Sub-harmonic welding

Date 07-10-2008 20:43

Yes, I've heard of it. Not in context of stress relieving, but the process of setting up sub harmonic vibration for stress relief was first pinned down by a gentlemen that worked at Lockheed. (forget the exact division). Prior to that, attempts at stress relieving were made using harmonic resonance, but that had a nasty, (and still does) habit of causing damage to the work piece from time to time depending on a number of variables.

As I recall it was the lower 1/3 of the leading edge of the harmonic curve that turned out to be the solution. Some bright souls are currently working on that with phased array modeling to try and tie it to inline weld inspection. (Inspecting the weld as it's welded in an automatic fashion). The thought being that the sub/sonic/infra/ultra etc vibrations set up by the vibration/weld solidification could be used to detect the crack if any as it forms. Unlike acoustic emissions that typically requires an initial loading if i recall correctly, this method would not.
When I last visited this process, the problem was not theory, it was computational speed. There are millions of possible permutations depending on the part. To capture them all in a coherent image real time would take a phenomenal computational speed. I believe it's just a matter of time if not already before this hits the market.
Getting back to your question, yes I've seen it, and test results were positive in that it worked.

By darren

Date 07-12-2008 10:46

aren't they using sonic/subsonic resonance to homogenize parts for formula one /indycar/space shuttle etc....
actually i know they are just trying to open the thread to interpretation and inference.
absolutely no idea what sub harmonic welding means. in this context i believe sub to "what" must be explained. are they speaking of "chaotic/non chaotic impulses" to try and homogenize grain structure. there are many orders of harmonic (actually its definition). so sub harmonic can mean way too many things. this thread lacks a definitive direction but if it can be discerned i am absolutely interested.
darren
cwi555 from what you have stated we are looking for currently incomputable information group that would indicate real time reflection as to the quality of a weld based on sonic information/interpretation.
p.s. hows that mini sub coming along, i believe that was you with the titanium questions, and a query as to how to get the signal to your submersible ship without tether

By CWI555

Date 07-14-2008 12:17

Minni sub is at the bottom of the river, the fail safe died on me and it never surfaced. Two trial runs no problem, the third was a disaster. I know where it's at, it's lodged in the mud at the bottom. Just got to figure out a way to get it out.
(don't have diving certs, though for the cost of that thing, it would be worth it to get them)

As for formula one cars, I can't speak for them. As for the shuttle I recall that being true.
Your right about the many orders of harmonics, the harmonics of a wave is a component the signal frequency that is an multiple of the fundamental frequency for which all will be the same period at the signal frequency. That takes you into Fourier's in which the periodic function degrades into the sum of oscillating functions. (sines/cosines)
This is where welding/metals bridged into sound and vibration when fourier was working up the heat equation in plate.
to understand what they are after for this will take a pile of reading.
In a nutshell (not perfectly correct but down and dirty summarized) Fourier was trying to model a complicated heat source as a superposition. http://www.kettering.edu/~drussell/Demos/superposition/superposition.html
In the process of this, a multitude of discoveries were made, the applicable ones relating to vibration analysis which opens up several worm cans. In short, modeling the interactions real time, and modifying the various solidification phases in that the sub harmonics vibrates them into a relieved state real time.

I don't purport to understand it completely and will therefore insert that caveat, but I do understand direct results. I've seen it work. I'll leave it to the scientist in the group to explain it in better detail.
It's taken me several years of study to get to the point I am at, and still I am lacking. You will understand why when you start delving into the various intricacies of the problem. (hilbert space, Euclidean space, Sturm-liouville theory, partial differential equations, Paley wiener theory etc etc) I simply don't have time to get into like I need to for a full understanding and still work. For me I have enough to understand that it can in fact be done, that real time as welded inspections is a distinct possibility awaiting the proper computational speed and programming, and that there is more to this than is generally realized in the industry as a whole.

Regards,
Gerald