Re: This will blow your mind...
TomD
There are design options that will add to the stiffness of the top. I don't understand enough what is really being done here to know whether there is a problem. For instance how long is the unsupported span, and does he expect to sit on it. I made this low coffee trestle table with a tempered glass top. Looked good to me. And every kid in the neighbourhood had a go at sitting on it while the kids were young. That was not part of my brief.
So off the bat I would want to know whether the design issue is stiffness, or tensile strength, or both.
If you go with the metal bar part, I would use a strap, like 1/8x3/4 and make a 3/4" wide groove. You set it in with a few blocks to support it, you should be able to get a half inch of arc in it. The whole thing gets hardware bonded with epoxy. So what you get are 1) a much wide support for the load, in stiffness you want the supporting dimension to be out there as far as possible, circular section are great for being the best in all directions, but you only care about one direction in this case, so align your material with that; 2) you want the maximum offset in the metal, like diamond wires on a mast; 3) You want the whole thing bonded in so the load is transferred to the whole bar. A bar of metal is not that stiff in these sections, but if you bond the whole thing in, you create a composite truss. Another material that might work pretty well in this application is one of those pieces of pultruded glass fiber they sell in the fencing section of hardware stores, they bond well.
Simplest method is to make your top 1 inch thick, and taper the visible edge. That way you would pick up nearly twice the stiffness, and almost twice the strength, depending on how abrupt the taper is.
Complicated method, that would make sense if the design is really pushing the envelop, is to Vac bag veneer over carbon cloth. That would be very stiff. Seems unlikely to be needed in this case. But if the desk had only a support at one end, and was bolted to the floor, and people would stand on it to change light bulbs...
If you put say a wire,or cable in a groove and tension it, that can really increase the stiffness and strength, even with a slight offset.
One technique that is used in bows is to, in this case, make you outer board of the top 2 x 3/8" thick pieces that are natively curved their length, then laminate them in a mold, such that the 2 Cs nest, but they are flexed past straight to the point that when they come out of the mold and spring back, they are straight. Very difficult to one-off. But it will be very stiff in one direction...
-------------
So a few examples of the above are examples of pre-stressed structures. Did you know that the bicycle wheel is a pre-stressed structure, and that the loaded axle does not hang from the spoke above, but stands on the spoke below it? Yes, you are supported by the column under the wheel, not hanging from the wire above. Apparently you can prove this to yourself by plucking the musical notes of a bicycle wheel (a guitar tuner may help), or using a spoke tension tool to read the tension. If you were hanging from the spoke you would hear a higher pitch above the axle, but what you hear is that the only pitch that changes is the spoke(s) below. That becomes lower in pitch. Or so the story goes.
Mind blown? If nothing else it is a great way to start an argument with cyclists.
