Casting a two stroke engine cylinder

Hello,

I recently found this forum and I have enjoyed reading it and learning new stuff about metal casting.
I have been metal casting for 2 years now. I have one main project that got me into metal casting: making a powerful cylinder for my race motorbike. I have cast many other engine parts too.

This is my second successful two stroke cylinder pour. I have cast them in iron which I think is rather uncommon for a hobbyist to do. I know many guys who have cast a cylinder in aluminium, but haven't seen anyone do it in iron. I decided to use iron because an aluminium cylinder requires a cast iron sleeve(with precision made ports) to be pressed inside it so I save a lot of effort not having to make the sleeve. Also a partial reason for using iron was that I was much more excited about melting iron than aluminium.



A two stroke cylinder has internal channels for passage of gases that make the engine work. In a simple low power air cooled two stroke engine there might be only 4 channels, but this one has 8 separate gas channels and a water cooling cavity. This is of course done in an effort to extract maximum power out of the engine. Much thought has gone into the design of the channels. There is a lot of science behind two stroke engine scavenging flow. The cylinder has only 50cc cubic capacity but if the channels are good and everything is setup well, it has potential to produce over 20 horsepower, engine turning at around 15.000rpm.





Making the cores (or the core boxes rather) was a crazy task but it turned out very well.



Some cores in the making.



I have made the core boxes by 3D printing. This is an excellent technique to make them, but takes quite a bit of thinking if the core is complicated.

I used epoxy glue as a core sand binder. I was expecting to get core blows but it seems that this is not a problem. I used around 2.5% of epoxy by weight, but the exact percentage varies as small batches are hard to weigh out. I have since bought a more accurate scale and planning to find the best mixing ratio. The epoxy cores are really easy to get out which was pretty much a requirement for these cores.

First cylinder (of this model) was defected:



I am not sure if this porosity is caused by core gas or just shrinkage. At first I was sure that it was core gas. Then I made the center core using sodium silicate and made a larger hole inside the core. That pour was successful and there was no large porosity there, just a small one but that wasn't too bad. But later I have started to think that it was shrinkage porosity and the second pour succeeded just by luck. I didn't put any thought on feeder design, I just stuck a simple feeder on the highest point of the casting, but it was far away from that point which got the porosity. If you have any thoughts about this, please share.

This is my youtube channel: https://www.youtube.com/channel/UCd8snQws0gTi8KbVdLNvl-g
I have videos about this cylinder casting provided with English subtitles.

I have cast my first cylinder a year ago and it worked, but it had much simpler channels and didn't have that much power. With this cylinder I wanted to make the channels as complex as they can be (actually can be more complex still), challenge myself with the cores and try to make it as powerful as possible. I have not yet put the cylinder into the engine as I need to cast many more ancillary parts and machine everything, but I expect to start it in a couple of weeks.

 

Excellent!

Very interesting work and great results. You are doing nice work.

Welcome to The Home Foundry.

There are so many things that can affect porosity. I don't know much abut iron casting but the few I have done have resulted in good castings. Did you include any ferrosilicon in your melt?

Your English is great, by the way.

 

Iron looks good...pretty sure those are core gas bubbles but without being able to hold it in my hand it is hard to say. It could be shrink but looks very smooth inside like a gas bubble from what I can tell. Nice job!

 

Wonderful work!

Please keep posting, it is absolutely fascinating to read of your progress,

Cheers Charlie

PS Just had a look at your Youtube movie about assembling the core. I notice that you have a lot of sharp edges on the core, which will mean that the casting will have sharp edges and this is not good for avoiding shrinkage.

I had problems for years with one of my castings; sometimes it would shrink, sometimes not. Then I read about avoiding sharp edges; so now I radius the edge of the sand-cores with an old hacksaw blade. Just a small radius of about 3mm has made all the difference – no more shrinks!

PPS Just watched your pouring movie. You can weaken sodium silicate with sugar in the mix – see Melterskelter's posts.

Also, if you vent the cope with a wire probe, before you seperate it from the drag, any gas can escape to atmosphere. A piece of 1.6mm wire works well; push it into the sand but stop short of the pattern if possible.

 

Welcome. That core is work of art!! I can't imagine how many hours you've got invested in the core boxes.

 

Wow, that is a real accomplishment. Congrats and welcome!

Jeff

 

Watched through a few of your videos. Super impressive congratulations on accomplishing all that you have

 

You have done some very fine work here. The complexity of the cores is amazing and the fact that they work is outstanding. I agree with OIF’s comments about radiusing all corners. I have not yet watched your vids, but intend to.

I can’t say what caused the voids. I have casted a couple hundred iron castings and I have casted and machined between 50 and 100 8 to 35 pound iron castings and have yet to see voids in the main portion of the casting. I’ve seen them on the surface from green sand that broke off and I’ve seen vacuum defects near the gates if I did not design my gates and risers well. Cast iron does not seem to have the propensity to form pores the way aluminum does.

Are you using filters in your pours? I ask because my first guess would be the voids may be due to slag inclusions. Before I started using filters (5 x 5 x 1 cm) I did have occasional slag defects on the surface of castings. No more slag issues since using the filters. I think the slag is of such low density it tends to float to a surface and not just freeze in the center of a casting.

I am looking forward to learning from your posts.

Denis

 

Thanks for your nice words!

I didn't add any ferrosilicon. I melted an old sewing machine (as luckygen1001 mentions they have good soft iron) and the cylinder was excellent to machine. For the second cylinder, I remelted some gating parts from the previous, added a little bit of Singer(about 500 grams) as I wanted to spare it and rest(about 1kg) was supposedly low quality iron from a barbeque cooking plate. In total the melt was 3.5kg. That cylinder was 99% excellent to machine as well, but I saw one spot of harder iron.

I have noticed the same thing, with iron porosity is nowhere near as bad of a problem as in aluminium.
I didn't use a filter but I would like to if I could get some. Seems the availability is like a 20kg box minimum, around here they only sell casting stuff for industrial foundries.

Sharp core edges could be one thing. The largest pore was next to a channel core, but I think it didn't have a particularly sharp edge. Interesting thing about this porosity is that it was on the drag side. In one of my castings, I got huge core blowholes because I had used epoxy as a core glue. These blowholes were all in the top part of the casting just below the surface. This makes me think that this porosity was from shrinkage and not gas bubbles. Also I read from Campbell's book that usually core blows are very large, but I wonder why there couldn't be a small core gas bubble.

 

If those are shrink defects they should occur in the portion of the casting that froze the very last. Without having the part in hand I can’t really say whether that’s the location likely for shrink defects or not but I don’t think so from what I can see. I suppose that bubbles from a core are possible but I also think it’s a little odd that those would be in the mid portion of the casting and not just under the surface. However these defects may actually be just under the surface with you having machined away the skin. It is an interesting question and it looks like not an easy one to answer.

I agree that in general it’s hard to find filters at a reasonable price in other than near case quantities. However I did see a near full case come up on eBay a while back and it was a pretty good price so I decided to go ahead and jump on it. I’m glad I did because I really like the filters.

I’ve never used epoxy for cores though I have toyed with the idea on more than one occasion. Do you recall what percentage epoxy you used to mix up the core? I think that Cobett recommends something on the order of 2% epoxy for an epoxy bound core.

Denis

 

Now that I think about it, the defects were probably 1-2mm at most below the skin that was against the core. Its just that I would expect a gas bubble to rise to the top part. Maybe the bubble was shot out of the core in a very late stage of freezing and couldn't move upwards any more.

This is a view of the casting from my 3D model. The red circles are where the porosity is. This side of the casting was in the drag, facing the ground basically.



I thought that it seems likely that the pores are due to shrinkage in a "remote" thicker section that got isolated. My feeder was not on that round part, it was on top of the bottom flange of the cylinder. But on the other hand it seems just as likely that they are core blows because they are simply near the cores.

In previous post I just said that the core didn't have a sharp edge, but I recalled wrong, actually it does(the core below the largest pore)

Yeah...impossible to say for sure. Just need to make more castings and see.

The epoxy percentage in cores. I used about 2.5%, but the net result will vary depending on how well you can weigh and mix the epoxy. I had an inaccurate scale(1g accuracy) and the epoxy that I used was very thick, which resulted in a lot of the epoxy mass being lost into the mixing spatula and scale plate. In practice, when I tried to measure 2% of epoxy, the cores were way too brittle. When I measured 2.8%, the result was workable but still a bit brittle and had to be very careful. I think the net amount of epoxy was 2.3%-2.5% then. With 3% or more, the cores are exceptionally strong and easy to work with, but I think there will be a great risk for core blows.

In Ironsides' video, he says that he used 1.5% of epoxy. I think the viscosity of epoxy must play a big role here. With 1.5% I had no hopes of the cores holding together, but I imagine with a more fluid epoxy the result could have been different.

 

Hi Asoftaaja, that cylinder is a work of art!. I have a similar iron casting which was going to be a cylinder with a flange on one end and using a polyurethane no-bake resin mould. Initially the casting looked fine but after machining it there were holes/voids similar to what you have. I'm having problems tuning my furnace to be hot enough for cast iron and as far as I can tell it was caused by the iron not being hot enough when poured and trapping gas from the pour. When I poured, I could see actual lumps in the molten iron pour.

 

The holes in the cast iron cylinder are from gas. The gas came from the burning epoxy. The binder balls in the sand cause localized high concentrations of gas. You should thin the epoxy with xylene, acetone, propylene carbonate or any commercially available epoxy thinner. This will eliminate the binder balls and you can use just 1.0% binder on the sand.

Regarding holes in castings, here are some guidelines.
1. If the hole is rounded, smooth on the inside and empty (nothing in the hole) it is from gas.
2. If the hole is jagged, rough on the inside and empty, it is from shrinkage.
3. If the hole is rounded, smooth on the inside and has a little BB in the hole, it is a Manganese Sulfide defect (only occurs in iron/steel)
4. If the hole is jagged or rounded and has something glassy in it, it is from slag.
5. If the hole is jagged and has sand in it, it is from loose sand grains.

Regarding shrinkage of cast iron, here are some thoughts. The shrinkage rate for cast iron depends on the Carbon content and the pouring temperature. If the Iron contains about 4.3% Carbon, we call it Eutectic iron. At this carbon content, the cast iron will have the lowest possible melting point (about 2100F or 1150C) and it will have no shrinkage. As the Carbon content goes down or up from 4.3%, the melting point goes up. When lower the Carbon content, the more shrinkage will occur. This is called Hypo Eutectic Iron. If you have more than 4.3% Carbon, the Iron will actually expand slightly as it solidifies. This is called Hyper Eutectic Iron and is very good for making pressure tight castings (like cylinders).

There are answers for every question!

 

Tom, is there any way someone casting at hobbyist's level would be able to estimate the carbon content of their melts? And, if necessary, adjust the carbon content?

 

Be sure to start with cast iron scrap that you know will probably have a carbon content if 3.2 to 3.5%. Most brake drums and discs as well as intake and exhaust manifolds should be in that range. Then be sure to melt with excess carbon floating on top of the iron. Use charcoal or graphite if possible. So long as you have carbon floating on top just before you pour, you can be assured of higher carbon content iron.

 

Hi Tom, I took a closer photo of a random void on my casting, there's small beads of slag in the void so it looks like it's number four on your list above. I added ferrosilicon right at the end and gave it a stir before pouring so it looks like I stirred slag into the iron.

I'm using brake rotors for my iron source, is it possible to heat the iron for too long and remove carbon?. I notice the iron pieces seem to burn and form a thick layer of oxides with sparks flying off just before the iron slumps and goes liquid.

 

Rocco, I adjust by pouring and breaking chill wedges. By adding carbon increaser or ferro silicon I can change the amount of free carbon for a given thickness and speed of solidification.
The thread for the Indian heads is lost in the AA fiasco and Photo Bucket but I needed the fluidity to fill the fins, but still be able to cut threads and the dogs for the domes.
Here are a couple wedges. In the lower one the white iron stands out, and in the other soft all the way.








The cages (the lower piece) that carried the intake valve is only about .120" wall thickness at the vents, but was still machineable.

And these are a head and the cores for the combustion chamber.






Not trying to hijack your thread. That is an impressive cylinder.

 

What carbon increaser are you using? Gorgeous heads.

Denis