Chromium carbide overlay microstructure

[bekks]

I tried to attach two micro photos, but I think only one is being attached. There is no scale bar, but it is taken at 100X magnification with Vilella's etchant. This micro is taken at the mid-thickness of the second layer of a CCO. I've circled two regions in yellow that I am curious about. I am questioning whether they are primary carbides as they are very ordered with a triangular overall structure. I see this morphology somewhat regularly and I don't have a confident explanation for it. I more frequently see this morphology near the interface of layers, so it is presumably affected by cooling rate and therefore I assume it is caused by something along the lines of increased nucleation with lesser growth time. Is anyone able to confirm that these are primary carbides? The implication is that we are doing an image analysis to count the percentage of primary carbides, so this structure gets included in the count. From this image we apply a threshold to determine what gets counted, so the grey eutectic matrix is omitted.

 

[MagBen]

what material? processed history?

 

[bekks]

Chromium carbide welded with mechanized FCAW. I don't have the welding parameters, but it is an an as-welded production test, so no service. Chemistry of the deposit is around 30% Cr and 5% C.

 

[MagBen]

Looks eutectic carbide to me, eutectic phase often shows some kinds of "ordering" structure. The matrix is martensite/ferrite/remained austenite depending on cool rate, so this hypereutectic CCO consists of three components. you could run an XRD to confirm.

There are big pencil-like primary Cr7C3 phases that can make the layer cracking prone. do you have micros at the interface (with base metal) to share?

 

[bekks]

Thanks MagBen,
I've attached another photo (still 100X mag) of the interface between the first and second layer. The ordered phase appears to grow in size as you move from the interface to the surface. This growth is what leads me to think that it may be primary M7C3 (or another composition?). I agree that an EDS scan would solve the question fast, but no one above me is interested to know the answer.

The first layer appears to be all dendritic austenite. There are cracks, but they don't penetrate into the base metal.

 

[EdStainless]

If you leave these out of your primary carbide count how does in impact the numbers?
They sure won't have the same contribution to wear resistance as the larger carbides.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[bekks]

This is getting to the root of my question Ed. Our image analysis (via software counting pixels in a two tone image) cannot differentiate this "phase" from M7C3s. A typical image analysis count is 30 - 45% "white" and I estimate that this ordered phase accounts for up to 10% of that white count, so 3 - 4.5% of the total. There is also a wear resistance test that is conducted and, as would be expected, a higher percentage count usually correlates to a better wear resistance test. Client specs are usually cut and dry regarding acceptable criteria, however, it seems the fabricators often have some room to argue the acceptability of their product if they don't meet the client's spec. I am working as a third party for the testing side of things.

 

[EdStainless]

Can you also set a size cutoff for the phase count along with the contrast?
The odd thing is that depending on the kind of wear the small carbides may be superior to the large ones.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[bekks]

Good suggestion. I'll look into an option for a size criteria in the software.

 

[MagBen]

there is a big different micro between the first and second layer, second layer shows more typical hypereutectic microstructure!
Strong dendrite may indicate a high cool rate in the first layer.
Are the compositions the same between two layers, any other strong carbide formers added (e.g. Nb, Ti Mo)? the "ordered" carbide does not seem to be primary but rather eutectic carbide, but maybe MC.

 

[bekks]

I agree. They have very fast cooling rates on their first layer and don't precipitate any primary carbides. But it is the same wire composition for both layers.

I was given another sample to look at today from a different manufacturer. I don't have the chemistry yet, but assume it is still in the hypereutectic region. This is a 3 layer overlay and the attached image is the interface of the 2nd and 3rd layers (larger carbides on the bottom are the top of the 2nd layer). The ordered structure is quite obvious in the third layer and I am still inclined to think that it is a carbide of some composition rather than a eutectic structure. I looked at this up to 1000X and can see the black centre of the carbide rod (in some instances). Below the interface appears more like eutectic matrix between the large carbides.

 

[MagBen]

for a hypereutectic composition, the basic structure is primary carbide + eutectic (carbide + austenite) + austenite. I cannot think of any other carbide. in terms of carbide composition, the primary carbide should be mostly M7C3, depending on cool rate and other alloying elements addition, it could be MC, or M3C (if peritectic reaction following eutectic solidification occurs.
the higher the cool, the less the primary carbide, which is not ideal for wear resistance but good to prevent cracking. since the eutectic is a Co-formation mechanism, by nature the nucleation and growth are contained each other, it is often finer and shows some kinds of pattern (ordered).

 

[bekks]

Thanks for the thoughts. I've got an OK to do an EDS analysis, but it might take a couple weeks. I'll try to remember to get back on here with the results.