"I'd ask that you look at the size of the holes allowed as you get closer to the middle. They increase usually by 1" diameter for every foot you go away from the ends."
Exactly - the amount of material you would be allowed to remove would increase toward the middle, and decrease toward the supports. So as not to dramaticaly compromise the shear value requirements that would be higher toward the ends. I have talked to alot of people about this... (In fact I ask every Structural Engineer I come in contact with the same line of questions, as I am facinated by the slight differences in opinion.) The idea that there is a zero area between the top and bottom areas that does little work is being dispelled by many people as "Olde School", all removal of material is a reduction to some degree. Imagine these I beam types as a trelessed bridge, with many cross members between the top and bottom. Imagine your hole is where two of them cross. Your hole is removing both of them. You are actually removing and reducing four areas of web tension from the top across to the bottom.
Some of the differences in engineering opinion that I find funny are with dimensional lumber and Glue/micro lams. As tension and compression are not consentrated in the top and bottom edges. Tension and compression values are more spread out in the body of the dimensions of the material. And both also have web tension values, and shear zone areas. The differences in engineering opinion that I find funny is how that web is formed, and the importance of holes in the inside 3rd as opposed to the outside 3rd's. That whole "3rd's rule" that has been drilled around for so long.
Here are the differences:
"Olde school", imagine the most important web tension from the top of one supported side to the bottom of the other side, so the web is one big X across the whole span, making large holes in the middle 3rd a tabboo.
"New skool" sees all spanning structers in a simular light, with many smallers X's from top to bottom, but with more importance toward the edges, top, sides (shear) and especially the bottom. (like that drawing I made) The reasoning for this is that they see wooden structure (Dimensional lumber and glue lam types) as being more dynamic, as the bonds of the grain of wood do not actually stretch from one corner to the other. But more closely together.
A compromise to this would be to split the span in fifths and avoid the two outside shear areas, and the center fifth. Like I said, I have an engineer friend, who likes to BBQ... Many drunken nights of talking shop.
Alan, very true, cantilevers and such reverse everything to some wildly varying degree. Tension and compression make wide sweeping arcs over the support and move shear area toward the center of spans. Several hundred years of engineering can not be expained simply... By me, (Not an Engineer) or any one else.
[This message has been edited by e57 (edited 01-08-2006).]