075" under the 0.25" thickness when you mortise to a 0.25" line of depth. If one didn't notice this, and made the pins exactly 0.25" thick, then they would be rewarded by a fitting situation in which the mortise, though cut to the lines, seems too tight for the pin you precisely sized (because it IS). If your cut out is sloppy anyhow, the foregoing is a moot point.

Another comment that could be made about the parallelogram arrangement is that the trench for the shachi sen has an acute internal angle, and I dare say this is a little harder to cut cleanly than one with an obtuse angle. Jabbing or slicing hard into the tip of the acute interior angle more easily risks that the chisel will tend to cut a little deeper than you might like into the surrounding wood, and this can precipitate a split later when the joint is loaded. The acute internal angle also represents also a more concentrated point load than when the abutment is square.

I have tended to prefer the parallelogram pins in that past, though they do make for a slight increase the cut out difficulty over the rectangular, both in making the pins and mortises.

There is an argument (well, one I am at least making) for a middle position between the rectangle and parallelogram sections of shachi sen. It's a potential refinement. The argument for this approach being along the lines of those used for detailing the critical and highly-stressed heel joint in a truss, between the tie beam and the principle rafter, or chord, connecting to it on (the beam's) top surface: you ideally bisect the angle between the two and that bi-section sets the abutment angle in both mortise and shoulder on the connecting piece.

Here's what I mean, taking the same boundaries of 1.0" for width and 0.25" for depth

perfect version.jpg
perfect version.jpg (76.73 KiB) Viewed 5927 times
The angle in question is the complement to the one in a right triangle having a rise of 0.25" and a run of 1.0", namely 165.964˚.

Bisect that angle and you obtain the ideal middle ground abutment angle of 82.982˚:

perfect version yes.jpg
perfect version yes.jpg (125.29 KiB) Viewed 5927 times
And very conveniently, the required pin for this is exactly 0.25" thick. Cutting an 82.9˚ abutment is not a technical difficulty if you can cut them for the parallelogram-shaped pin and mortise shapes.
User avatar
Brian
Deshi
Posts: 1090
Joined: Sat Nov 29, 2014 11:46 am

Re: Hip Roof Model 2, mailing 2: keta intersection

Wed Jul 12, 2017 8:26 pm

Interesting, slightly more tools involved in the markout but the cutout is pretty much identical for me at least.

Thanks for discussing these effects to such detail, it is greatly appreciated and of course makes for interesting reading since I will certainly apply these details to future use of the pins.
User avatar
Chris Hall
Site Admin
Contact:
Location: Greenfield, Massachusetts
Posts: 2385
Joined: Mon Nov 22, 2010 12:46 pm

Re: Hip Roof Model 2, mailing 2: keta intersection

Thu Jul 13, 2017 8:29 am

I can add that if the stock is relatively wide in relation to the tenon, then there is less concern about the rectangular pin forcing the joint seam open when it is driven in. with thinner mortise sidewalls, however, the rectangular pin is best avoided I think. Here's an example of a situation where there is plenty of support around the pin:

a0322824_16212838.jpg
a0322824_16212838.jpg (58.05 KiB) Viewed 5904 times
User avatar
Yxoc
5
5
Posts: 395
Joined: Mon Jun 27, 2011 7:43 am

Re: Hip Roof Model 2, mailing 2: keta intersection

Thu Jul 13, 2017 8:32 am

Chris,
your knowledge in these matters is only exceeded by your generosity in sharing it. Thank you, sincerely, for the time and effort you put into these explanations.
User avatar
Chris Hall
Site Admin
Contact:
Location: Greenfield, Massachusetts
Posts: 2385
Joined: Mon Nov 22, 2010 12:46 pm

Re: Hip Roof Model 2, mailing 2: keta intersection

Sun Jul 16, 2017 8:19 pm

Let me know when I need to get rolling on the next mailing.
User avatar
Yxoc
5
5
Posts: 395
Joined: Mon Jun 27, 2011 7:43 am

Re: Hip Roof Model 2, mailing 2: keta intersection

Mon Jul 17, 2017 7:50 am

Sorry, a bit slow posting...

On the weekend I started with larger and longer sen stock and tapered slowly with the plane with trial fittings between each pass. I continued in this fashion until they each inserted 45mm of the 50mm depth, using only moderate hand force. I expect the last 5mm would go fairly easily with the mallet, but would be more permanently attached:
DSCF4466.JPG
DSCF4466.JPG (190.36 KiB) Viewed 5885 times
For my first try it was ok but I hope that in future I can do much better.

With Chris' details of the sen geometry from his carpentry text, I was quite happy with what I was trying to achieve however constructing such a shallow taper accurately was challenging. To lay out the trenches, I used a try square with the width of a hacksaw blade just under the edge of the stock, such that the stock was slightly lifted at the far edge. To taper the depth I subtracted a 1/2 mm from the depth measurement on the inner most edge of the mortice and tenon.

In retrospect, during the layout phase, I should have used a more mathematical approach to ensure that the taper in width and depth were equivalent. By eyeball it looked pretty close.

When tapering the sen to fit the mortice, I took alternating passes with the plane on the cheeks and edges to adjust the fit. Again I didn't do any measurement to check that the taper of the sen matched the taper I had established in the trenches, it was a trial and error approach.

Any tips for fitting these things would be welcomed.

Chris, I'm happy to move on with the next mailing now.

Return to “Carpentry Drawing”

Who is online

Users browsing this forum: No registered users and 3 guests