YAGT: If you ever wondered what your brass REALLY looks like

[Fenixgoon]

So let me preface this with the fact that besides being a budding firearms enthusiast, I'm a materials engineer/metallurgist by training - so I study the science of metals and how the processing, structure, and properties of metals are all related. My specialty is steels, but there's a lot of knowledge that's applicable to all metals, including brass.

Structure, you say? Why yes indeed. Metals actually have a structure to them on the microscopic scale. They're not just one nice big blob of continuous material, even though 99.99% of us consider them that way (even all you structural folks!).

Metals, with few exceptions, are crystalline. The atoms arrange themselves into certain crystal shapes. Most common are body-centered-cubic (BCC) and face-centered-cubic (FCC), but there are others. When you see a giant hunk of metal, it's not one continuous crystal (...unless you happen to be looking at a silicon wafer or a turbine blade). Most metal products we deal with on a day to day basis are composed of millions/billions/trillions of crystals, called grains. Each grain is a crystal with a certain orientation in space. When grains with differing orientations meet, they form a grain boundary.

Now, how do we observe all this? Basically, sand paper, diamonds, a few select chemicals (usually particular acids), and a really nice microscope. You take a piece of metal, polish it to a near-perfect, near-deformation-free surface finish. The polish of your chrome rims? That's got nothing on a proper metallographic sample.

Metallographic samples are generally cut and then ground with progressively finer grits of sand paper (320grit --> 800 or 1200 grit). While 800 and 1200 grit give a nice finish, the sample surface still has a lot of deformation in it. To relieve that, metallographers use very fine diamond and silica (sand) suspensions - I used 6um diamond and 1um diamond. I would have used 0.02um silica, but I needed to clean out the solution, which I didn't feel like doing

For reference, 320 grit is about a ~50um particle size. So 6um diamond is ~10x finer, 1um diamond is ~50x finer, and 0.02um silica is 250x finer. Silica gives an incredible polish, but for my purposes it wasn't necessary. Also, if the silica crystallizes, it forms sand particles, which then scratch the crap out of your otherwise nicely polished sample and you have to start all over.

Shame on me for not taking a picture of my nicely polished sample because it literally has a mirror shine to it. I'll post an update tomorrow.

With an as-polished sample, the one thing that generally can be observed in metal pieces are inclusions - things we don't want in our metal but are there anyway due to things we can't get rid of during processing (or things we don't feel like paying to get rid of).

But if we really want to see the structure, we have to attack our nicely polished sample with acid (or in the case of brass, an ammonia-based solution).

After that is where hella-expensive microscopes come in (as in..people can drop up to $100k on a light-microscope).

Ok, so finally to spring some pictures upon ATOT

I focused down on the base/firing pin area of my 308 cartridge since that's where I spent most of my polishing effort. It looks the nicest, and shows the most interesting stuff.

So here you can see an overview of the cartridge, as-etched

http://pics.bbzzdd.com/users/fenixgoon/00-01-25x.jpg
http://pics.bbzzdd.com/users/fenixgoon/00-02-25x.jpg
http://pics.bbzzdd.com/users/fenixgoon/00-03-25x.jpg
http://pics.bbzzdd.com/users/fenixgoon/00-04-25x.jpg

What you can kind of make out in these pictures are the flow lines in the firing pin area. So when the cartridge is formed, some grains are stretched out or curved. We don't see the actual grains themselves, but all the inclusions (junk) get stretched along too, so that's what the etchant brings out mostly.

Now if we go to the top of the firing pin area, we can really start to see how the brass flowed during the forming process:

http://pics.bbzzdd.com/users/fenixgoon/01-50x.jpg
http://pics.bbzzdd.com/users/fenixgoon/01-100x.jpg
http://pics.bbzzdd.com/users/fenixgoon/01-200x.jpg
http://pics.bbzzdd.com/users/fenixgoon/01-500x-BF.jpg

Sadly, there didn't appear to be much at 500x. Until I used dark field (turns dark things bright and bright things dark).

http://pics.bbzzdd.com/users/fenixgoon/01-500x-DF.jpg

My guess here is that there's so much deformation that the grains are severely stretched out, with a very high aspect ratio (long/thin). Or it could just be my polishing/etching job - I haven't worked with brasses extensively, so I don't know what the best etchants are. I used a general one that's good for a variety of things. Plus I just find this all fun and interesting

I used other imaging modes in this area, but offhand I can't make anything additional out of them. But hey, they still look pretty:

http://pics.bbzzdd.com/users/fenixgoon/01-500x-DIC.jpg
http://pics.bbzzdd.com/users/fenixgoon/01-500x-polarized.jpg

Now, if we move down towards the base area:

http://pics.bbzzdd.com/users/fenixgoon/02-50x.jpg

At the bottom of the cartridge, we can actually start to make out grains since there's not much deformation in this area.

http://pics.bbzzdd.com/users/fenixgoon/03-100x.jpg
http://pics.bbzzdd.com/users/fenixgoon/03-200x.jpg
http://pics.bbzzdd.com/users/fenixgoon/03-500x-BF.jpg
http://pics.bbzzdd.com/users/fenixgoon/03-500x-DF.jpg

It looks like the grains are about 50um and equiaxed - which means they're roughly the same size in all dimensions (unlike in the firing pin area, where there would be a lot of stretching out of the grains). There also appear to be twins (which we'll ignore for the moment). The other thing we can see in this area is that some areas are lighter than others. This could mean that we have two phases - crystals with two different structures (FCC vs. BCC) and/or composition (say, one that is ~35% zinc vs another that is ~45% zinc).

Not knowing the composition of my cartridge brass offhand, its exact processing history (heat treatment and forming), or having much experience with brasses, I'm hesitant to say what phases we do or don't have. I could find out the composition roughly if I really wanted to, or pay ~$25 to get it exactly since I don't have access to the necessary equipment.

What I can say is that the forming of the cartridge appears to be done after any kind of heat treatment, since some grains are heavily deformed (in the firing pin area) while others at the base are equiaxed.

So, now let's move on to the firing pin area where we can really see the brass flow:

http://pics.bbzzdd.com/users/fenixgoon/00-02-25x.jpg
http://pics.bbzzdd.com/users/fenixgoon/04-50x.jpg
http://pics.bbzzdd.com/users/fenixgoon/04-100x.jpg
http://pics.bbzzdd.com/users/fenixgoon/04-200x.jpg

Aaaaaand that's all I got for now. If you happen to blow up a barrel or something, send it my way and I'll throw it under a microscope for ya when I have spare time

 

[slugg]

wat

 

[Howard]

Cool, I guess, but what's the point?

 

[SKORPI0]

What etching reagent was used on those samples? I don't see much contrast on the grain structures. It's been quite a while since I've used a metallograph.

 

[norseamd]

So let me preface this with the fact that besides being a budding firearms enthusiast, I'm a materials engineer/metallurgist by training - so I study the science of metals and how the processing, structure, and properties of metals are all related. My specialty is steels, but there's a lot of knowledge that's applicable to all metals, including brass.

Can you tell me how good 4340 would work for arms and armor and other uses like automobiles? And also why do they not make steel alloy with tungsten or titanium considering their huge strengths?

 

[OverVolt]

Well that is pretty cool, thanks for the pics.

 

[Fenixgoon]

What etching reagent was used on those samples? I don't see much contrast on the grain structures. It's been quite a while since I've used a metallograph.

i used ASTM 30, which i can't remember what it is offhand. ammonium hydroxide, hydrogen peroxide, and water, maybe?

i looked up some typical brass structures that were etched with Klemm's III, and looked way better. i want to say it's kind of nasty stuff, but at the very least i didn't have those chemicals handy.

so ASTM 30 it was. pretty generic, but probably not the best etch. certainly not as colorful as you linked (but i think yours may have be annealed )

Can you tell me how good 4340 would work for arms and armor and other uses like automobiles? And also why do they not make steel alloy with tungsten or titanium considering their huge strengths?

for hobbyist purposes, i'm sure 4340 would work fine. it's relatively lean in alloy content, but gives you enough to be able to quench and temper in thick sections. in fact, it might be overkill, since arms and armor tend to be relatively thin (and the whole point of 4340 is to go thick). but if you still wanted to use it, there's not reason off the top of my head why you couldn't. it might just making forming/heat treating a bit more of a headache.

you'd want it to be in the normalized or annealed conditions to do any sort of cold-forming on it. during hot work and heat treatment, you'd want to avoid decarburization (unless you plan on grinding it all out or don't care). tempering times and temperatures should be used to avoid temper embrittlement and blue brittleness, which as i recall are in the 500-800F range.

regarding armor as in ballistic armor...i have no idea. high-strain rate events are their own animal!

4340 and similar alloys are probably not pervasively used in the auto industry (say, sheet metal) simply due to cost, and most likely formability as well. any time you add alloying content, the price goes up significantly. and in the quantity of material that the auto industry uses, it's simply not economical. on top of that, if 4340 has a higher yield point or strain hardens faster than typical sheet alloys, then completely different equipment would be required, which is a huge amount of capital investment required to move to higher strength alloys.

that being said, some pretty neat alloys are used in say...side impact crash beams, where high energy absorption is a necessity.

and i know i've seen aftermarket 4340 connecting rods. not knowing exactly what the design-limiting feature of a conrod is, it's hard to say how much of an improvement it would be. compared to titanium(6-4), you're adding mass to a moving part - generally not good. on the other hand, you're probably getting a good bit of a strength bump, which means you can probably shave a little meat off to reduce the weight. stiffness increases, which should reduce the tendency towards any buckling criterion provided you didn't turn you conrod paper thin

titanium and tungsten actually are both used in steels. titanium often serves as a grain refiner, and can also precipitate out carbides during heat treatment. when present, you'll probably see it added in small weight percentages. tungsten would be used to precipitate out carbides as well, but as far as i know it doesn't serve much other purpose (e.g. chromium imparts corrosion resistance, nickel stabilizes the FCC phase and tends to add toughness, Mo adds corrosion resistance in 300 series stainless steels).

the other thing is that unless you really *need* alloying elements to achieve a certain goal, there's not much point to using them. steel by itself is pretty flexible as a material. you can get variations in strength of 2-3 just by changing heat treatment.

so if you look up an alloy like A286, it actually has quite a bit of titanium, but for a very specific reason (precipitation of TiC/N for strength). but A286 is a pretty high end alloy compared to what's going to be used in the auto world!

oh yeah, everything here is "to my knowledge as of this post" since it's not like i've been in the metals industry for a lifetime

 

[gorobei]

do you think cryo forging would do anything to improve reloadability of brass? while unsupported chambers and overpressure tend to be the main factors, would a more uniform grain reduce the chances of bulged side in a casing?

 

[olds]

I think he's messing with us and those are pictures of kitchen tile from the 50s.

 

[norseamd]

oh yeah, everything here is "to my knowledge as of this post" since it's not like i've been in the metals industry for a lifetime

Honestly you just gave me some good knowledge about questions I have had about particular materials for a long time now. As for 4340 being lean in alloy content I thought it was basically 4140 with lots of nickel added in for strength?

Also looking at that A286 holy shit that is a lot of fucking nickel that alloy uses. Still the titanium percentage is like only 1.9% to 2.3%. Any idea why there are never any alloy materials that use 40% to 50% of iron and titanium each? What would that material be like and what would be the strengths and weaknesses?

Ever used T1?

 

[Hayabusa Rider]

I'm considering ordering a knife which would be able to take and retain an extraordinary edge. Sandvik 13C26 seems a likely candidate, but not being my field of expertise there may be better choices that I'm missing. Opinion?

 

[gorcorps]

Honestly you just gave me some good knowledge about questions I have had about particular materials for a long time now. As for 4340 being lean in alloy content I thought it was basically 4140 with lots of nickel added in for strength?

Also looking at that A286 holy shit that is a lot of fucking nickel that alloy uses. Still the titanium percentage is like only 1.9% to 2.3%. Any idea why there are never any alloy materials that use 40% to 50% of iron and titanium each? What would that material be like and what would be the strengths and weaknesses?

Ever used T1?

I think I can add to this conversation, since I'm also a metallurgist and work as a process engineer at a steel mill.

When it comes to steel, alloy additions are often used to control grain size. If you look at the pictures the OP showed (or SKORPI0) you'll see a good image of grain structure. The area surrounding each grain is called a "grain boundary". When you deform materials (bend, stretch, etc) you create what's called "dislocations" in the atomic structure that flow through the material. These dislocations flow easily through the grain themselves, but pile up at the grain boundaries and can't flow between them. The pileup of dislocations causes a resistance to deformation (ie: strength). If you use things like titanium to keep your crystals smaller then you have more boundary area, which is pin those dislocations more often, and you end up with a stronger material.

 

[Dirigible]

Nice post. Been almost 20 years since I was taking pictures like that.

 

[franksta]

Have you done any EDX analysis in your line of work? Or another way to check the composition? We got that add-on to our SEM maybe 9 or 10 years ago. It seems to be mostly used it to verify our processing tools weren't contaminating our product.

Also, in your image file names is that the microscope objective's magnification you're specifying or the total mag?

 

[Fenixgoon]

do you think cryo forging would do anything to improve reloadability of brass? while unsupported chambers and overpressure tend to be the main factors, would a more uniform grain reduce the chances of bulged side in a casing?

i've never heard of cryo forging. and generally strength rises and toughness falls as temperature goes down, so i'm guessing it's a cryogenic heat treatment (dunking your sample in LN2 after quenching to room temp).

depends on the type of brass you have, i guess. a quick google shows some brasses might benefit, but for cartridge brass, i don't think it would be beneficial unless it can/does undergo a martensitic transformation.

the purpose of cryo treatment is to complete the martensitic transformation (usually in steel). the reason for this is that after heat treatment and quenching, some alloys have a fraction of their structured that is still un-transformed at room temperature. in the case of steels, this is called retained austenite. by cooling to cryogenic temperatures, the retained austenite then transforms to martensite, since martensite reactions (at least in steel) are athermal or diffusionless transformations.

Honestly you just gave me some good knowledge about questions I have had about particular materials for a long time now. As for 4340 being lean in alloy content I thought it was basically 4140 with lots of nickel added in for strength?

Also looking at that A286 holy shit that is a lot of fucking nickel that alloy uses. Still the titanium percentage is like only 1.9% to 2.3%. Any idea why there are never any alloy materials that use 40% to 50% of iron and titanium each? What would that material be like and what would be the strengths and weaknesses?

Ever used T1?

well, lean relative to other steels i'm used to - custom 465, A286, 13-8Mo, 17-4/15-5, Aermet100 (good shit right there) Those are all much higher alloy content steels

so the big advantage of 4340 over 4340 is the added nickel content further delays diffusional transformations and stabilizes the austenite phase. what this means is that 4340 can cool at a slower rate than 4140 and still transform into martensite (as opposed to getting austenite, ferrite, bainite, or pearlite). this not only lets you work with thicker sections, but can also help reduce residual stresses (since you can cool slower).

here's 4340 versus 4140

so what you can see in this Time-Temperature-Transformation (TTT) diagram is that if you were to try and quench 4340, you have about 10 seconds to get your entire part below the Ms (martensite start) temperature - the goal generally being to avoid any "noses" in the diagrams. Compare that with 4140, where the "nose" of the curve is back at 2 seconds on the time axis. This is what makes 4340 better for thick-section applications than 4140.

regarding why there's not a 40-50% Ti/Fe alloy is because for an iron-base alloy, you only need a little bit of titanium to get the most benefit. for a titanium-base alloy, iron actually becomes an impurity that you don't want, so you definitely don't want 50% of it

looking at the Fe-Ti phase diagram, my guess would be that Fe2Ti is a brittle intermetallic phase (i don't actually know offhand), which means it would be bad news bears during any kind of forging or rolling if you had it.

I'm considering ordering a knife which would be able to take and retain an extraordinary edge. Sandvik 13C26 seems a likely candidate, but not being my field of expertise there may be better choices that I'm missing. Opinion?

so it looks like that alloy might be something like a 420 steel in composition, but with some extra carbon - i think the high-carbon grades of 420 are usually 0.4-0.5%.

http://www.smt.sandvik.com/en/products/strip-steel/strip-products/knife-steel/sandvik-knife-steels/sandvik-13c26/ versus
http://www.aksteel.com/pdf/markets_products/stainless/martensitic/420_data_sheet.pdf

It can definitely achieve better hardness, owing to the higher carbon content, but the Ms temperature will be lower. That may not be a bad thing, since if you were to quench it edge first, you'd get an extremely hard edge and a then softer core (whether it would actually be softer than 420c i don't know). i'm not sure what else makes for a good knife edge, but it looks like the sandvik steel would certainly do the job.

 

[Fenixgoon]

Have you done any EDX analysis in your line of work? Or another way to check the composition? We got that add-on to our SEM maybe 9 or 10 years ago. It seems to be mostly used it to verify our processing tools weren't contaminating our product.

Also, in your image file names is that the microscope objective's magnification you're specifying or the total mag?

we have an EDS on our SEM, but i was too lazy to do it on this EDS can be hit-or-miss, so we tend not to rely on it for precise compositional values. it is a good indicator though, as you said, for checking to see if there's contamination, or if you want to get a rough idea of the sample chemical composition.

AES/OES gives very precise values (but we don't have one of those).

file names are total magnification. objective values are filename/10 (so 2.5x, 10x, 20x, 50x).

I think I can add to this conversation, since I'm also a metallurgist and work as a process engineer at a steel mill.

When it comes to steel, alloy additions are often used to control grain size. If you look at the pictures the OP showed (or SKORPI0) you'll see a good image of grain structure. The area surrounding each grain is called a "grain boundary". When you deform materials (bend, stretch, etc) you create what's called "dislocations" in the atomic structure that flow through the material. These dislocations flow easily through the grain themselves, but pile up at the grain boundaries and can't flow between them. The pileup of dislocations causes a resistance to deformation (ie: strength). If you use things like titanium to keep your crystals smaller then you have more boundary area, which is pin those dislocations more often, and you end up with a stronger material.

*like* :thumbsup: what mill do you work at, and what kind of product do you guys produce? i always thought working at a steel mill would be badass.

 

[Cerpin Taxt]

Can you tell me which exact microscope and camera you used?

 

[Fenixgoon]

Can you tell me which exact microscope and camera you used?

zeiss axio-observer D2(m?) with a mrC5 camera. it's like $50k!

 

[Fenixgoon]

here's the etched cartridge. it's not as shiny as the as-polished condition - the acid attack tends dull the surface a little:
http://pics.bbzzdd.com/users/fenixgoon/mount_308.JPG

 

[Howard]

I think I can add to this conversation, since I'm also a metallurgist and work as a process engineer at a steel mill.

When it comes to steel, alloy additions are often used to control grain size. If you look at the pictures the OP showed (or SKORPI0) you'll see a good image of grain structure. The area surrounding each grain is called a "grain boundary". When you deform materials (bend, stretch, etc) you create what's called "dislocations" in the atomic structure that flow through the material. These dislocations flow easily through the grain themselves, but pile up at the grain boundaries and can't flow between them. The pileup of dislocations causes a resistance to deformation (ie: strength). If you use things like titanium to keep your crystals smaller then you have more boundary area, which is pin those dislocations more often, and you end up with a stronger material.

Hey man, I need some AISI 1005 or 1008 steel plate. Do you know where I can get my hands on a small quantity (say a piece of 8"x8"x1/2") without paying an arm and a leg?

 

[JEDIYoda]

I think he's messing with us and those are pictures of kitchen tile from the 50s.

or your basement tile...

 

[Sukhoi]

Cryo temperatures can also be used to refine grain sizes more quickly during severe plastic deformation as the recovery proceeds much more slowly.

 

[Fenixgoon]

Cryo temperatures can also be used to refine grain sizes more quickly during severe plastic deformation as the recovery proceeds much more slowly.

care to explain? i don't understand how that happens off the top of my head

 

[Cerpin Taxt]

zeiss axio-observer D2(m?) with a mrC5 camera. it's like $50k!

Oh believe me, I know. I never did much materials analysis, but in biological science the imaging methods are highly complex. With complexity comes a wide assortments of lenses, filters, multiple camera setups, lasers, nipkow disks, etc. That stuff adds up really quickly. Confocal setups quickly reach into the several hundreds of thousands of dollars. Hell, I used to carry a camera that itself sold for $45K.

I left before I ever got to work with the newer Zeiss inverted platforms. I'd set up an AxioImager a couple of times, and plenty of Axioverts. Most of my experience was with the comparable Olympus platforms, though.

Cool stuff. I posted this thread back in the day, but unfortunately TinyPic has since deleted everything I hosted there.

EDIT: Actually, this pic and this pic were hosted on bbzzdd so they're still with us.

 

[Fenixgoon]

Oh believe me, I know. I never did much materials analysis, but in biological science the imaging methods are highly complex. With complexity comes a wide assortments of lenses, filters, multiple camera setups, lasers, nipkow disks, etc. That stuff adds up really quickly. Confocal setups quickly reach into the several hundreds of thousands of dollars. Hell, I used to carry a camera that itself sold for $45K.

I left before I ever got to work with the newer Zeiss inverted platforms. I'd set up an AxioImager a couple of times, and plenty of Axioverts. Most of my experience was with the comparable Olympus platforms, though.

Cool stuff. I posted this thread back in the day, but unfortunately TinyPic has since deleted everything I hosted there.

EDIT: Actually, this pic and this pic were hosted on bbzzdd so they're still with us.

wow, that's pretty incredible! looking at nerve tissue/cells?