Hey, cool to see my article on the front page of hacker news. Never expected THAT article to hit hacker news.
Got a question, I could answer.
Quick question: do you recall what wood you used for the box/dovetail article? It looks similar to the spruce from the earlier joint strength article. I suppose using a stronger wood would just get you closer to where the glue is failing more than the wood, but I'm just curious.
Less-quick question, feel free to punt/ignore: any thoughts on deliberately building in a very, very slight gap between box joint fingers? I would expect this to reduce joint strength (thanks to reduced force between fingers and surface area in contact) but potentially also mitigate wood stress from movement in service due to humidity. Though I imagine if a work piece is big enough for me to be concerned about movement in service, I probably shouldn't be using box joints!
As for gaps in joints, if its a small gap, it doesn't hurt. I actually did some tests in that regard, linked from the article. http://woodgears.ca/joint_strength/glue_methods.html
A small gap actually helped. Probably because the glue soaks into the wood, so a joint without any gap sort of runs out of glue too quick (my theory). But it suggests that putting glue on and leaving the joint open for a minute may be beneficial.
It's great that you continue to post videos. Thanks so much for combining two of my favorite things: woodworking and engineering!
Have you ever used your joint-strength rig to determine the minimum M&T geometry or dowel configuration required to make the joint fail before the surrounding wood fails? It seems to me that if all the joints are stronger than necessary anyway, your tests were really testing the joint, plus the wood (which you tried to control), plus your own assumptions about the equivalency of different joinery techniques, rather than just the joints.
I'd also be interested to see a test of hybrid joints, such as a screw at the end of the joint under compression, and an M&T or dowel at the end of the joint that's under tension. A screw-head won't pull out through the wood, if it's mostly being pushed and sheared rather than pulled. I'm thinking a dowel at 45 degrees on the tension side, and then drive a perpendicular screw into the end grain through the compression side.
Your tests also suggest that, since the joints tended to fail along the glue line, dimpling the tenons, or adding ring-shaped grooves to the dowels--rather than the grooves parallel to the long axis or slightly helical that are typical in pre-cut dowels--would increase the joint strength. The interior surface of the mortise or dowel hole would have to be similarly roughened. Perhaps thread the dowel hole, as if for a larger bolt, and thread the dowel rod, as though for a smaller hole, and let the wood glue bite into both like the threads of a screw?
<< ###wood### << ###wood###
<< ########## << #/\#/\#/\#
~~~glue~~~ VS ~~~glue~~~
########## >> /#\/#\/#\/ >>
###wood### >> ###wood### >>Weakening the dowel is okay, since it wasn't the dowel that failed. It's probably more important to rough up the inside of the hole anyway, since the failure photos mostly showed that the glue was still stuck to the dowels, and it was wood from the inside surface of the hole that broke away. And the end-grain in the rail somehow frustrated the cracks better than the grain in the post.
1. After you moved out to the country, and discovered you did not want to handle the lifestyle of maintaining all that land (esp. w/ tendonitis + family health issues), did you take a financial loss selling the land? I could see that experience being frustrating.
2. Would you be financially independent regardless of your youtube income and other income (selling plans) from your website?
3D 'Tracing at Scale' with light
Apparatus is two rotationally-locked turntables and a can light (fairly directional), all mounted in a line. A model is placed on the center table, working material is placed on the table further from the light. The light and model are aligned so a shadow of model is cast on the working material. Remove everything that is illuminated, rotate, repeat, until you've created a copy of all (convex) curves of the model. Because the tables are rotationally locked, you can reliably reproduce the corresponding shadow for any rotation - can work your way around multiple times or easily come back to tricky angles. Distance between light and model can be adjusted to change scale resulting shadow.
Could be fun way to guide yourself with modeling additive materials like clay as well as woods and such.
Thank you for all the hard work.