Inspecting for Porosity with a Microscope

Interesting subject. When I was pouring through giant sprues with a pouring basin at the bottom, I had a healthy gate going into the hammer head but often had defects at the gate/hammer junction.

In trying to learn, I have to form a theory then implement the theory to prove it. The theory I'm working with now, like Denis says, is to provide a reservoir on the inlet to the pattern reasoning that after all the metal has passed by that will be the hottest part of the pour and should solidify last. To me, your riser on the right has a small connection to the part which is likely to freeze first. I am currently trying to provide a big connection. In the last few weeks I have poured several castings with a bifilm basin and small runner. That design loses a lot of heat and I have had lots of shrinkage. The inlet riser has stopped that. An outlet riser didn't shrink much but the part did.

The first feed to the farside of a part will be the coolest so it should solidify first.

I'm guessing by postulating a theory then investigating it with multiple pours. I did notice olfoundryman putting a riser at the entrance to a part and I think I know why now. but give me a new part to cast and I'll be confused.

 

I would like to share my 2 cents.
First,

This is theoretically not correct . Let me explain. You pour your castings at some level of superheat. You have 3 kinds of melt contractions:

1. from temperature you finish pouring to the liquidus temperature;
2. during solidification from liquidus to solidus;
3. solidified casting contraction.

One wants to pour at the lowest temperature possible to minimize 1st kind of contraction and maximize metal yield. Additionally, at higher temperatures difference in hydrogen solubility between molten state and solid state is higher => thus increasing gas problems (can be solved by proper degassing that is hard to be done in hobby conditions). But, you need just high enough temperature to have adequate fluidity/castability => fill the whole mold and achieve sharp corners/letterings and similar.
The second kind of solidification can cause shrinkage. The kind of shrinkage depends on the properties of the Al alloy being poured:

1. If the alloy has a low modulus of elasticity and low yield strength, the shrinkage will be on the outside of the casting since the inner vacuum caused by melt contraction will be sufficient to deform initial solid skin formed at the surface of the casting (this is often the case for compositions close to pure Al).
2. If the alloy has wide solidification range (for example AlSi6Cu4), you will always have some dispersed porosity (a lot of small pores and most are visible only on higher magnification) since the feeding paths are blocked at some point due to the developed dendrite network. This does not necessarily mean that these alloys have higher shrinkage, it just means that this shrinkage cannot be solved by risers and is unavoidable.
3. if the alloy/metal has narrow solidification range (AlSi12 for example), the shrinkage can be solved by proper feeders and chills

The shrinkage of pure metals and narrow solidification range alloys is solved by directional solidification: you cool one side of the casting (the casting has thin walls on that section, you put a copper chill in the mold, you place several cooling fins to increase surface area), and you heat another side of the casting (put the feeder). For feeder to work it should be at the top of the casting since the metal flows from the feeder to the casting by gravity (do not bother with atmospheric risering at home). To achieve proper directional solidification Heuvers circle method (not ideal, but great for hobby casting) is often used to achieve proper directional solidification.
Therefore, if one wants to feed through the ingates (Cambell does not approve this ) => use large cross-section ingates. Rule of thumb is following (this is theoretical value, if your pouring temperature or times are high, you will heat up the mold around the ingate and dimensions can be lower than stated here):

• If one wants for the gate to be thermally neutral the thickness of the ingate should be half of the mold wall it is connected to in T connection type: Tcasting=Tingate/2
  
• For the ingate to be feeder it should be 2x the section size of the adjacent mold wall: Tingate=2xTcasting
  
• for the ingate to be a chill it should be as thin as possible

I hope you now understand why Al Puddle had problems he had with thin ingate and riser put at the thinner section so that nothing was feeding his thickes part of the casting and therefore he had a sink.

Now, to differenate the shrinkage and gas porosity. Shrinkage has elongated irregular form (interdendritic shape), while gas pores are more round due to surface tension. Gas pores caused by mold gases or improper pouring (blowholes) are usually larger the gas pores due to dissolved gases (which are regularly dispersed).
There is much much more to be said on the subject, but for small social media post.... I hope I was thorough enough but not been too long and boring

 

Not long and boring at all, but my level of ignorance is astounding. If you don't mind, some comments.

I was totally lost by this statement and would like more detail. You are saying that it is theoretically not correct that going from 0.25 sq in runners (1/2"x1/2") to 0.04 sq in runners (0.2"x0.2") does not lose more heat into the sand? It seems logical to me that a larger cross section causes more heat to be retained in the metal due to the smaller perimeter per sq in of cross section.

1. I'm stubbornly pouring extrusions which I believe to be mostly aluminum
   What range of diameters are you expecting here?

Thanks for that mini-dissertation. I had never heard of Heuvers Circles and don't fully understand (do we ever) but am grasping the concept.

• Did you mean Tingate=Tcasting/2?

Unfortunately your text does not change my take on Al's shrinkage. His riser could not feed because the thin gate connecting to it froze before the bulk of the casting froze. Similarly his sprue could not feed because the gate froze first, but the sprue had a better opportunity to feed because all the metal that flowed through the ingate warmed the sand and provided a delayed cooling. Is this incorrect thinking?

Can you comment on the shape of pores I am seeing? I see nothing I would call round, and they are relatively evenly dispersed, which makes me think they are oxide cavities. I appreciate your time and effort.

 

Are you blinding me with science? I noticed you were working out a theory. Your theory is a little different from mine.

It may turn out that riser placement can be effective at the inlet or outlet of the part. I have more experiments to perform.

The shrinkage on the part was very slight. I had to stop pouring when I saw metal in the riser. Upon cooling, the riser shrunk slightlly and the sprue shrunk a bunch. This leads me to believe the feeder to the riser froze the instant sprue and riser height was the same because there was no flow. Since I'm trying to cast a big blob maybe I need to increase my cope height, and riser height, to allow more time for the part to solidify while still maintaining flow through the part.

 

This did get a little complicated, Al. Like you, I'm trying to pour a blob, not as large, but still pretty thick for it's size.

My take on your gating is that the coolest metal enters your riser, it has already traveled across your pattern, warming the sand as it goes, and losing heat. So it will be the first to freeze, not because it stops moving, but because the metal front has already cooled.

Try moving your riser to between your sprue and your pattern, then make the connection between the two as big as you can, as thick as your part is not too big, we're trying to keep it liquid while the part solidifies and that thick part will take some time to solidify. A blind riser will work fine, but by being between the sprue and the pattern it will be filled with hot metal, hotter than the pattern. You don't need pressure to fill a shrinking part, just contact with liquid metal. Do that before you build a higher flask, I think you'll be surprised. A blind riser on top of your pattern would not do as well because it is filled with cooler metal. I had a blind riser on top suck metal from my pattern below it.

Your sprue shrank more because it was hottest, therefore cooled later, and the gate between the sprue and pattern had been heated by all the metal flowing through it.

 

No, I am stating that loss of heat decreases the shrinkage. I understood your cited comment as you claiming that higher heat loss (lower temperature) causes higher shrinkage.

You mean diameters for risers?

The concept is the basic one in casting methoding and is an extension of Chvorinovs rule. In essence, it means that cooling time of the particular casting section is proportional to its modulus defined by: modulus=volume/cooling surface. Basically => the more surface to cool the metal > the less time it takes. To avoid complicated math, the simplest implementation is to correlate modulus with the circle drawn in the cross-section. For most alloys that solidify in narrow solidification range one wants for modulus to increase up to the risers and then: (modulus of the adjacent casting section) : (modulus of the feeder neck) : (modulus of the feeder) = 1 : 1.1 : 1.2 => feeder should have 20% higher modulus than the casting section it is connected to (this method is not valid for cast irons because of graphite expansion, If someone is interested > we can discuss that also). That is the way you calculate riser dimensions. This has interesting implications: high and narrow risers are not beneficial since you decrease its modulus. Commonly used risers are cylinders where Height = 1.5 x diameter > but I have seen many ratios in practice since things are not straightforward as I explained here.

Yes. Sorry for the typo

I have the same feeling. One can calculate the required feeder neck thickness (in this case he wants ingate to serve that function) by modulus method I explained earlier.

The pictures are not great so it is hard for me to tell exactly. But it seems you have everything there.

This cross section defects look like gas porosity

These look like shrinkage

And one picture from your first post looks like traped dross. But the optics of your microscope are so poor I do not know what I am seeing .

 

I'm a little late with this one.. But Andy said something on the last page...

I'm probably all wet and will admit I have not personally poured brass, but from what I've seen, I don't consider brass to be anything remotely like bronze. ZINC IS SHITTY, stinky and volatile.. Silicon bronze is literally like pouring hot honey. I've seen zero porosity in my work so far. Ceramic shell might be helping with that too. Shell is gas permeable, probably more so than packed sand? I don't know. But I do know you'll never see me melting brass. I've had one bad zinc headache and it was a mother. Never again for me. Only thing I've ever seen is a void here or there up by my cup or in the main sprue. Careful spruing can control where you put the hot metal and when. Now that I said that, I'll probably fck up my tree. First time for me running something totally hollow.

 

I used to pour brass and have seen holes in it. Not sure if they were shrinkage porosities or degasing issues but I haven't seen them in bronze. Maybe the holes are being smeared. But maybe bronze is less full of holes than other metals?

 

Well, my comment intended to say higher heat loss resulted in lower temperature in the gates causing them to freeze before the pattern could suck up liquid from the riser. For the same pouring temperature smaller runners result in cooler metal in the pattern, right? (slower flow rate, more heat into the sand)

I meant for porosity. How small is still considered porosity, and reasonably how large. I know how large is unlimited but for small round porosity do you have a feel for expected size range? What I am getting at is my magnification good enough to see porosity or am I just seeing giant holes?

Thank you, I'm learning. by trial and error I figured out a squatty blind riser trumps a narrow open riser to feed.

Since modulus is volume divided by surface area, it's units are linear, cu in, sq in, modulus is in inches. Very logical that a cube has the highest modulus and a thin sheet has the lowest. A thin sheet can freeze even before if fills. But then you said "the more surface to cool the metal the more time it takes", unless I didn't understand the symbols. arrows or greater than symbols?

Thanks for the riser ratio guidelines. Right now I'm trying to cast solid cylinders and my feeder neck (gate?) is the same size as the cylinder and works, but I get the point.

Don't give up on me, we're making progress.

We all make them, and in this case it helped me to think

So now the rubber hits the road, lets say his casting is 3"x4"x1/2". Modulus is 6 cu in/31 sq in=0.19 in. Or do you take out gates in the area calculation so you only have sand contact area? So if he has a 1" wide by 4" long feeder, it should have a modulus of 1.1x.19=0.21, and a 1"x4" needs a thickness of 7/8" to get a 0.21" modulus. Probably it's only the sand contact area so pattern modulus becomes 0.21 in and a 1" x 4" feeder should be 11/16" thick to get a 0.227 modulus. I guess it's reasonable the feeder should be thicker than the pattern. Then the riser should have a modulus of 0.25 in. An arbitrary 1-1/2" square cube has a 0.25 in modulus.

Are we getting there?

And I thought I had great pictures! I really don't think the pictures are that bad, the surface is just not well polished. I think

Thanks again. I'll take some pictures of something which is cleaner at the photographed level, like the blue tablecloth.

 

Yeah, I misspoke, I have never poured tin based bronze.

You need a good respirator to safely pour brass, and that goes for lost foam as well. Just trying to stay our of the fumes is like texting and driving. Usually works.

I've had zinc fever many years ago welding on galvanized structural. Now I wear a respirator, bu tif you get caught drinking lots of milk immediately vastly reduces the effects of the zinc poisoning. Old welder thing, don't know if doctors recommend milk.

 

Today is a busy day for me. I do not have time to check the math, but I can do that later.
Not to lose momentum of discussion => read this: https://massalotando.files.wordpress.com/2018/10/projeto-livro-wlodawer.pdf
It is for steel casting, but the principle applies for all narrow solidification range alloys. In chapter 2 you have worked examples of modulus calculation. Chapter 8 gives you an idea how to use chills to reduce the necessary number of risers (increase metal yield). For Al alloys you use copper chils since Al reacts with iron.

Everything is considered porosity. It is just a matter of the application to acknowledge what size can be tolerated. Those small pores you see under higher magnification in a microscope is just called microporosity > but it is still a porosity.

Post edited. You notice my every typo. I type fast with both hands, and in some cases the fingers are quicker than the brain

 

Interesting video. I've done some etching with caustic, but mine is weak so I let it set. I haven't had a chance to listen to it yet, just scan through. I've never looked at an etched part with the microscope though.

After CLR questioned the optics of my microscope I thought it might be appropriate to do some tests so this is what I came up with. I tried to find something small but well finished so here are two old rulers.



Fine scale on the lower is 1/64th inch marks, and on the upper it's 1/100th inch marks.

Here's the center of the A in U.S.A.



And the 1/64th marks



And 1/100th marks. Each stripe is therefore approximately 0.005". A 0.006" porosity would be 20% larger than one stripe.



Then I switched to my knife. I buff it on red rouge on a cardboard wheel. It will cut you with light pressure with no pain and no blood, for a little bit. Kind of like a scalpel.



Here's the edge. Scaling off the computer screen (all images are unchanged from the microscope image so they may be distorted between vertical and horizontal but should all be proportional. There were 7 marks 0.005" wide in 100 mm, so 0.035"/100 mm or 0.00035" per mm) the scratches are about 0.0003 inches. That jibes with moving my mill table 0.0005" on the DRO.



I am more convinced the microscope is doing OK, picking up features as small as 0.0001", and I need to learn to polish better.

 

A HUGE, HUGE thank you for this link!

Now I know why my castings have been shrinking and how to fix ingate problems. A superb read for anyone who hates making scrap!

Cheers Charlie

 

I'll second that, thanks for the link, it's going to change the way I think. It will make some sense out of risers and gating. Sulzer is a great company.

I'm getting a lot out of it, and they simplify things nicely.

 

A little update and maybe some progress.



The first hammer head, with all the porosity, was poured from a crucible with lots of swarf and I let a steel screw get in with a window frame. That makes the aluminum boil and may have been a major source of the porosity. The second head was poured from window frames and the usual sprues and a few cans. The riser was too small and there are shrink defects on the surface of the hammer. The third head was poured from old muffins I made mainly from cans a couple of years ago.



When I poured it apparently I failed to clean out the runner and the sprue was barely connected to the runner. It's a wonder it poured.



Here's the fracture



I only showed the screwup because it may have somehow helped. The second head has some visible porosity, but under the microscope it's pretty nasty.



The pictures are about 0.045" wide, and 0.035" high for reference.

The third hammer has very few porosity holes, and they are hard to find, maybe five total. This is a larger one, about 0.007".



I've been pouring ingots with my spout, and the muffins seemed to have done well on the hammer head, with the spout and a pouring basin, and small runners. So I cut apart a recent ingot and compared it to the box section from a diesel pickup.



Very little porosity, and very little, about 0.003" was the largest.



The box section has lots of porosity, 0.003" to 0.008". This lower one's about 0.0075".



I'm using the DRO to measure with.

So, did I get lucky on the third hammer head and ingot? This is all extruded material, and no degassing.

What I'm thinking at this point is the spout and pouring basin all but eliminates turbulence, and the foam plug in the sprue puts a slug of vapor ahead of the metal front, shielding it from oxygen. I'm also using a large attached blind riser.

I don't stir my crucible, but neither do I clean it out much between pours (it's 304 pipe with holes and the aluminum oxide is keeping the liquid from seeping out the pinholes), but if I can get the level of porosity in the hammer head and ingot I'm pretty happy.

The first hammers I made, when I was dumping dross and all into a giant sprue with a settling basin at the bottom, I would get some giant holes where the gate met the hammer face, 1/8" and more. It really doesn't matter for a hammer, it soon gets peened over anyway. But I sure like these results.

 

Sorry to be so uninformed, but what is the truck box section from—-the body, engine, frame???

Denis

 

Yeah, you got lucky. That kink in the sprue/runner connection slowed down the flow and helped eliminate turbulence. Do it again.

 

Sorry, you're not uninformed. It was off a GM Duramax, I believe. Castings given to me by a guy who repairs them. I think it was an air passage because it was black inside, but maybe water.



It was several inches long off the piece in the middle. Looked good when cut so I used it as a commercially produced sample. I've also got some suspension parts, the end of one is just to the right in the picture. I need to polish one of them to look for porosity too.

With 3/16" square (5 mm) runners and about a 2 inch sprue I really don't think I'm getting much turbulence. Didn't in the exploding glass video. But, nonetheless I'll do a few more to make sure I can replicate these results. That run was sure slow, the open riser at the end of the runner didn't even fill.

When I started down this path I wanted to definitively determine if salt degassing would do me any good with my setup. If I'm getting clean castings without it, I may not go there. When I first started I was making muffins thinking I was cleaning up the metal with a melt. Then I started reading about hydrogen absorption and how remelting was bad for you. So I only made ingots when I was bored or had extra melted metal. Now I'm not so sure, but having the tools to investigate lets me try and examine. Air in ice escapes as the ice melts, I suspect hydrogen in aluminum pockets may do the same thing right along the melt front. But, then again, I run my furnace rich to limit oxide production. Burning used motor oil the hydrocarbon chains may not be releasing much hydrogen. My level of ignorance is astounding.