should metal be scuffed or as flat/polished as possible for brazing/soldering?

[coppercone2]

i think you can do vacuum carburization though.

[IanB]

Well yeah but that's the thing, the vapor pressure of carbon is SFA until well into the white-hot range.  How could it be gaseous diffusion if the vapor concentration is ~nil?  Surely it should be evaporating out of the iron, if anything!

Yes, but diffusion doesn't have to be in the gas phase. Diffusion can happen in any phase, gas, liquid, or solid, where there is a concentration gradient. Isn't that how doping occurs in semiconductor manufacture? Aren't the dopants made to diffuse into the pure silicon to produce N- and P- type materials? The silicon is solid, but the atoms diffuse through the crystal lattice.

[T3sl4co1l]

Well, the point I was making is, it's not just carbon granules and iron, in the traditional method.  Some gas pressure is apparently required, preferably a bit of CO (or air, to burn into CO all the same).

Also, that carbonate salts probably aren't a suitable catalyst, but some kind or another of salt is effective, cyanides being the best known (AFAIK).  That gives a liquid phase option to carry that concentration gradient.

In semiconductor manufacture, when a solid-phase method is used, the first step is evenly coating the surface (e.g., uh... phosphate glass, spin-coated from solution??).  It's hard to do that with elemental carbon, even if you, say, blacken the surface directly (lampblack is still very porous).  Maybe decomposed petroleum pitch would do -- that's the stuff used to stick together artificial graphite, and it can decompose to an adherent layer of amorphous carbon.  Hmm, I wonder if that would work?  Maybe it's too thick of a layer, too hard to control?  Maybe it cracks off before it can form a diffusion bond and eventually dissolve?

Vacuum carburization is a bit of a misnomer of course, as it's actually a low pressure gas (the vacuum is relative, as a "vacuum arc furnace" is as well).  Looks like they use methane and such, which makes sense -- under various conditions (including surface temperature and preparation (catalyst), gas mixture, ionization and more), methane decomposes to form various allotropes of carbon.  Any of which, when formed directly on a hot iron surface, would dissolve readily.

Tim

[coppercone2]

are you sure? i think methane would foul a kiln .

Do they use the kind with sealed elements or something?

Won't that carburate the elements? Or do they have a inner shell?

[coppercone2]

I want to try to boron-diffuse metal. I have some boron. I wonder if you can just do it ghetto by smearing boron on something and putting it on a BBQ burner for a while. No kilns yet. Wrap in copper sheet. It's boron in oil, maybe you can just paint it on and heat it red. Should get carbon from the oil. 

[SparkyFX]

Would polishing lead to a work hardening of the surface or alter the grain structure of e.g. steel at the surface?
I can imagine both might influence its ability to be brazed (coming from a diffusion layer approach to the problem), so the suggestion to try brazing gauge blocks could provide usable results, otoh gauge blocks are usually hardened and therefore brittle. A test would need same hardness and tensile strength base material.

[T3sl4co1l]

Depends on the polishing process.  Lapping is a very low-impact process as far as I know, so if there is any microstructure change, it would be to less depth than the diameter of the particles used for the final polishing step.  So, fractional microns most likely.

Also, as it's already hard, as you note, there wouldn't be much room left for work hardening.

It's not a bad observation though -- in fact it's very relevant sometimes.  I've seen copper sheet physically curl up as it's stressed from glass bead blasting -- it's a much more aggressive process than most would think!

Brazing temperature is around where steel's microstructure itself relaxes (annealing, normalizing, austenizing, whatever it's called in the case), so you wouldn't expect to have hard parts when you're done, at least in the average case with common alloys and no special treatment.

You can quench harden steel after it's been brass brazed, as it freezes at a higher temperature than heat treating requires.  Of course, this is a bit harder to do with a filler that melts below the austenite transformation temperature!

Tim

[coppercone2]

I thought your not supposed to cool brazes because it can fracture?

Do you mean heat treating tbe entire part after it cools naturally, not like squirting water on it after you done?

Also to be more technical lapping and reaming are scraping prkcesses. I think grinding makes alot more heat. As far as working metal goes lapping is super cool because its slow, high thermal surface area and coolant between your work piece and the lapping plate.

I think processes that use machine rotation are inherently more heating because you get a particle with kinetic energy of wheel striking the plate like a meteorite.

I find what you say interesting because of wood. If you compare precision planed wood (which is a scraping process) its actually hydrophobic. If you sand it its hydrophillic.

How ever i am pretty sure carburized parts grow at least a surface oxide layer, so any tiny amounts of work hardening should turn to rust?

[coppercone2]

Can someone explain work hardening with a diagram showing as few atoms as possible?

[T3sl4co1l]

Yeah, so it's not suitable for all brazes -- many alloys are indeed quite hard (silicon bronze? nickel boron?).

It can be done with brass braze, which is soft enough not to care.  After brazing, either cool it and do whatever finishing steps you need (remove slag, grind to near-net shape?), then heat to austenite temperature (coincidentally about the Curie temperature as well) and quench.  Check hardness, then temper, then grind to final shape and finish.

Tim

[coppercone2]

does metal ever work harden from any metal cutting processes like milling?

[David Hess]

does metal ever work harden from any metal cutting processes like milling?

Stainless steel sure does when you drill it necessitating cobalt steel drills but maybe that is due to the temperature rise?  Softer metals might but you would never notice because the tool bit is so much harder.

[T3sl4co1l]

Yes, the surface is work-hardened under a cutter, in the shear zone where metal is pushed aside.  The depth of this layer corresponds to material properties, depth of cut, and geometry of the cutter (worse for a low rake angle).  The amount of work hardening depends on the material.  (Copper and stainless are examples that go from fairly soft to surprisingly hard with not much shear, making them difficult to machine.)

Doesn't have to do with temperature, as far as I know.  If anything, that would make things easier?  Which brings to mind, turning hardened steel with a carbide tool at high speed.  Seeing glowing hot, stringy chips flying off is an awesome sight...

Tim

[coppercone2]

does it ever harden in a axis that is not the direction of material removal?

Like drilling makes sense. But if your just cutting or grinding to the side, does force get transferred underneath? I thought it goes into tearing/stretching.
'
With something fast. I feel like a lathe is more likely to do it then a fast milling bit that shoots chips off?

[T3sl4co1l]

Shear at the surface.

So, the wall of a drilled hole is worked, the side and bottom of a milled slot is worked, the side and face of a turned cut is worked, and same for grinding, sandblasting and so on.

Tim

[coppercone2]

so can you imagine it as 'necking' like when you put a steel sample into a yield test machine? That each 'instance' of the blade cutting through the material when a curl/chip is formed, you get the metal in the yield mode directly near the cutting edge of the tool, so when the tool cuts it, its like a spring that unyields and the metal actually recoils back downwards and hardens as it impacts? Or is it not really recoiling back down, and its just yielding and it gets hard as this happens?

Can you almost imagine it like directly above the cutting surface, its like putting a rubber band between your fingers and smacking it down every time it tears? But its a continuous time process some how (weird to think about) since its rotating)?

If you had a powerful pulsed laser you could clean it up from the mechanically deformed stuff after if you blow up the surface atoms layer? Or do you get some kind of laser peening too from the vapor explosions that happen (more like sand blasting then the yield tearing)?


Maybe you just need to dissolve it in acid or base? That should be gentle.


I have trouble because with sandblasting you get a downward impact like beating it with a hammer, With the machine process it seems like your just i dunno, like pulling it out. Does the metal get hardened the same way? or is there a counter force from spring like I said? In the yield test machines you can see the shockwave after it tears so you might imagine it hits it just like a hammer.

lasers out

[eKretz]

Any metal that will work harden with cold working (like hammering) can work harden when machined - with any type of cutting operation, whether drilling,  milling,  turning, whatever. To avoid this, it's important to keep cutting edges SHARP so that they don't rub. If a dull cutting tool is used on something like the right grade of stainless steel and it does more rubbing than cutting, work hardening can definitely occur, even to the point of a HSS cutting tool being unable to further cut the material.

[David Hess]

So can you imagine it as 'necking' like when you put a steel sample into a yield test machine?

Absolutely, it is still strain hardening where the metal is pushed past its yield point.

If a dull cutting tool is used on something like the right grade of stainless steel and it does more rubbing than cutting, work hardening can definitely occur, even to the point of a HSS cutting tool being unable to further cut the material.

I have seen this happen many times with inexperienced operators on drill presses.  The cutting tool needs to be kept cool to prevent loss of hardness which means low pressures and low cutting rates.

[coppercone2]

but is the type of hardening different for downwards drilling and the resulting hardening from a surface pass?

With the drilling I imagine the force being heavy and downwards. With the surface pass I just imagine it stretching out.

When it 'tears' is there a downwards shock wave when it actually breaks? On the necking machines you see that it gets long and then it snaps back. Can you imagine that during milling, you get a continuous 'shock wave' happening as the metal tears? Or is the mechanism different?

Does anyone have a ULTRA slow motion video of a necking test? I can't find what I am thinking about. Maybe I saw a weird material in SEM. Does it just get super dampened?

[eKretz]

The hardening is work hardening.  It doesn't matter what operation is being performed. It isn't about keeping the cutting operation slow or cool, it's about keeping the tool sharp so it shears the material. Sometimes this requires slowing the cutting tool or keeping it cooler so it doesn't dull, but it's the dulling that causes the problem. A dull tool pushes the material around more rather than cutting it cleanly, effectively cold working it. When a milling cutter dulls, it pushes against the bottom surface as well as the side surface.

[T3sl4co1l]

No, no shock waves.  There may be some cracking of the surface as the material tears (in tension) as well as shearing of the chip from the base material.

There are micrograph videos of cutting tools in action, check those out.

Actually, there may be shock waves, in certain materials.  Zinc and tin come to mind, as materials that can exhibit supersonic slippage.  In effect, you're "cracking" a crystal along a plane, and it slips very suddenly as the tension is relieved.  The shock is limited to the width of a given crystal.  In bulk, this manifests as "tin cry", a crackling sound emitted as the material is bent.

The way that deformation works, is slippage planes in the crystal lattice.  The reason work hardening occurs, is those planes do not slip evenly, creating dislocations.  The paths of dislocations (like line faults) flow through the crystal in 3D; if two paths slip into each other, they can annihilate, or they can entangle.  The general accumulation of overlapping defects is what gives rise to work hardening.  Eventually, so many defects accumulate that the material tears apart (it is too congested to slip in an orderly manner anymore), and a crack forms.

What distinguishes ductile metals from brittle substances, is the number of coordination sites per atom, and the electron density.  Ionic crystals are brittle because they have relatively low coordination; when a slip occurs, the planes rarely line up correctly, and a crack forms almost immediately.  Metallic crystals have many sites so that, even if the planes no longer line up properly, the atoms can still share electron orbitals, and maintain a bond.

I think.

You can still have ductile motion in typically brittle substances, but it's much rarer to see, most often on the microscopic scale where more perfect motion is possible.  Example, the marks left from grinding and polishing.  Sometimes a smooth track is left, implying a fairly smooth cutting process; other times, it's a jagged track where material was effectively chipped away from the base.  I would guess the relative amount of each type depends on the angle and pressure of the cutter, and the material properties.  In a typical ground face, the cutting angles are random (angular bits of grit), so the relative mixture of both speaks to the properties of the base material.

Tim

[coppercone2]

Surface roughness reduces the effective contact angle causing more spreading and scratch mark or “lines” act as capillaries causing directional spreading in one direction but also act as dams reducing the spread or flow of alloy in the perpendicular direction.

http://www.altairusa.com/wetting-braze-flow/

almost wonder if you are doing a deep lap joint, can a specific scratch pattern be used to flow solder correctly.  like if a carbide micro engrave CNC scratch thingy could be used

would be difficult to figure out the correct dimensions though.. maybe very sharp sand paper (SC based?) of a particular grit would work

[SparkyFX]

The graining of metal to increase shear resistance is quite common, some establish a rasp like structure before e.g. glueing. It is hard to tell when effects go from the micro- to the macroscopic scale, such as work hardening.

The point of brazing is mostly to connect two dissimilar metals, with the braze often the toughest spot. So it becomes the question which part develops the highest shear strength via surface before the braze fails. If a braze fails right at the surface of a part people would assume the surface was not clean enough, hence would need better preparation (=sanding).

From an economic perspective any sanding/polishing step brings new problems, like additional cleaning to get the polishing/abrasive compound off and out of the surfaces (it shouldn't contaminate the brazing, but might have been embedded in the material).

So it might be a trade off between removing impurities before brazing and just brazing. The flux does also play a significant role anyway, etching the surface a bit or at least preventing oxides from forming while heat is applied.

[KL27x]

almost wonder if you are doing a deep lap joint, can a specific scratch pattern be used to flow solder correctly.  like if a carbide micro engrave CNC scratch thingy could be used

I think it would be more useful to have ridges (thin lands, wide grooves) than scratches/grooves (wide lands, thin grooves). Ridges a few thous tall would prevent (larger areas of) tight spots that close up the gap between materials.

The flux does also play a significant role anyway, etching the surface a bit or at least preventing oxides from forming while heat is applied.

Agree. I think in many or most brazing operations, any fine sanding/polishing is moot, because the flux (+ heat + atmopheric oxygen) is going to etch/obliterate it by the time the joint flows. Acid etching increases the surface area on a microscopic level to a much greater degree than mechanical sanding. Etching leaves a delicate "fluffy" surface. The dimensions of the material can actually initially grow a bit rather than reduce, cuz of the fluff. This etched surface sucks the braze material into the surface of the metal.

In fluxless TIG brazing, you don't get any capillary action. The gas prevents oxidation, but there's no etching. And in TIG brazing, you just lay the bead/filet across the two materials to be joined. You don't set a gap to try to get the filler to flow between materials, cuz it's not really gonna happen. That is partially due to the focused/localized heat source, perhaps. But the TIG torch can penetrate deep on a weld, so heat isn't necessarily the limiting factor. Just imagine trying to solder without any flux.

You might think that the flux/acid etching is a significant component to capillary/wetting action. All this scratching/polishing idea is perhaps nonsense, because it can't produce the type of surface that acid etching does.