weakening a cantilever beam in one direction 9but not the o

huangyhg

weakening a cantilever beam in one direction (but not the ot
i have a unique application that essentially boils down to a cantilevered beam with a load acting perpendicular to the end of the beam.  the unique part of the problem is this: we want the beam to be stout with the load acting in one direction (let鈥檚 say downward), and we want it to fail with the load acting in the other direction (let鈥檚 say upward).  the cantilever is made of steel and is rectangular in cross section.
our concept is this.  we鈥檙e thinking of making a thin cut (using a 0.001鈥?thick edm wire) near the base of the cantilever.  the cut would be perpendicular to the beam and would start at the bottom and move towards the center of the beam.  at the end of the cut we would make a circle with the wire to reduce stress concentrations.  so, looking from the side, the cut would look kind of like a lollipop.  if necessary (and feasible), the thin cut would be filled with shimstock.
with the load acting downward, the part of the beam that is cut will be in compression, so our hope is that the effects of the cut being there will be minimal.  this, of course, implies that the cut must be well on the compression side of the neutral axis.  with the load acting upward, the part of the beam with the cut is in tension, and we essentially have a thinner beam.  also, we have a nice stress concentration at the end of the cut that will help with failure when the load is acting upward.
one complication is that the load is acting dynamically (think of hitting the end of the cantilever with a sledgehammer).  that means there will be shock waves and rarefaction waves running around.
what are your thoughts on this concept?  its simplicity is driving me nuts.  i can鈥檛 help but think there鈥檚 a 鈥済otcha鈥?somewhere!
thanks,
haf
note: this question is also posted in the "mechanical engineering other topics" forum.
my first impression is that you've covered the fatigue aspect of it with the crack arresting hole.  and using the shimstock seems reasonable too.  i guess that you intend to stitch weld the shim plate to the tube in order to hold it in place.
if you've got the time, build one and test it.  the test tube can be smaller so that emulating the fixed connection at the base won't be too difficult.  and sledgehammers they're always around somewhere where iron is.
you should fill the cut with metal material, else the endgrain can embed in the endgrain on the opposite side.  sounds like a plan...
why not use a mechanical hinge with pins the will shear one your reached the desired load?
this is very simple and easy to compute and fabricate. there are so many pins with different strength.
good luck
sorry to say the obvious but rcc cantilever beam is inherently so...
thanks for the input.
we are planning on testing this concept.  unfortunately, the test is extremely expensive (well over $100k usd) and high visibility.  there are no inexpensive tests we can perform to accurately achieve the environment.
unfortunately, hinges, rope supports, support brackets, etc. are not possible with our design constraints.  likewise, relatively complicated cross sections (t-shape, h-shape) are not an option.
let me be more specific with my question.  let's say we have a 6 inch thick cantilever beam.  we make the edm cut 2.75" long, so that a 3.25" thickness is left.  will that beam be stronger than say a 5" thick beam with no cut?
haf