Deep scratches? Those you see as horizontal dark places are probably about few micrometers deep. The marks in the top-right corner are from a needle.These are not exactly stress tests, but micro-imperfections tests. The dark/white areas are 'bumps' on the contraction surface. This 'alloy' (for the lack of better word) was made by blasting copper and alluminium into eachother. So there is a place on the border between materials, where they are slighltly difused (from 'difusion') - a thermodynamics/quantum physics effect.

AHHH!! thanks for clearing that upi looked up diffusion...and i gotThe scattering of incident light by reflection from a rough surface. The transmission of light through a translucent material. The spontaneous intermingling of the particles of two or more substances as a result of random thermal motion <-- these dont seem like the most accurate but i think i understandso you are testing the "alloy" by looking for the extremely tiny imperfections and to discern from those imperfections how long the delamination process takes? or how much force is needed? or how to speed up/ slow down that processwith something so small i wonder how you almost messed up the microscope lense :roll: :wink:

Diffusion is very time-expensive effect. It takes very long to actualy observe any changes in material, unless You heat it. The higher the temperature, the faster and with higher probability the particles intermingle. If Youre up to mild quantum physics, i could clear it even more. So far we are testing it just to see what usefull things can be made out from the research. If we stumble on something new, thats bonus for us. The lens is very close to the sample (.5 cm - focal point). To get the best, and the clearest results, we have to set the height to the first maximum of Rayleigh wave. What we did week ago, we set it to the 5th or 6th maximum. That meant the lens was much closer to the sample, than it should be. And when we did V(z) [V - amplitude, z - height] test, the lens banged on the sample.

:shock: :shock: :shock: Very lucky you didnt damage it then!!!and yes i am up for some mild quantum physics

Okay, i will try to make it as painless as possible (which means 0 equations ).Every particle creates in its vicinity a 'barier of potential'. This potential has its specific height, which defines the energy and the effect it creates. Now in metal we have particles crystalized into their positions very tight. But they have some freedom of movement (3 degrees of freedom). They oscilate around their 0-points. Even in 0K they would move (its called 0-energy oscillations), becouse if they wouldnt, we would be able to see exactly WHERE they are andwhat momentum they have (0), and thats impossible, from the Heisenberg Theory (position differential * momentum differential >= h/4PI). Returning to the point, You have to know that the situation, on which particles of two different types are crystalized into the same bpdy sepparately (there is a border between them), is not stable from the view of thermodynamics. Stable situation is when the Gibbs Potential (Gibbs Free Energy) is minimized (the chemical potentials are equal), in other words, when the particles are places in their natural (from crystalization point of view) positions. These positions can be really odd sometimes. But generally the crystalline base (the smallest part of the crystal, from which can we build the whole crystal) is something of few (up to hunderds) particles of both kinds spaced in some shape. I see this gets complicated so heres the 1D example:O - particles of 1st kind0 - particles of 2nd kindInitial positions are such:O O O O O O O O O O O O O O 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0Now, after diffusion we would get something like this:O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 O 0 So the base in this crystal is composed of two particles: O 0 . When we repeat this base over and over, we get whole crystal.Whats the reason to say all this? To show, that particles 'long' to form a uniform (thermodynamicly) crystal. Thats natural. But to do that, they have to conquer the bareer of potential first. So now we know they oscilalte. These oscilations have some energy. Sometimes this energy is enough to 'jump' above the top of barier potential into new place:|O O| 0 |O O==>|O O| 0 |O OBut if they dont, you have to add their energy. And the simpliest way to do that is to heat the material. When we add energy, the particles oscillations have higher amplitudes (energy), which means, its easier for them to 'leap'. Thus, very slowly we get the diffusion. In metal crystals, the range of this effect is short. The crystals are too 'heavily guarded' by potential bariers. What we get is VERY thin layer, where the particles are intermingled. Phew...PS Thank heavens for "solid state physics". Thanks to that, i didnt have to go deep into QP (i dont really want that - it is hard to explain it all in ASCII - range, i would have to copy a book here). All i needed to do was to remember the past two sessions.

i c, very intresting....What exactly are you studying?Becouse i dont want to get a question regarding this on my exame....

Im studying Applicational Physics. More precicely i will be choosing specialization this semester, and i think it will be 'solid state physics'. Part of that is what i wrote above.

ok that was a good explanation i understand what your doing (but thanks for keeping it very basic ) keep me posted on what you find thats interesting in the patterns could you simulate the diffusion process on a computer? and accelerate it to get results faster?i guess not...justa thought thoughand is heat the only catalyst that speeds up the process?

Yes, You can simulate that. But You have to forget or simplify some processes. But it can be done with high probability of getting the right result in the process (mostly the shape of potential bariers). For example: You would need to have initial pattern, Gibbs Function algorythm (temperature related), diffusion energy algorythm (temperature related). Doing it numericly wouldnt be hard, but the results could be difficult to understand (i.e. which particle ends where after given time). Doing it graphically.... it could be problem to me, becouse i am not VG programmer.Is heat the only thing to speed up the process? Well, what i meant was to give energy. Energy, which can be translated into higher oscilations. I dont believe heat is the only thing that COULD do that. Some quantum effects could have enough energy to do that, although i think those cases are very rare (1 in a couple of billions, trillions even :? ). In this example the energy needed to complete the task is quite high, so the heat IS the most effective way to speed up the process.Heres the diffusion experiment You can do at home, if You have bathtub. First open the Heat valve and wait, till the tub fills to the half. Then close the Heat and open Cold valve, but VERY slowly (not that its barely sipping, but the water is coming slowly). Wait at least until the bath fills itself. Sometimes it would require to leave the valve open for half an hour. Anyway, if You do it right, then when you put a hand inside the tub, You should feel hot on the top and cold on the bottom. There will be two layers of water particles in the bath - of different energy. Between the layers there will be border, where particles will exchange energy - the border of diffusion. The process will continue, until the temperatures will equalize. In this example the diffusion-driving process is the difference of energies between particles of the same kind. Heating will speed up the process (eventually), but the results will be not what we want to get (the water will boil). In this case the only way to speed up the process is to stir the water.