Casting a Model Westinghouse-Type Twin Steam Engine in Iron

[vtsteam]

Yesterday I cast a block for a model Westinghouse style twin in iron. I'd tried this a few weeks ago with some questionable iron using an old pattern and without cores, just as an experiment to see if the iron would cast well. This second time around, I made a new corebox, and a baked sand core (mentioned in another thread). This time also I used brake rotor metal, which I'd tried successfully earlier in casting a lathe chuck back plate.

My furnace has been acting up a little for the last two casting sessions -- not sure of the reason why other than repairs to the lining lately, but I'm beginning to suspect the jet in my burner might be partially obstructed. Symptoms are, the furnace is harder to light, and  slower to melt. I was earlier getting phenomenal melt times of 20 minutes for a pour of this size. Now it's more like 40  minutes -- still reasonable, but I'd like to know the reason for the difference.

The other possibility is that disk rotor metal takes longer to melt. Or a maybe a combination of all of the above.

Anyway, I made up the mold and placed the core in it without any hitches. The facing sand was a little dry (I should have checked in advance). So there was a little breakaway/flash at the seams, but no biggie.

The melt took much longer than expected, as mentioned, and I took some pains to slag the melt twice and thoroughly. Perhaps too long with the crucible out of the furnace, or too cool in the melt, because when I poured the sprue filled and stopped accepting metal before the riser was full to the top. I don't think I've ever had that happen before.

But I was still hopeful because the riser had filled to well above the casting's height.

Here's the breakout about 4 hours after the pour. Notice how well the cylinder cores in the block have worked, and burned to black sand:

[vtsteam]

I was pretty hopeful, the detail looked good and the core had worked perfectly:

[vtsteam]

But nope, the metal just wasn't hot enough to fill the thinnest area furthest from the sprue. The riser had been closer to the sprue and in a thicker part of the casting, so it filled. I'm pretty sure this would have been a perfect part if the metal had been hotter. Oh well, next time.....

I'll relocate the sprue and give it two gates to either end of the thinner part, plus, of course, pour hotter.

[tom osselton]

That is pretty close to working.

[awemawson]

Do you have any way of checking the temperature of the melt Steve?

I used to use a ‘disappearing filament optical pyrometer’ ,  basically just an incandescent filament run through a rheostat   Tweak the rheostat until the filament disappeared against the melt and read off the temp from the rheostat dial. Sadly when I dug it out a few weeks ago I found that it had succumbed to dampness and ended up chucking it out. 

[vtsteam]

Andrew, I've actually got instructions for making one of those, which I downloaded awhile ago. I haven't made it yet.

But there's also the Terry Aspin method, which actually works if you pay attention to it strictly (I didn't this last melt because of a dumb belief that it had taken so long it MUST be ready) And that is, just look at your 1/4" dia steel stir stick. If you give the pot a poke or stir and there's a glob of slag or metal stuck to the end, it's not ready. If it comes out perfectly clean and coming to a sharp point like a pencil because it melted a little, well, time to pour!

Steel melts at a bit higher temp than iron, so it's a simple perfectly straightforward thermometer. You do have to pay attention to it though, for it to work.

[tom osselton]

That’s what it says to check the temp in the metal casting bible by C.W.  Ammem

[vtsteam]

I had a point and a gob. Shoulda just been a point. I now think the point was from the prior melt. Gob shoulda melted off, point shoulda been shiny and sharp
Lesson learned: No gobs, period.
 

[ironman]

[quote author=vtsteam link=topic=13496.msg162471#msg162471 date=1
If it comes out perfectly clean and coming to a sharp point like a pencil because it melted a little, well, time to pour!
Steel melts at a bit higher temp than iron, so it's a simple perfectly straightforward thermometer.
[/quote]
A low carbon steel rod will quickly absorb carbon from molten cast iron and it will lower the melting point so that is the sharp point that appears after 10-20 seconds  of stirring. If you hold that stirring rod above the crucible for 10-20 seconds it will not melt because the flame temperature is not hot enough. I do have a R type pyrometer and the final temperature before pouring is 1550C-1570C above the crucible which is just above the melting point of a low carbon steel.

[vtsteam]

Ahhh, interesting, Ironman, so you're saying that the best place to check (without a pyrometer) would be just above the crucible, rather than stirring the molten iron itself.

[ironman]

Pushing a steel rod into molten cast iron is still the best way to test for temperature just do not stir it otherwise it will dissolve. If you pull out the rod and slag, iron sticks to it your iron is too cold to pour. Another way to test the temperature is to look at the crucible to see if is brighter than the molten cast iron, if it is it needs more time to get to pouring temperature.

[vtsteam]

Oh, okay, great tips Ironman. I'll be sure to check both next iron pour.

I have a couple questions about slagging. I know you slag in the furnace at least once -- I'm guessing just before adding ferrosilicon. What is the end of your slagging tool made out of? It looks relatively thin, and must be some heat resistant material -- stainless steel?

I find that in my furnace, the ITC HT-100 reflective coating soon gets covered with a brownish glaze -- I'm guessing from vaporized slag, mainly from slagging in the furnace. A little slag always escapes the slagging tool.

The brown slag glaze on the furnace walls reduces the reflective property of the HT-100 and melting time goes up from the initial very fast melts. I think my first few melts were 20 minutes long, now that are around 40 for the same 6 to 10 lb melts in an A6 crucible.

To try to see if I could improve that, I've re-coated the furnace with Satanite as a base over the blanket, and then a paint coat thickness of ITC-100 over that.

But, because of the slag glazing problem I'm wondering it the ITC 100 really is worth it on a foundry furnace (as opposed to a forge where people use it a lot), and maybe just the Satanite would give about the same results after the first few melts. It's much less expensive and also seems tougher and more heat resistant than the ITC-100.

Finally, speaking of looking at the crucible and iron to compare temps, I'm curious about what you use for eye protection. I see you use a face shield with a filter -- I've tried filters and it seems like a shade 5 is too dark for safe maneuverability, so I'm thinking of a shade 3 filter. What do you find works for you?

Thanks for all your advice and information over the years! I still have a copy of your early website from nearly a couple decades ago when I first saw it, where you were melting with oil, and also a cupola.

[ironman]

Yes you are right it is stainless steel, 2mm thickness 40mm wide and is bent to a simple "L" shape. Welded to it is 6mm dia rod about 2 meters long. I do scrape off the slag just before adding ferrosilicon but I try to keep my hands as high as possible on the steel rod to keep them as far as possible from the exhaust vent. I try not to put the stainless too far into the molten cast iron as chromium absorbs carbon many times quicker than steel does. After about ten melts the stainless steel has absorbed enough carbon to make it brittle and also looks like swiss cheese so I cut that piece of and rebend another "L" shape. The stainless scraper part starts out at 200mm long but ends up a lot shorter after many melts.

I have never tried ITC HT-100 to coat kaowool and the same goes for satanite but I have heard a lot of good reports about satanite. I still have half a 25kg tub of zircoat and would love to try satanite but it will not be happening soon.

My face shield is a shade ten filter and is great for working around the furnace but when I get into my dark shed with my pouring trolley it is a blackout, way too dangerous. So before I push the trolley into the shed I put on cheap safety sun glasses which has a light tint and a clear face shield. I have tried lower filters but my furnace gets so hot and the glare hurts my eyes plus they are briefly dazzled which can cause a nasty accident. Here is a tip when using a shade ten filter, as soon as I don't need to look at the furnace I lift up the shield to see where I am walking.

That was not my website, I was featured on another metal casters website who took all  the photos.

[vtsteam]

Thank you again, Ironman.  I think the shade number on some of my headgear is off. I have some gas welding/brazing goggles from Harbor Freight that claim shade 5, but they are darker than the Shade 9 setting on my same brand auto-darkening arc welding helmet. Hard to see anything out of the goggles except the furnace barrel. Too hard to maneuver,  even outside

Actually I've discovered that the auto darkening helmet seems to work well for me in iron melting. At the lowest setting, it dims the open furnace enough not to be blinding, and when the furnace is closed it reverts to a light green shade that seems right for carrying the crucible to the mold.

I have a couple more questions for you if you have the time:

First, do you adjust your furnace for a neutral flame, or do you vary that a little either way?

Second, I remember that you use copper wire to lower additional iron into the melt. I'm guessing transformer wire from scrapped gear. Is there any advantage to copper wire vs thin ferrous wire (like thin uncoated florists iron wire -- which I happen to have -- I don't have any scrap transformers).

Thanks!

[ironman]

When melting iron I adjust the flame to be slightly reducing.

The copper wire is from the field winding from a car alternator. I found through trial and error that if the wire is too thick it takes too long to melt so burning my fingers even though I have gloves on. If the wire is too thin there is not time to aim for the crucible and the wire melts too quickly.

[vtsteam]

Thanks again, Ironman. 

Ah,slightly reducing probably explains why you have so little loss in your pours compared to mine. 

I take it there is no problem with a small amount of copper in the iron melt. I guess the wire probably mostly vaporizes anyway. I'll have to experiment with the wire types I have and see what works.

If you get a chance, could you measure the diameter of your alternator field wire? That would helpful as a starting point for finding similar copper wire.

I'm going to give the Westinghouse engine block another try when I get another core made up.

[ironman]

The wire I use is 1.15 mm Dia. Small amounts of copper do not matter in a iron melt. I look forward to seeing you to recast the engine block.

[vtsteam]

Great, thanks so much I'm glad you measured it as I imagined it was quite a bit finer. Now I have a good starting point. Here that would be 17 ga. AWG.

I'm re-thinking the corebox. The plaster of Paris version has a crack in it -- I can repair it, but I'm thinking now about fabricating or casting one in metal. Since I'm using a baked sand core for this engine, it would be convenient to bake the core in a corebox, or corebox half, unless that causes release/clean-out problems. Or I could  make a core plate for that, since I'm going traditional. Then the corebox could be wood, which is easy for me------- well, still thinking about it.

It's interesting to think about....making a corebox from the solid has pretty much the same shaping considerations as making a die, the alternative being molding it with some compound from a plug, eg. plaster, etc

[vtsteam]

I repaired the plaster corebox and baked a new core in preparation for another go at the small Westighouse steam engine casting. This is really a test piece, rather than an engine I want to build at present. But it's an interesting challenge, and it helps me understand how to design a better engine pattern, core and corebox, and I can also try different types of scrap iron I have on it.

To me the ideal corebox for baked sand cores would be made of aluminum so you could not only mold it in the box, but bake it there as well. And the best aluminum corebox would be relatively thin around the core -- say 1/4 or 3/8 inches thick (6-9mm). That way it would heat up quickly, and also use the minimum amount of aluminum.

Plain Plaster of Paris is fairly resistant to heat, but is still rather delicate, and the temperature must be kept around 200F if the Plaster is to last for repeated use, when ideally you'd be able to bake at 350F. Also plaster coreboxes like the one below, poured into a containing box, are massive. That bulk plus the fact that plaster is an excellent insulator means extended bake times for the core.

I would think that an aluminum core box would bake a small core out like this one in about 15 minutes in a small toaster oven at 350F. The existing plaster box took about 45 minutes, and I still had to finish the ejected core for ten minutes at 350.

[vtsteam]

After replacing the furnace lid with a new one built from scratch, I did a melt this morning. I think everything went right as far as I could tell, plenty of metal and it was hot enough. I melted 7 lbs even thought the casting only needed about 3. I useddisk brake rotor, and 11 grams of ferrosilicon was added. It took about a half hour to melt from a loaded cold crucible, and then another half hour to add more iron, slag the melt, add ferrosilicon and get it hot enough to pour.

Not sure why my earliest melts with this furnace this year were about half that time, but I'm starting to suspect the iron itself. I think the first iron I used was an easy melting variety, and the latest disk rotors are not. That might fit with disk rotor metal tending to chill easier, increased toughness and hardness. Maybe rotors are lower carbon. I dunno.

There was a fair amount of really thick chunky slag in this melt. Disk rotors in Vermont get a lot of salt corrosion --especially in the hollow ventilated part.

Anyway, here's the pour....nice and neat. The ingots, less so -- I need to level the base here better...

[vtsteam]

While I'm waiting for it to cool, I broke one of the ingots on half to see what it looks like inside. Well, nothing like I've seen before...

I don't really know what I'm looking this time around for cast iron -- it's very uniform and silver, not white and not gray.

It isn't really chilled because I can file it, though it is tough. It was VERY hard to break. I had to go 4 tries with a sledge. Also it rang when it fell on concrete, In other words, it sounds like white iron, breaks like semi-steel, files like somewhat hardened steel.

Some new kind of animal....

The photo just doesn't show the shiny silvery quality...

[vtsteam]

The shake out. Looks like the core did its work:

[vtsteam]

Free of the sand the casting looks good:

[vtsteam]

I cut off the runners -- they were in awkward locations and didn't respond well to a hacksaw blade -- chilled at the junction, so I used an abrasive cutoff wheel.

Otherwise I found that most of the casting filed well, with the exception of a couple of small chill spots. I took off most of the flash with a file, and roughed in some other areas just to see how workable it was.

I think I'll try annealing the whole casting tomorrow. No idea if this will work with this disk rotor metal, whether the general hardness is an alloy contribution, or whether like other types of cast iron, it will respond to annealing....guess we'll find out tomorrow....


Anyway, success on the third attempt at the Westinghouse Twin casting!

[tom osselton]

That is a good casting I would expect it to anneal with no problems.
Would coating the pattern with graphite smooth out the casting?