%Martensite and grain size determination 2

[Best99]

I would like to know the method and/or standard which can determine %Martensite and its grain size in heat treated steel

In my opinion, martensite gran size cannot be measured due to its needle shape.

Thank you for your suggestions

 

[metengr]

Bodyslam;
You can determine volumetric percentage of martensite using TEM. Grain size would refer to prior austenitic grain size, which can be determined with a chemical etchant using optical microscopy.

 

[Neubaten]

Maybe it's too kitchen-sink, but you can determine the martensite % using hardness (if you know the composition, of course)

It's not TEM, but it could serve.

 

[metengr]

Neubaten;
The problem with this approach is that you may have other phase constituents of similar hardness, like bainite, intermixed with martensite that could result in inaccurate results.

 

[TVP]

Bodyslam,

There are a multitude of etchants for steel, some of which allow for easy identification of martensite separate from ferrite, bainite, etc. What you are referring to is the appearance of quenched and tempered martensite when etched with nital, a solution of nitric acid and ethanol or methanol. When etching microstructures that are mostly martensitic, it is difficult to determine the martensite phase fraction using only nital. I suggest reviewing some of the following articles on color metallography and tint etching for more information. The article by Amar De, et al is especially useful.

http://asmcommunity.asminternationa...oken=e31290abe427ad0c0ef4105e87a9fae1ec2b2392

http://www.metallography.com/etching/tint.htm

http://www.buehler.com/technical_information/Buehler Book/etching.pdf

http://asmcommunity.asminternational.org/portal/site/

http://www/AsmStore/ProductDetails/?vgnextoid=ff2f2b41ce0f8110VgnVCM100000701e010aRCRD

 

[redpicker]

I was under the impression that the relationship between hardness and percent martensite for a given carbon content was well known and that as-quenched hardness was the preferred method (as opposed to optical metallography methods) for quantitatively determining the amount of martensite. I know that some API specifications use this method.

rp

 

[metengr]

redpicker;
I had not seen this correlation related to the volume fraction of martensite. I would be interested to see this information. As you are aware, the relationship between hardness of martensite and carbon content is well known.

Getting back to etchants. Yes, this approach can be used for specific applications. However, I have yet to find a met lab that can distinguish between tempered martensite and tempered bainite, other than using TEM, for some of the creep strength enhanced ferrite steels.

 

[NickE]

Ahh... I had a very long discussion with someone about this once....

Martensite grain size, as opposed to %-vol, can be measured.

Since it is a needle shape, obviously you are not actually measuring a real size. Similar to the common untwinned grain wall charts for measuring the size of prior austenitic grain size there is also a chart for twinned grains.

(I think its ASTM E35, but I could very easily be wrong.)

I would use the twinned grain chart, at 750x and a 10x divisor applied to the scale on the chart.

IE: use a well polished sample, etch with picral. Possibly use an etch attack method to increase contrast. The wall chart I speak of is at 75x so 750x provides the easiest magnification step. Examine a large area of the surface and determine the closest match to the wall chart. On the chart is a value for the typical grain size at each size class. Take that value and divide by 10.

I dont think this is an actual ASTM method. However it does allow the average grain size to be estimated.

An alternate method would be to use image analysis software to determine width/length/orientation/aspect ratio, from these you could develop a mechanism to apply quantitative results on the morphology.



Nick
I love materials science!

 

[Neubaten]

Ok, hardness discarded.

Now, another wild-guess: Wouldn't it be theorically possible to indirectly measure martensitic volume fraction by measuring magnetic permeability?.

 

[metengr]

Neubaten;
Yes, as long as other phase constituents are not present, like ferrite and bainite, because these phases are ferromagnetic and will interfere with your results.

If you are thinking along the lines of a ferrite gage this works well for austenitic material, because the austenite phase is not ferromagnetic, so ferrite can be detected and measured.

 

[TVP]

The method to which NickE refers is ASTM E 112 Standard Test Methods for Determining Average Grain Size using either Plate II (flat etch) or Plate III (contrast etch) for twinned microstructures. ASTM has a number of other standards for evaluating microstructures. Use the following links for more information:

http://www.astm.org/Standards/E112.htm

http://www.astm.org/Standards/E562.htm

http://www.astm.org/Standards/E1181.htm

http://www.astm.org/Standards/E1268.htm

http://www.astm.org/Standards/E1382.htm

 

[redpicker]

metengr,

API Specification 5CT (ISO 11960) requires Grades C90 and T95 to have a through wall hardness test performed in the as-quenched condition and the "Mean hardness numbers shall equal or exceed the hardenss corresponding to a minimum of 90 % martensite, as determined by the following equation:
HRCmin = 58 x (% carbon) +27"

It gives a reference of "J.M. Hodge and M.A. Orehoski, Relationship between hardenability and percentage martensite in some low alloy steels. Trans. AIME, 167, 1946, pp 627-642 and limits the carbon content to 0.15% to 0.50%
For other grades, it requires "50% martensite, as determined by the following equation :
HRCmin = 52 x (% carbon) + 21"

These equations appear (to me) to match up favorably with the curves found in the Metals Handbook for the carbon contents covered (probably from the same source).

rp