Let's say that I'm building a sofa or hall table, 48" wide by 12" deep by 32" high. Make it Shaker to keep things simple (jointwise), with an apron all around joined by mortise and tenon. Assume the M&T to be the same all around the table.
Which of the joints will endure the most stress over time? Front and back stretchers, or side ones? Neglect for now that the table might end up against a wall (or sofa), which would reduce the stress on the side stretchers.
I'm curious as what you will say (I don't have a solid answer myself), please justify your answer if possible...
>Bearing in mind that the table will most likely be against a wall, and there probably won't be much movement... my thought would be that the problem would be front-to-back instability simply due to the legs being so close together.
>The stresses on any structure are dependent on how it's used. If you have a family of happy rambunctious, high-velocity teenagers who tend to bump into obstacles at a run and collapse into chairs then the materials and joinery of anything exposed to them should be selected accordingly. OHT if yours is a quiet family you can pursue a lighter, more delicate construction.
M&T joinery when made of strong materials glued together is surprisingly resistant to destruction unless cruely forced. A Shaker chair if swung at the head of a western movie hero will be far more likely to produce bust his skull than fly into slivers.
>My instincts tell me that, in a free-standing table such as you describe, the tenons on the front and rear rails will endure more stress. Applying a force to the front or rear of the table and it will be more likely to tip over than if the force is applied to either end and this alone would suggest that the tenons on the front and rear aprons will take more stress. This assumes that the tables feet are anchored against sliding. I have made a table a bit like you describe but six feet long and it is far more likely to simply slide on the hardwood floor generating little stress on the joints.
>Denis, It seems to me that if the joints are the same, the strength would be the same, but since the feet are different distances apart the forces on them will be very different. If the floor were "perfectly slick" there would be very little force on the joints, on a high friction surface like carpet the force could be huge. It seems that the force that would cause the joint to fail would be raking or turning the leg like a wrench while the rail is anchored. Sort of like holding the rail in one hand and the leg in the other and bringing them together (or apart) like when you use two wrenches on a nut and bolt.
In this case, since the feet are four feet apart one way and one foot the other, I cannot imagine a way you could apply as much force to the narrow direction without over turning the table as you could to the other. If, with the feet firmly anchored, you put your hands on the top, pushed downward, and left to right, you might be able to break the M and T joints, but if you pushed front to rear, the table would probably turn over before the joints failed.
This is a case for Vector Analyses, but I have forgotten anything I ever knew about that. I think the joints attached to the longer rails will take the most abuse while the table is standing just because it is harder to put stress on the shorter ones without turning the table over.
(In the real world what might happen is one of my son's stoned or drunken friends would knock the table over and then fall on it. All bets are off. There no Fat @$$ factor even in M&T joints.)
>I would definitely say the front to rear. I made a bunch of those tables over the years and because I was new to woodworking back then I used all sort of joinery. Mine were usually 38 to 48" long and 1 foot wide. I gave them to family members and as far as I know they are all still in use. One of them, that I think was assembled with 2 dowels per leg just went through a fire and subsequent cleaning of smoke damage by a professional company. It still looks great and is solid. It grew up with a one boy who is now in marine boot camp.
The Oak one I gave to my parents had floating tenons cut with a router. It's still as solid as day one and served many a party as the snack table.
>Let's replace that sofa table with a dining room table, let's say 3' x 5'. To isolate the problem, assume it's sitting on thick carpet (never mind the practicability of such a setup for the moment), so that the legs can't slide on the floor.
My guess would be that it will be the long rails that will bear the most stress, as the lever arm in that direction is longer. But I'd like your opinion on this...
>Denis, If the table were square and the lengths all the same, and the people bumping into the table were all the same, etc, the stresses would be equal. As the table becomes more rectangular, the stresses change.
If a person bumps into the table or leans on it from the end, the table would offer more resistance without over turning. The forces would be "resisted" by the joints, primarily the M&T joints in the long rails to the legs. Bumping it or leaning on it from the side would impart those stresses into the joints of the short rails to the legs, but there might be some differences due to over turning. IE, If the legs closer to you lifted off the floor they would not be under stress much at all, but the set that are still on the floor would have double or more stress in them.
If you just loaded the table heavily The long rails would tend to bend or deflect more and would probably stress the joints more assuming the feet are restrained by the carpet.
If the rails are securely attached to the top they might operate slightly differently and offer some resistance, but those joints really should be floating and their contribution would be negligible.
At least that's the way I see it until somebody 'splains it to me better.
>I am not a physics minor or major, and have never even played one on tv. But the angle at which the force is applied determines the amount of force needed to make the joint fail, everything else being equal. The angle of force would be a diagonal line from the bottom of one leg to the top of the other. As the angle increases towards 90� the amount of stress on the joint increases. The short apron would have a low angle of applied force, therefore would take more force to make the joint fail than the longer apron. Would love for some physics expert to tell me if I am wrong on this.
>As I recall my 101 class in engineering mechanics, the forces on the joint are expressed in 3 linear directions, x, y and z planes that are orthoganal (90 degrees to each other) plus a turning moment force (pushing the table when it is on a carpet).
When the table is loaded, a downward force is applied to the M&T joints along the z axis. If the object is placed in the center of the table, the downward force will be equal on all M&T joints. If the table is pushed sideways, a force is applied in the x and or y planes that results in a turning moment on the M&T joints that is proportional to the length of the table leg. Longer legs will exert greater force on the joint, but there should be no greater force on either the short rail joints or the long rail joints because the legs attaching those joints to the source of the twisting force ( the floor ) will be equal.
In actual practice, however, the joints on the longer rail will experience more stress, because we can usually push a narrow table along the axis of its long side. Pushing in the short axis direction will often tilt the table over before it slides on the floor. The tilting prevents the M&T joint from experiencing the stress equal to the long rail joint when pushed along that axis.
>For simplicity sake, lets say that your tenon is full height of the apron, and call it 2".
Each time you double the length of the lever side, the forces are doubled. So if a force of 100# is applied 2" down perpendicular to the leg, then the top edge of the M&T would feel 100#. Or for each 2" increment down the leg, just add another 100.
four inches down the same force would cause 200#, and eight inches down is 400#, 16" down get you up to 800#.
So, by the time you are up to the 32" height of your table, a 100# force on the end of the leg is ? to 1600 pounds of tension on the top side of the tenon, and the same in compression on the bottom of the tenon, or in shear on the sides of the tenon.
It is more complicated than that, but as you can see, the numbers start getting up there when you get the force down to the end of the legs. Stretchers just share the force on one leg with the others.
I agree that the tip-over of the short plane makes the long axis more important.
Since the tenons are competing for space in the top of legs, I usually try to give the tenon of the long rails greater consideration, so they will get the top and bottom portion of where they meet, and the short rails will use up the middle of that, and may be more-or-less a box-joint within the leg, if I feel it needs that.
Having said all of that, by adding height to the apron rails changes those numbers considerably fast. If the joint is 3" high, on a 30" tall table, the same 100# on the end is ? to 900 across the joint.
This was a good question for the forum. I hope it is more than just chewing gum for the mind, but this kind of analytical type of thinking can make better woodworkers out of whomever takes it seriously, and can be taken much further.
When you start looking at the properties of each species of wood, and their crush factor on the compression side, split along the grain, and rupture across the grain etc.
When the weakest of these starts to fail, then the zipper starts.