Jump to content
Aiden CC

Getting into Metallography

Recommended Posts

This morning the materials science lab director at my college trained me on the equipment to mount, polish and polish samples as well as look at hem with an optical microscope. My firs sample was a piece of saw-blade I plan on using for fillet knives. I am completely new to interpreting micrographs, (a little later on I'm planning on meeting with a professor who is specialized more in metallurgy to take a look at the sample), and if anyone has any ideas about these, I would love to hear them.

5x_unetched_with_scale.png

Un-etched, low magnification. I think the comet tails mean I need a better polish. The scratches definitely mean that.

20x_unetched_with_scale.png

A closer look at the little inclusions.

5x_etched.png

Same magnification after the etch.

100x_etched_with_scale.png

Much higher magnification with an etch, maybe this is martensite? I'll probably re-polish and etch a little later on today.

 

Thanks for looking!

Share this post


Link to post
Share on other sites
30 minutes ago, Aiden CC said:

I think the comet tails mean I need a better polish. The scratches definitely mean that.

Yep!  Common for starters, generally looking pretty good.  

31 minutes ago, Aiden CC said:

Much higher magnification with an etch, maybe this is martensite?

Let me edit that for you:  Much higher magnification with an etch.  This is martensite.  

Share this post


Link to post
Share on other sites
2 hours ago, Jerrod Miller said:

Let me edit that for you:  Much higher magnification with an etch.  This is martensite.  

Yeah, that seems to be the consensus. I re-polished (forgot to take pictures before etch, but no comet tails!) and did a gentler etch since it was a bit over-etched at the tooth end. 20x_tooth.png

Not sure if it's anything other than an uneven etch, but there is some un-evenness where the tooth was.

50x_tooth_with_scale.png

2.5x closer

50x_far_from_edge.png

A picture showing some of the small particles which weren't removed by the etch this time. Not sure what they are. It's hard to see in this pictures, but there are little spheres throughout which looked like carbides to me and the professor I showed the sample to. X-ray fluorescence found only iron and manganese, maybe the steel is hypereutectic  and they're iron carbides. XRF can't see carbon, it looks like the best thing I can do is anneal the sample and look at the relative amount of ferrite/cementite. Might be able to make a quick python script to do it for me...

Share this post


Link to post
Share on other sites
25 minutes ago, Aiden CC said:

Might be able to make a quick python script to do it for me...

ImageJ does that kind of thing automatically for you (calculates area based on contrast parameters that you set).  It is fairly simple and it is freeware.  It is also what I use to add scale bars and such to my micrographs.  

Share this post


Link to post
Share on other sites

The low power photos with the 'comet tails' are significant to me. I see  a few of these in every batch as I am polishing out and always wondered if they were voids or inclusions. I would sure like to get  a good close look at them to help determine when in my processes they are occurring. And how to improve my abrasive stages to avoid seeing them at final polish. I can see the streaks fairly clearly with my 20x or 40x hand lenses, but not clear enough to tell what the void or inclusion is.  I have tried to figure a way to macro-photograph these but the lighting is seemingly quite a trick, esp. when I try over a 100x in a small microscope.

Any hints?

I googled searches such as  'metallography' and ' metal surface macro photography' but so far found nothing applicable to a DIY settup.

Share this post


Link to post
Share on other sites
Posted (edited)

rh1g.jpg

A streak of comet tails, taken through a hand lens. A streak like this is quite unusual in my work as I normally only see an occasional single one, but this more extreme example makes for better study. This pic is from an iPhone mounted on a tripod looking through a mounted 40x hand lens, with side light provided by a small led spot.
The tiniest changes in lighting dramatically change the image seen, and made me appreciate metallurgical microscopes that have the light coming through the upper lens, not just externally upper lit.

rh1h.jpg

Here’s the same setup with a slightly different light angle.

rh1a.jpg

To try to sort out the source of the comet tails I gave the blade a quick touch with 1,000 grit.  Now the inclusions appear as a row of dark dots, the lines are the 1,000 grit ‘grooves’. The color is a streak of felt pen to help me locate what is essentially invisible to my naked eye. The run of ‘dots’ is about .040” long, each individual dot is about .001”.  

(Thanks to learning that  ‘comet tails’ is indeed an industry standard descriptor - I used it to find other troubleshooting pages. In one page by Struers  https://www.struers.com/en/Knowledge/Grinding-and-polishing#grinding-polishing-troubleshooting    is  an image fairly like this one that they call ‘pull-outs’.   However, my conclusion so far is that any time I am getting comet tails they are most likely inclusions of bits of friable abrasives. My best guess is that they are occurring when the steel is still soft, but after forging, at stage 3 below.  (My original concern was that they are inclusions happening when I hot-shape the blades to a precise form, or were voids happening in the salt bath, or came on the ground sheets from the steel supplier. )

Stages:

 1. Ground flat stock (O1) is rough ground to shape

2. Heated in propane forge at low end of heat range and bent around a precise form to establish its shape. 

3. Inside is ground with 120 grit to remove any decarb / scale and to establish that the surface and tool itself is ready for the austenizing salt bath.  

4. Heat treat is 15-minute soak at high 1490F…salt bath marquench at 425F…cool to 125F and temper at 395F.

 

 

 

rh1b.jpg

Here is the same photo with a only tiny light change, the dots go from dark to light.

rh1 k.jpg

Here’s the setup I used with 40x hand lens mounted and adjustable. The aluminum tube on its side is my favorite tool, it is a visual microscope ‘Micro Mike’ It goes in the small hole above to be held securely.

 

rh1d.jpg

Trying to take a pic down the tube of my ‘Micro Mike’ isn’t easy! The lighting here is distorting the apparent dimension of the inclusion streak, but generally the dots are  in the .001” range. 

 

set3curves.jpg

I produce about 40-50 knives a week for orders of 12 different models, half of them being bent, so I pulled these 3 from today’s batch for a photo to show the bended shapes.  The only place I ever see comet trails (though they are rare) is on the inside surface of the bent knives.

This surface, unlike most knives, has no secondary bevel or micro-bevel…it eventually becomes the actual edge, thus it must be close enough to a perfect surface that, as the wood carver sharpens, they will never encounter any inclusions or voids, however microscopic, that will eventually show up at the edge. 

But, I have sent this picture also to show the difficulty I have in getting proper magnified photos, or even manual viewing, as all surfaces are curved. The curved shapes also pose some difficulty in getting lenses and light where they are needed – esp. in a microscope setup.

I have a 40-900x hobby level microscope and tried to use it to take photos for this post, but I found the lighting almost impossible. I see that there are a number of metallurgical microscopes on eBay from India under $600. That is still a difficult $ expenditure for me to justify unless I am confident it can work with these curved blades.

The comet tails are not really difficult to get rid of once I see them, but it means going backwards in my processes.

Any thoughts?

Are there scopes out there that are made more for lower power surface study than microscopic metallurgy?

900x.jpg

Hopefully a new metallurgical scope has more open focal length / air-gap-to-object than this student 40-900 microscope has. I trust the issue I have in getting a good image is obvious!

Edited by Del Stubbs
duplicated pics x4

Share this post


Link to post
Share on other sites

Del, I'm definitely not an expert, but you might consider making a test blade, then carefully cutting/polishing a section of it for viewing. All of my pictures are taken on surfaces which are fairly flat, as they are mounted in resin then lapped/polished.

I got a bit more done today. I mounted and polished some samples form 1095 and 440C as they came from the mill. I also learned how to program the furnaces and ran a single annealing cycle on some of the scrap steel I'm working with (also got my waterjet work order for the saw-blade fillet knives completed, but that's for another thread). Now for the pictures.

20x x2 scaled.png

This is the 1095. Looks like pearlite.

50 x x2 scaled.png

A bit closer, also looks like pearlite, but I'm no expert.

440C 50x scaled.png

Here's the 440C, I think it needs a more aggressive etchant than the 2% nitric acid in ethanol I've been using. 

 

On 2/23/2018 at 6:16 PM, Jerrod Miller said:

ImageJ does that kind of thing automatically for you (calculates area based on contrast parameters that you set).  It is fairly simple and it is freeware.  It is also what I use to add scale bars and such to my micrographs.  

Do you think these pictures would be good enough (at least for the 1095) to use ImageJ to calculate carbon content? I can see what look like pearlite grains, but can't really see what's going on inside them. The plan is to polish some samples from the spring (which I found with XRF to contain a bit of chromium) and saw to calculate carbon content (probably just figure out if it's hyper or hypo-eutectic).

Share this post


Link to post
Share on other sites

Yes, the 1095 is spheroidized annealed. It (sometimes) comes that way from the manufacturer. 

Share this post


Link to post
Share on other sites
30 minutes ago, Skip Williams said:

Yes, the 1095 is spheroidized annealed. It (sometimes) comes that way from the manufacturer. 

That would make sense, hopefully I'll get to see some pearlite in the samples I normalized this afternoon, plus I care more about their carbon content, this was just a test.

1095 etch 100x center processed.jpg

Processed this using imagej, using the auto-threshold after finding edges. In this image the black portion (presumably ferrite) makes up 77% of the area. Using the densities of ferrite and cementite, as well as the weight percentage of carbon in cementite, I got a percent carbon in this image of 1.05% be weight, pretty close to the actual value of 1095.

Share this post


Link to post
Share on other sites
Posted (edited)

After etching with nital both ferrite and cementite will be about equally bright. You'll have to use Picral, Klemm's, or Baraha's reagent to stain them differentially.

Revealing the Microstructure of Tool Steels, March 15, 2012 by George Vander Voort, Figure 3
https://vacaero.com/information-resources/metallography-with-george-vander-voort/1174-revealing-the-microstructure-of-tool-steels.html

There is a cheat called 'heat tinting' (J. E. Stead, 1905?) which is to heat the sample until the temper colors show. IIRC, ferrite will oxidize at a much lower temperature than cementite. But this obviously doesn't work if your sample is mounted in resin.

Edited by Skip Williams

Share this post


Link to post
Share on other sites
2 hours ago, Skip Williams said:

After etching with nital both ferrite and cementite will be about equally bright. You'll have to use Picral, Klemm's, or Baraha's reagent to stain them differentially.

Revealing the Microstructure of Tool Steels, March 15, 2012 by George Vander Voort, Figure 3
https://vacaero.com/information-resources/metallography-with-george-vander-voort/1174-revealing-the-microstructure-of-tool-steels.html

There is a cheat called 'heat tinting' (J. E. Stead, 1905?) which is to heat the sample until the temper colors show. IIRC, ferrite will oxidize at a much lower temperature than cementite. But this obviously doesn't work if your sample is mounted in resin.

That's good to know. I'm not sure if I could get my hands on any of those etchants, I would have to talk with the materials/chemistry lab director. The heat tinting might work, the mounting press runs at 180C to melt the resin, so it's possible I could color the steel by heating just the surface without degrading the mounting that much. The SEM was under repair when I started this project, but now I believe it is working, so I might look into that route for composition analysis as well. 

My hope is that the samples I'm actually analyzing are pearlitic (which would mean that my heat treatment of them worked). Would that be any different than the spherodite to analyze? It seems like I should at least be able to determine what side of the eutectic point it is by how the micrographs look (pearlite + ferrite vs pearlite + cementite).

Share this post


Link to post
Share on other sites

I think you 'll be just fine. Jerrod and others can tell you how long and hot to soak the speroidized material to put all the carbon back into the equation. You're doing good so far!

Share this post


Link to post
Share on other sites
Posted (edited)

Looks like you are on the right path here.  Here are a couple images I took (over a decade ago) of mild (I think 1020 and 1040).  Both were polished then etched with a weak nitric/ethanol mixture, but it may have been up to 5%.  You can see that there is certainly good contrast between the cementite and ferrite.  Also, keep in mind that grain boundaries are going to show up black, but should not be considered as cementite.  You may want to play with concentration and etch time to improve your contrast.  Stains like Skip mentions are super useful, but can be a pain; one more step to go wrong. 

square3.tif

 

circle2.tif

 

Edited to add: Not sure why the images are showing up as links to download rather than just images. Oh well.  

Edited by Jerrod Miller

Share this post


Link to post
Share on other sites

Thanks guys! I did a hardness test on the pieces I heat treated, and it turned out I forgot to hit save after entering the program into the furnace, so they only got brought up to 400C. Today I re-did it and confirmed that he furnace reached a maximum of 925C with a brief soak to make sure the pieces came up to temperature. The saw blade was 45 HRC and now measures between 15 and 18 HRC. Hopefully I'll get some pictures of some samples (which I'm hoping are pearlitic)  before I leave for break. 

I got some budget from my school which I used to buy a piece of CPM 154, I'm going to try and make some blades out of it over break and heat treat them when I get back. Turns out I can also put in work orders on the water-jet for this kind of stuff, so I designed some fillet knife blanks and got them cut last week, which I may leave as is or re-harden, depending on how hard fillet knives ought to be (they are currently 45 HRC).

Share this post


Link to post
Share on other sites

Just mounted and polished some samples of the saw blade, and I'm pretty happy with how it came out.pearlite.jpg

It's hard to see in this picture, but the iridescent surface after etching is really cool, apparently that's why it's called pearlite (I wounder if it would be possible to get a blade to look like that...).

5x.png 

A low magnification picture showing the decarburization around the surface (I probably should've wrapped this in foil before having it in a furnace so long). That would explain why the hardness measured so low (15-18)

20x edge.png

Closer up picture of the edge showing the decarb creeping into the sample (I'm assuming the white is ferrite).

50x center.png

This is the picture I think speaks the most. Higher magnification, near the center. From what I can see, the structure is almost exclusively pearlite with little to no ferrite. Am I correct in thinking this means the sample is close to 0.77% carbon? Maybe it's some old form of 1060/1075 steel? The only metals are Fe and Mn according to XRF analysis.

50x edge.png

Same magnification, showing carbon loss at the pearlite grain boundries.

100x center.png

Maximum magnification showing grains of pearlite, showing that there is essentially no ferrite between the grains in the center of the sample.

I'm also still debating if I should re-harden some fillet knife blanks made from this stuff. They are currently 45 HRC, which seems a little soft for a knife, but I'm not sure.

 

Share this post


Link to post
Share on other sites

Looks like you are getting the hang of it!  Hard to polish it so everything is in focus at that magnification, isn't it?  

Share this post


Link to post
Share on other sites
13 minutes ago, Jerrod Miller said:

Looks like you are getting the hang of it!  Hard to polish it so everything is in focus at that magnification, isn't it?  

Yes, quite! I probably should change the disk on the grinder the pad on the coarse polisher, which might help a bit. Are these pictures enough to tell that I’m looking at a roughly eutectoid steel?

Share this post


Link to post
Share on other sites
55 minutes ago, Aiden CC said:

Are these pictures enough to tell that I’m looking at a roughly eutectoid steel?

Looks like it to me (well, slightly under, as you note above).  I'd guess treating it like 1070-1075 would be appropriate.  

Share this post


Link to post
Share on other sites

It's been a little while, but I have some more pictures! This time, I normalized the spring in a furnace and then did some hot work on it. For now, I just have pictures of the normalized sample and an "as forged" coupon.

First, the "control"

20x Scaled.png

50x Scaled.png

Definitely pearlite, looks slightly hyop-eutectoid, maybe around 0.65-0.75% carbon (I seriously need to get SEM trained).

Large Piece 10x scaled.png

Now after being aggressively forged and dropped on the floor after the last heat. Not sure what the black streaks are, maybe carbides, or dislocations between grains. There are more of them by the surface (bottom)

 

Large Piece 50x Scaled.png

A little closer (I know, my polish needs to be flatter, I was lazy and didn't want to change the abrasives on the rotary/roll grinders, that probably didn't help). Beside the strange dark streaks, this looks a lot like martensite, which would definitely explain the difficult of working steel right out of the forge. I have some more samples, which have been normalized 1, 2 and 3 times in a forge as well as a third "standard" heat treat of three normalizations followed by a "by eye" quench in canola oil and tempering at 375 F in a toaster oven.

Small Piece 50x Scaled.png

An interesting comparison; this section was cut 90 degrees to the other, such that he primary direction of the drawing out was the vertical direction of this photo. Does anyone know what the dark streaks might be? It's interesting how similar work hardening looks to a quenched microstructure.

 

Share this post


Link to post
Share on other sites

Sure looks like martensite.  I'm glad you're getting the chance to do this, and a little jealous!

Share this post


Link to post
Share on other sites
6 hours ago, Alan Longmire said:

Sure looks like martensite.  I'm glad you're getting the chance to do this, and a little jealous!

I'm definitely enjoying it! Plus, I got some money for steel and a bit of credit for it. My college is focused on hands on learning, so for a small school it has a pretty well equipped machine shop and materials science lab. Also, of the matsci professors who has been here for a while specialized in shape-memory alloys before he shifted his research to education, so there is a pretty good casting setup, which I've used to make some ferules for a set of fillet knives.

Share this post


Link to post
Share on other sites

The black really reminds me of bainite.  

Share this post


Link to post
Share on other sites
Posted (edited)
3 hours ago, Jerrod Miller said:

The black really reminds me of bainite.  

The MatSci professor I asked thinks so too, my current guess is that this is a mixture of work and air hardening.

Edited by Aiden CC

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×