Jump to content
Harper John

Structure Of Martensite

Recommended Posts

Hi jhobson,

 

btw, dendrites are formed from solidification process. .

 

 

OK. So if we harden a cast piece of steel, out of he mould, then the martensite plates will not form across a boundary because of different crystal orientations (on each side of the boundary).

 

Grains can be formed from other processes.

 

As for example when we change the grain size by thermal cycling. Now a new question arises in my mind: Do these new grains have necessarily differing crystal orientations? Or as the larger grains change into bigger or smaller ones, is the same crystal orientation retained? What's driving the formation of the new grain boundaries?

 

If in the above we can have adjacent grains with the same lattice orientation, then would it be possible for a martensite plate to jump across a boundary?

 

I don't know if it's my browser, but couldn't make that link work.

 

Cheers

John

Edited by Harper John

Share this post


Link to post
Share on other sites
I don't know if it's my browser, but couldn't make that link work.

It is a microsoft powerpoint file. If you don't have powerpoint or equivalent then you won't be able to view it :(.

If you were really keen, then you could download a free equivalent from openoffice.org, but a lot of bother just for one picture.

However, the site the link comes from does have other resources available.

There are a number of videos of the actual cambridge grad course which I find hard to follow, and one of them seems corrupt to me, so I miss half the lecture and I gave up at that point. My conclusion is that you would gain little, if any, benefit to knife making by watching them. But, if you want to take the next step in understanding i.e why the crystals undergo a phase change, and how the change propagates, then I think you need a good understanding of the energies involved and these lectures are a good introduction.

There are also a couple of other lectures given by Harry Bhadeshia to other groups (e.g a vist he made to Australia) and these are much more accessible. Still, not a huge benefit to knife makers, but generally interesting. e.g he has created a steel that transforms to bainite at room temp (no salt pots required!). Trouble is, it takes a hundred years :( Or, did you know that iron has another another crystal structure? Difficult to measure it's properties as it is only stable below 800C but at thousands of atmospheres of pressure.

 

 

Dendrites is the term used to describe crystals as they solidify. The process results in direction of crystal lattice, and separation of impruities/alloy elements. These can be obvious in fresh wootz or meteoric iron, but the main influence that have on processed steel (i.e. a finished knife) is the separation which can be sometimes seen as alloy banding or the woots pattern. Big alloy elements don't move much during normal processing.

Share this post


Link to post
Share on other sites

Hi jhobson,

 

 

 

 

OK. So if we harden a cast piece of steel, out of he mould, then the martensite plates will not form across a boundary because of different crystal orientations (on each side of the boundary).

 

 

 

As for example when we change the grain size by thermal cycling. Now a new question arises in my mind: Do these new grains have necessarily differing crystal orientations? Or as the larger grains change into bigger or smaller ones, is the same crystal orientation retained? What's driving the formation of the new grain boundaries?

 

If in the above we can have adjacent grains with the same lattice orientation, then would it be possible for a martensite plate to jump across a boundary?

 

I don't know if it's my browser, but couldn't make that link work.

 

Cheers

John

 

 

Hmmm, not quite. Dendritic growth from a liquid is a very different way for crystals to from but does have one or two similarities to changes insolid phase causing crystals to grow through solid state reactions. Solidification structure has some relevance but is a whole new topic, and for this discussion it is perhaps best assumed that in the steels suitable for exploring the current discussion, it is sufficient to assume that the coarse and segregated as-cast dendritic structure of steels produced by liquid metal metallurgy (as opposed to powder etc) is carefully homogenized before and during hot rolling to make the steel suitable for use. There should be little enough "dendriticness" in these steels to ignore it in this discussion - at least for now.

 

The short answer to your question about the grains necessarily having different orientations is yes. Each crystal will grow from its seed crystal in the orientation it was seeded in(this starts at the first few atoms arranged in a particular FCC, BCC, BCT or CPH lattice - i.e. less than a handful or atoms already having a crystal orientation).

 

Grain boundaries arise where there is a buildup of dislocations as two misaligned crystals "clash" and the energy required to alleviate the dislocations or slightly re-align the molecules of one crystal to match that of the adjacent one is not available (i.e. grains won't "grow" or combine to form bigger ones when the metal is too cold). If two perfectly matched crystals grow into each other a perfect alignment would mean no grain boundary will form and they will just become one crystal. Grain boundaries are classified and differ in energy according to how misaligned the crystals on either side are.

 

Because of the misalignment, built up energy and "space to move atoms around" due to high dislocation density at grain boundaries, the formation of new crystals during for instance a low driving force phase transformation ( austenite from pearlite) or precipitation (cementite in hyper eutectoid precipitation) occur at grain boundaries more easily.

 

So more grain boundaries = more places for easier re-arranging or atoms for whatever wants to happen to start happening.

 

The martensite reaction has a very very high driving force and thus does not really need a grain boundary to initiate twinning or needle growth (which happens at speeds in the vicinity of the speed of sound), but it is none the less influenced by grain boundaries. In equilibrium driven reactions and transformations, whatever is easiest will happen. So crystals always form where it is easiest for them to nucleate and grow. Growing through a grain boundary takes energy so when a martensite needle "can choose", it will take up an open space inside an austenite grain rather than go through a boundary. By the time all the open spaces are taken, much of the driving force has been spent on transformation, and it becomes even less likely for needles to go through grain boundaries.

 

So even though the exact effect (especially on physical properties) is complex, the influence of austenite grain size on martensite properties is in part based on the above principles.

Edited by Bertie le Roux

Share this post


Link to post
Share on other sites

. Now, squash and pull these BCC cubes until they are true cubes - equal in all 3 axis, .

 

I've not really understood why BCC is less dense even though I 'lucked' into the explanation with the above post. I found the unit cell sizes of BCC anf FCC (both at 912C) here: http://www.cmse.ed.ac.uk/MSE2/IronStructure.pdf

 

A simple bit of Pythagoras shows that in going from FCC to BCC the FCC unit contracts in one plane but expands in the other 2, resulting in a less dense BCC but the space in between atoms for Carbon to dissolve is now squashed so although more volume, it isn't tall enough. And a single Carbon atom can't squash.

I drew it out on a bit of paper...

IMG_20150118_0001.jpg

Share this post


Link to post
Share on other sites

The reason behind the relative densities of the face and body centred cubic structures is based on the different ways regular, perfect spheres can be fitted into a perfect cubic space. For the sake of metallurgical chemistry, the lattice nodes where the nuclei of atoms rest are represented as one of those idealized spheres. For any orientation, the theoretical maximum volume:empty space ratio mathematically achievable is pi*(3)^(.5)/2 which is approximately .740 and coincidentally happens to be the way face centred cubic lattices are packed. Body centred are more space-inhabiting with a ratio of about .68 (see sphere-packing). Bear in mind that, while the size of carbon and iron atoms do not change, and the relative amount of volume for the voids does not change with size of the spheres packed into a space, the orientation of the spheres in the cube will change the size of the voids. Because carbon is dissolved into the lattice interstitially, and carbon is incompressible, the only way to make it more spatially efficient is to change how the interstitial voids are sized. As a result, carbon can be more easily consumed in the face centred lattice than the body centred (which is represented by the lower carbon content ferrite (.021% max C) and the higher carbon Austenite (~2%) lattices). The lattice bonds between atoms are forced farther apart in the body centred lattice, which also makes it less dense than the face centred. As you said, carbon doesn't squash.

 

John

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...