Melterskelter You were talking about ferrosilicon causing shrinkage in cast iron. In my latest video I pour two exactly the same castings and the first one has a lot of shrinkage and the other does not. The third and last casting has some shrinkage but not as much as the first one. There are two possible explanations as to why it happens. Pouring temperature is too hot or fading of the ferrosilicon effect. I used 40 grams of ferrosilicon to 14,000 grams of cast iron. I am not sure which one is causing the shrinkage.
Very strange the way the petrobond casting came out. So many things spinning in my head right now I dont know how to comment...
If I am doing the math right, you used 1.41 oz of ferro for 31 lbs of iron, which to me is basically no ferro. Here is a flywheel I poured in 2012 in petrobond, and it turned out ok, but had a few sand inclusions where the petrobond failed. I did not use any ferro, and it machined very easily. The iron was clean Class 40. I did not use a riser on the hub, and notice that there is no shrinkage there, and no riser required at the hub, perhaps because of the rather tall sprue. And here is a second flywheel from the same pattern, poured in 2013, this time using resin-bound sand, and also without ferro, and it too machined beautifully. This was a crazy flask layout, and lots of parts in a common mold (big mistake), so I used a giant riser to keep the other parts from drawing from the flywheel. There seem to be some persistent myths about iron, such as: 1. Iron must have ferrosilicon added to it to make it easily machinable. A myth circulating on ytube is "all cast iron parts must be left in the mold and allowed to cool slowly, else they will not turn out machinable". Both of the flywheels above were removed from the sand within 15 minutes of pouring. And another myth on ytube is "all iron castings must be tempered and annealed in order to be machinable". Again false. I did not notice much shrinkage in the risers for these parts, but the risers were oversized for insurance purposes. .
If it was the fading of the ferro, then how do you explain my flywheels above that had no shrinkage and no ferro? or maybe you were referring to the plaques? My pouring temperature for both flywheels I am pretty certain was on the low side of the pouring range for iron. .
Obviously you missed what ironsides was referring to... Look at the pour sprues.... Also his pours were from the same crucible on the same day not a year apart That would depend on the carbon content of the iron to begin with as well as the thickness of the casting, the medium it is poured into, and the rate at which it is cooled. In my mind the lower classes of iron like 30 or 40 would show less benefit from a ferro silicon addition than that of a 60 class (due to the carbon content) So no its not a myth, it just depends... Not so much a myth as it is a good practice to allow them to cool slowly. Rapid cooling can and will induce a chill. Well not all anyways.....but it will in most cases improve machine ability, especially if you are dealing with a hard casting...
I agree with much of what you are saying. Not to take away anything from ironsides. He is the king of iron on y-tube, there can be no doubt, but there are some here doing some very nice iron work too, and not everyone gets the same results, so I think it is good to compare results when possible, and perhaps we can all learn more about iron, and hopefully reach a consensus about what works well and why it works well. At 11:45 he stated that he has found from experience about the riser on the flywheel hub, to avoid shrinkage defects at the hub, but I have photo proof that the riser on the hub is not necessary. This is the kind of stuff I am talking about, and I point it out not just to be obstinate or obnoxious, but just to mention that not everyone experiences the same effects, and others are doing exactly the same thing that he is doing with the same type of sand. "Rapid cooling" would be defined as a quench into oil or salt, or perhaps some sort of mechanism introduced into the mold to cool the metal rapidly, but pulling the casting out of the sand after 15 minutes definitely does not induce a chill in the metal. It may induce some internal stresses that could be alleviated with tempering/annealing, but no chills. It is not recommended that iron be chilled in water since the shock induces too much stress in the part. Did he have different shrinkages between petrobond and greensand or between greensand and greensand? Its really hard (for me) to tell how exactly what he is referring to. Is this a sand issue? Did it happen on plaques only, or flywheels only, otherwise you are comparing apples to oranges. Edit: Not to sweat the details, but with iron, the devil is in the details. .
rapid cooling will exasperate internal stress. a casting that sits to below 400 maybe fine while one that is broken out at 1097, may crack. the first can be greatly improved with an annealing process. though the best way to mitigate internal stress is in design, but there are obvious practical limitations. this applies to all metals, but is critical in cast Fe V/r HT1
I compare notes from several people I know who do a lot of iron pours, and one thing that is for certain is that iron seems to have a mind of its own. Often the smallest of details will cause a part to have a fatal flaw. Other times, a part will cast perfectly one time, and under seemingly the exact same conditions produce a failed casting the next time. I hear this even from the professional iron foundry guys, like from Germany. Iron does some strange stuff when it is cooling, such as contracting, and then expanding. The collapsability of the sand mold can be critical, as can be the rigidity of the mold, and how well it is held in place during the cooling process. There are a LOT of variables in iron casting. If you can get it sorted out and get consistent machinable, strong iron parts in a variety of shapes, with consistent hardness, and no chills, you have really accomplished something. You can dump a lot of phosphorus in iron and get it to flow pretty much anywhere, but that does not mean it is usable for a strong machinable part. .
PatJ Not all cast irons shrink as much as other cast irons so using scrap can be a bit of a gamble. Also adding ferrosilicon can alter the way cast iron solidifies.
I nearly missed this thread entirely and just noted it this morning. I have tried to observe carefully the effects of FeSi on cast iron in my foundry setting where I have cast the same patterns 20 to 30 times each over the course of several months. I have observed that large amounts of FeSi on the order of 1 pound in 50 pounds of iron does increase shrinkage compared to no added FeSi. I am using grey iron 40 returns purchased from a commercial foundry in all cases and poured at 1500 to 1500F. This difference in shrinkage is quite evident in the pouring basin/sprue and in the risers. These castings are 20 to 40 pounds respectively and are of moderate to relatively thin section. With no FeSi very little shrinkage occurs. With 1.5 pounds added to 50 pounds rather striking shrinkage occurs. Most recently I have been able to reduce my FeSi added to around 4-6 ounces in 40 to 50 pounds and still achieve uniform softness close to Rc 0-3 and experience 1/3 to 1/4 the shrinkage noted with larger additions of FeSi. I have unequivocally found that faster cooling results in increased hardness of my cast iron. That is not to say that the harder areas are not machinable. (Whether a casting is "machinable" or not is a rather crude measure of hardness as anything from Rc 0 to Rc 20 will be difficult to differentiate on the basis of machinability.) On the same casting I have used my calibrated Wilson Junior Rockwell hardness tester to examine various portions of the same casting. Areas that are at the periphery of the casting particularly those that are angular may measure as high as 15 when a more central area may measure 0. Obviously all that iron entered the mold at the same pour temp with the same mix of carbon, iron, silicon, manganese etc etc. the only difference is rapidity of cooling which is greater in the more peripheral areas. Wedge testing (in this case simply scoring and fracturing the casting) showed a nice fine grey crystal structure visually indistinguishable in the harder vs the softer areas. I have also found that addition of relatively small amounts of FeSi moments prior to pouring (many industrial consultants recommend adding the FeSi in the ladle rather than in the furnace to allow as little time elapsed as possible from mixing in the FeSi to actual pouring) to result in essential elimination of differential hardening. So that testing the same pattern casting but with late addition of 3-6 ounces of FeSi results in it testing Rc 0 at the periphery and at the mid portions of the casting. There are many many papers on the subject showing the effectiveness of added FeSi in preventing hardness including actual formation of extremely hard white iron. Here is one I just came across that looks a little bit at FeSi and grain structure and hardness: https://file.scirp.org/pdf/JMMCE20110300003_64896406.pdf The other factor that rapid cooling (early shakeout) clearly affects is internal stresses in a casting. The Navy published an article on this subject in 1948: https://apps.dtic.mil/docs/citations/AD0620556 So, early shakeout should generally be avoided to reduce retained internal stresses. Again this retained stress may not be obvious on machining as the metal will not be unduly hard. But it may confound the machinist as the part may spring between passes resulting in out of round parts or out of flat parts as various surfaces are successively machined. For the parts I am making retained stresses and variable hardness are a big deal. So, I try to be very careful of all the above factors. In addition, prior to final machining I do formally stress relieve my parts. Denis
I have considered this, but it seems like the mixing would not be even in small pours, and I think the late mix in the crucible that you do is the way to go.
Not knowing the amount of Silicon in your melt to start with is a problem. I don't think any of us have access to a Spectrometer. Rusty scrap will lower the Si. Your air/fuel mix in your burner could effect the Si. Also other elements, Mn, Cu, Cr, etc., will influence shrink. If you have a 100 pound melt and assume 0% Si in it, then about 2.56 lbs of FeSi (with 75% Si) would give you about 1.5% silicon in the final cast. Most FeSi alloys are 75% Si and you can expect an 80% recovery rate. This is all from my time at the Steel Mill but I think it would transfer over to Cast Iron. The formula we used was APa + B(.80Pb) / A+B = Pf I think I saw the same formula in one of the Navy Manuals. A=amount of Melt, B=Amount of FeSi, Pa=% of Si in Melt, Pb=% Si in FeSi, Pf = % of final Si desired. 100 x 0 + .60B / 100+B = 1.5% Solve for B... .60B = .015(100 + B)... .60B = 1.5 + .015B... (.60 - .015)B = 1.5... .585B =1.5.... B = 1.5/.585... B = 2.56 lbs If you wanted 1.5% in your 31 lb melt just take the 2.56lbs (41 oz) of FeSi times .31 = 12.7 oz. If I did the ounce calc right and didn't mess up any decimal points. ??
I guess it is difficult for me to understand why one would assume zero Si in the cast iron you are melting.(Maybe you were saying that at the mill you were using steel scrap and converting it to cast iron??) I know, for instance, that the returns I use have close to 2-3% silicon as I receive them---the foundry supplying it does have and does use their spectrometer and I did have them analyze spot samples of stock they sold me. And I would think it reasonable to expect most cast iron scraps or whatever to have a modest amount of silicon. Secondly, I think there may be two factors in play with silicon. One is a sort of needed steady state content of silicon necessary for proper balancing of iron and therefore a matter of just making sure adequate silicon is in the mix in the first place. But then I also read and been instructed by the metallurgist at the commercial foundry that has mentored me that for improved fluidity and prevention of hardness, silicon added just moments prior to pouring has best effect. So there is also some sort of dynamic effect that is transient following its addition to a melt. I observed that I could still get undesirable hardness (mild in most folks' book) in my castings even if I add fairly large amounts silicon 20 minutes to half an hour (should not have been timing it that way I now know) prior to pouring. This became evident as at one point I progressively added more and more silicon to overcome the hardness but I was getting up into the 1 .3 pound per 50 pounds of melt and started to run into shrinkage issues though hardness did come under control. But now, I can avoid hardness adding very small amounts like 3 to 6 ounces of 75% FeSi to 40 to 50 pounds of returns if I add it just before pouring. How these results will transfer over to other peoples' experience I don't know. Denis
All this information is good to know, especially for those who are trying to make machinable parts with good strength. Thanks to all who contribute their time and effort towards a better understanding of how to cast gray iron. .
We were making Steel from Steel scrap, the max Si we wanted was.20 percent. We would melt and hit it with an Oxygen lance until there was less than .01% Si. Silicon and Carbon are the first elements to oxidize out in an electric arc furnace. Some could be lost in a home furnace as well. You could start with 2 or 3% Si, but how much is left after the iron is all melted may be different. I would think 1.5% would be about the highest Si we would want for home castings. Your fortunate to have a starting Chemistry for your Iron. Most of us are busting up scrap bathtubs and such.
I have read that the high phosphorus gray iron was used for high fluidity for very thin parts such as radiators, and maybe bathtubs?, and an iron book say avoid using this scrap at all cost unless you are doing artwork castings, in which cast it is ideal. I may have mentioned this somewhere above already. .
Why did it say not to use Phos for cylinder castings? Thin castings probably have some Phos, thick castings are hard to bust up. Most any scrap is an unknown.
See page 227, "Stove-plate Scrap" https://archive.org/details/principlesofiron00moldrich/page/226 I also heard this from someone at the local Metal Museum who does a lot of cupolet iron using broken radiators. These work well for them since they do art work and the metal they get is very fluid. But remember that the proof is in the pudding. Undoubtedly someone will make a video to disprove the phosphorus thing. Without a spectral analysis, all of the iron alloying videos are just speculation anyway. If you can break the iron cleanly without chills, it drills easily, and appears to have good strength without being brittle, then use it for whatever, and ignore the composition of the scrap. A discussion of the term "semi-steel": https://archive.org/details/principlesofiron00moldrich/page/166 Edit: Page 103 talks about fluidity and phosphorus. https://archive.org/details/principlesofiron00moldrich/page/102 Page 127 talks about shrinkage, and what the sprue should look like with normal shrinkage. https://archive.org/details/principlesofiron00moldrich/page/126 Pages 128, 129, hardness and machinability: https://archive.org/details/principlesofiron00moldrich/page/128 Page 143 talks about how the correct usage of rounded corners and fillets: https://archive.org/details/principlesofiron00moldrich/page/142 Page 167, avoid sharp corners and drastic changes in section. https://archive.org/details/principlesofiron00moldrich/page/166 Page 174, use hard iron for cylinders and soft iron for piston rings. https://archive.org/details/principlesofiron00moldrich/page/166 Note that modern studies don't necessarily agree with this approach given high tech materials for cylinder and ring coatings, but I think it is a good approach for model engine work. .
