Some feedbck on the strength and rigidity of the table. This follows a dinner party last night, where guests danced on the table top, used the edges to do push ups, and then proceeded to lean on their elbows while eating.
Okay, so only the last bit is true.
My aim was to build a table with slender proportions. In part, the elliptical shaping created the illusion of slimless, but the dimensions were not over-generous to begin: in the main, mortise-and-tenon joinery 13mm (1/2") thick and 35mm (1 3/8") deep x 70mm (2 3/4") wide in 100mm (4") x 30mm (1 3/16") posts. Most of the (especially USA-made trestle tables) use sections about double these dimensions! I was clearly asking for trouble.
The wood used is in the base is Jarrah and Rock Maple for the top. Jarrah is hard and stiff, but enough? The Maple top looks slim in the photos, but is 30mm thick with edges chamfered to a 10mm edge. That 10mm edge is a repeated theme (also present at the ends of the elliptical stretchers), and comes up again in a short while.
Now there is a part of the build which I have not mentioned. After assembly, there was a slight amount of movement which may be described as the top twisting. It was slight, far less than the trestle table on the patio which I built 25 years ago (which often seats 10-12), dismissed by my wife, but unacceptible to me in my hyper-sensitive state. What I realised was that the width of the uprights was sufficient to prevent downward movement when pressing on the table top edge between the two uprights, but pressing down at the corners forced a slight rotation, and this was experienced as the top twisting. Consequently, I decided to reinforce the top with bracing.
Now here's the dilemma - I did not want to ruin the slimless of the construction with braces which contrasted with the existing lines. I registered what many have done, such as adding angled sections or a rail between the two uprights. I could have added a second elliptical cross rail. I wanted none of that - the cracing needed to be invisible.
Cross bracing: this is in common use in building construction and is strong as it utilises triangulation in its design. Here is what I did ...
The sections are 10mm thick and 50mm (2") wide. The 10mm thickness was chosen to align with the edge of the upper cross supports.
The ends were dovetailed and glued ...
The cross centre is a lapped joint.
The final construction sits neatly under the table top, and is also screwed at the centre ...
The result is completely invisible from the front of the table. The top of the table is now very rigid and without any appreciable movement at all. Not the slightest!
Regards from Perth
Derek