Looking for info on Sodium Silicate Mold/Core Washes for Cast Iron

I am getting very close to pouring cast iron into sodium silicate-bound olivine sand molds. I am nearly certain I will get burn-on of sand if I don't use some sort of wash. I am thinking of trying a mix of dextrin, plumbago and isopropyl as a spray-on wash. But I have no recipe for such and don't even know if those are good candidates for ingredients. I have found some washes in the Naval Foundry Manual and on Sandrammer's videos. But they all seem to be aimed at green sand applications. I am willing to just start from scratch, but would love to not have to reinvent this wheel.

When casting iron in green sand I have found sea coal to be the answer as an additive to the face sand to prevent burn-on. My sodium silicate-bound molds do not incorporate coal as I need maximum strength (I think) for my application.

Any help would be appreciated.

Denis

 

I have seen some foundries or people brush on dry graphite to the interior of the mold.
I have some graphite and I am gong to try that, but I will have to make two identical molds I guess; one with and one without, to get a accurate comparison.

 

Well, brushing on powdered graphite would be easy enough. I have a pound sitting on the shelf. My "control" sample could be one of the runners, with the other runner brushed and the mold itself powdered. Then there is the question of how much to apply...

Denis

 

I am going to try a light brushing, with no excess to fall off the walls of the mold.
Just guessing at this.

I have also noticed a difference between the finish on the top and bottom of an iron part (without graphite).
One or the other is usually better but I forget which.

 

For me, it has been the upper portion having the better finish on iron pours. I attributed ( this to the mold filling from the bottom. As the metal flowed in more, it had cooled a bit and so had less tendency to burn on. This is most obvious if a mold pours short. The very last metal to enter the mold is just above freezing and leaves a nice shiny surface.

Denis

 

Try Graphite and Western Bentonite equal parts suspend it in alcohol spray it on and burn it off... start with equal parts if it is settling out add western bentonite... if you can find silica powder that can be used in the mix, but I seriously doubt you can find it anymore.

V/r HT1

 

Thanks, HT1. Could diatomaceous earth be used in place of silica powder?

Denis

 

When I cast the 1914 Indian heads I used graphite lubricant spray from NAPA to coat the linseed oil bonded molds. I had a noticeable improvement in the surface finish of the cast iron. With petrobond molds I dust with locksmith graphite in the little squeeze bottles and then backprint the pattern. For hand cut gates and runners I use the slick and oval to smooth the surface. This also seems to improve the finish of aluminum castings.

 

Thanks to all for their responses.
The graphite spray sounds convenient to use.
I wonder if the bentonite suggested above might lend some smoothing effect to the mold surface.

Denis

 

yes, but it can effect finished dimensions if they are critical

 

OK, taking the various comments above into consideration, I mixed up some isopropyl and powdered graphite into a between milk and cream consistency and painted it onto my mold. I did leave one runner unpainted to look for a difference in surface. I did rub out brush marks with my fingers. I will find out in a day or two if I have wildly too much graphite or not. I am concerned that the graphite my close up the surface of the mold preventing proper breathing of the mold. To that end, I did drill four 1/16 through holes in the upper reaches of the mold to allow gases to exit. The mold is otherwise vented from the outside to within a 1/4" or so of the cavity in many places. Will that be enough??? Time will tell.

Even more important I will find out if the silicate/bentonite/diatomaceous earth glue holds the two halves of the mold together and whether the mold itself has enough hot strength to not break out. And I will be attaching a sprue to the basic mold with silicate/bentonite/DE glue as well. Will that hold?

Then there is the question of whether all this fooling around will all result in a softer thin-walled casting or not. I intend to "bury" the mold in a perlite containing box to act as an insulator to slow cooling.

The hope is that the relatively low mass dry mold and perlite will significantly slow cooling.

It seems almost certain that some parts of this setup will fail. But, hope springs eternal as they say. ;-)

Denis

 

Even though I cement my resin molds together, I still use weights on top of them because sometimes the cement will break and the metal will run out.

I have not had a sprue failure (using a cemented-on elevated sprue).

I assume you are going to use ferrosilicon?

 

Yes on the FeSi just not sure of amount

What are you using for cement and how long do you bake out the glue? From some testing I did, I think I need to bake the glue line at least 3 hours as the glue line itself seems to set in
less time, but moisture migrates a mm or two and needs to be eliminated.

Denis

 

I use Rutland furnace cement, and I use it on the outside edges of the mold only.
It does not take too long to dry.
I keep it to a very thin layer; just enough to seal the mold; so that it dries quickly; in perhaps an hour or so; air dried.
A few photos attached.
Iron is much heavier than the aluminum used in these photos, and so you have to probably use a bit more weights.
The wood strips are just dams to keep the very fluid aluminum from running down the side of the mold when it comes out the vents.

It is not a strong joint, and thus the metal weights on top of the mold.
I don't use cement in between the molds, but rather only along the edge of the closed mold.

Scavenger (the guy who made and sold all those antique hit-and-miss mufflers) uses approximately 0.75 oz of ferrosilicate per pound of iron.
I think it would be a good idea to use an amount that has been proven to work well for somebody else, at least as a starting point.

 

Thanks for the above info, PatJ.

I intend to use external pressure to hold the halves together, but also IMAGINE I need a cemented joint. I will be packing the mold in a wood box with pearlite packed around the mold and pressure applied to the pearlite as the top of the box is screwed down. How this will work in practice remains to be seen.

Those are nice looking molds and appear to be pretty thin----maybe an inch or so?

I see the Rutland furnace cement is rated to 2000. So, I am not sure how reliable it might be for iron.

Denis

 

The pearlite and box thing sounds like a lot of unnecessary work, but I don't know exactly how your molds are configured, so perhaps necessary.

The resin molds are very strong, so they can get pretty thin, and mine are perhaps 1/2" thick in places. (more photos below)
Resin is expensive and difficult to acquire, so I am transitioning to sodium silicate and/or water-based green sand.

I have seen commercial foundries apply a bead of sealant on the face of the molds, but for small parts like mine, that can distort the part dimensions, plus anything added to the face of the molds generally does not get fully compressed, so it can cause a lot of flash around the part.

The Rutland cement is not rated very high in temperature, but you have to keep in mind what its purpose is, which is to stop any leaks/gaps between the mold halves from draining the mold cavity (been there and done that), and to hold the mold halves together while you are handling them.
By the time aluminum or iron travels across the face of the molds and hits the cement, it has generally cooled to the point where the cement will stop the flow.
The cement may smoke, but it will not immediately fail if exposed to metal, and it will last long enough for the metal to solidify.

The bound mold is the same way. It only needs to last long enough for the metal to solidify, and there is significant scorching/burning of a resin-bound mold during an iron pour, but that has no effect on the casting quality, and there is no sense in wasting expensive resin and sand. Bound molds in general should be as small and as tightly conformed to the pattern as practical.

Some parts of my resin molds may get as thin as 1/4", and that works fine as long as the area is not too large, and as long as the mold still has enough strength to be handled.

I have used sodium silicate with brass/bronze, and it degrades more quickly and more completely after a pour (with sprue extenders often breaking into pieces after a pour), but again, the mold only needs to last for perhaps a minute or so (iron takes a little longer to solidify), and if the mold fails after the part has solidified, that has no effect on the casting.

You can see in the first photo that the 1" thick cope is plenty strong enough to support the significant part that protrudes down and acts as a core.
Sodium silicate molds are not quite as strong as resin molds, but I don't anticipate changing the thickness of my molds due to using sodium silicate, although I will handle the molds a little more carefully.
The trick with sodium silicate is to not over-gas the mold with CO2 (if using that process). They do have SS products with a hardener that self-harden without CO2.

The resin molds have to be broken up with a hammer after a pour.
I use a hammer to break up the SS molds too, but they break up more easily than the resin ones.
I do intend to recycle the SS sand. The resin sand requires burnout, and so I have not recycled it.




Here is an example of not quiet enough cement, leading to a runout.
I still got a usable part out of this since the entire mold cavity was located in the drag.
I started using weights after this mold.






With aluminum sometimes you can get away without using weights, but generally I always weight things now.






Weighing out the ferrosilicon:






These molds are thicker than they need to be for a resin-bound mold, but they worked:
I had problems with the Adolf's alignment inserts, even though I used a lot of care when drilling the holes for those.
I ended up omitting the Adolf's inserts from future molds, and use two dowel rods stuck into the sand for exact alignment.

 

Well, my attempt at gluing the Sand Sodium Silicate cope to my drag failed to provide the strength I had hoped for. I had dried both by cooking them at 225 or so for 4 hours. Then I mixed up a slurry of diatomaceous earth and bentonite with SS Type RU thinned 15% with water to a consistency of something like common yellow glue or perhaps just a bit thicker. I spread a 1/16" to 1/8" layer on the mating surfaces and then put them together carefully and recooked the whole works for about 3.5 hours at 225-250. The glue hardened up to a rock-like firmness. But as I took off the wood frames holding the cope and drag the weight of the mold stressed the joint enough to cause separation. The glue itself did not fail but the sand of the drag failed a mm or so away from the glueline. I do not think I overgassed the cope or drag---I have made that mistake in the past and recognize the crumbly sand result. I just think the sand bond is affected by the gluing process to weaken its bond near the joint line. Could be a lot of things like differential stretching and contracting due to moisture in the glue migrating into the nearby sand. Could be just inherent lack of adequate strength in the sand and a zipper-like failure of the sand bond. Regardless, I cannot rely on the glue to hold the cope and drag together. It would have done a nice job of filling any voids between the cope and drag though.

So, next is just to mold a cope and drag and rely on weight or clamping or both to prevent the cope from floating. On my mold, it looks like there would be something on the order of 50 to 70 pounds of float as the area of the mold is 54 sq inches and the head pressure from the sprue will be on the order of 3-4 inches of iron at about 0.3 pounds per inch of head.

Mulling whether to simply use weights as PatJ does or to also clamp and weight it. It would not be hard to crack a cope or mold by clamping. Weights could also cause problems if the mold is not evenly supported.

Here are a couple of pics of the results of my efforts. This attempt at making SS-bound molds has tested my patience!:



Edit: the black areas on the Cope and Drag are areas "painted" with powdered graphite to prevent burn-on as discussed in above posts.

 

Sodium silicate bound sand is not as robust as resin-bound sand, and especially if it is not cured quite right.

I would not try to adhere the cope to the drag using cement on the mold faces (some do, I would not).

I have tried weighting the flask, and also clamping.
The clamping works ok if you have an elevated flask support which gives you room to get under the flask.
I found the clamping to be such a pain that I just use weights, but for a large enough mold with iron the uplifting forces are significant, and the clamping may make more sense, especially for repeat castings of the same part.

With resin-bound sand, I often make twin molds, and furnace cement the molds together to cast multiple parts at the same time while having only one pattern.
So what that means is that if a resin mold (and perhaps a ss mold) breaks in a non-critical area, just glue it back together with furnace cement.
There is really no limit to the shapes you can combine with furnace cement when using bound sand.

You could try the ss that hardens with a hardener, thus giving a consistent hardness to the sand.
That is a pretty big mold.
I think this mold could work with ss sand.
I have seen much larger molds made with ss sand (36" long or longer molds, and perhaps 24" tall), but again the larger molds lend themselves to the ss/hardener approach for consistency throughout the mold.

And I have seen people use a combination of ss sand and green sand, with ss sand next to the pattern, and backfill with greensand.
I think I would go either/or, but you could save a lot of ss using the combo method, which could be critical in a money-making operation.

This is what I bought, but I have not used it yet.
http://shop.clay-planet.com/chembond-4905-no-bake-binder---5-gallon-bucket.aspx

The hardener comes in various speeds, but Clay Planet only sells the 210 unless you want to buy 5 gallons of hardener.
The hardener they sell (they call it catalyst 210, but it is really a hardener) has a set time of 75-85 minutes, which is a very long time for an attention-deficit person like myself.

The difference between hardener and catalyst is that set time does not vary with how much hardener you add, such as with the Chembond 210.
Set time does vary with a true "catalyst", and that is why Chembond 210 is not a "catalyst" but actually a non-time-varying hardener.

The guy who does the motorcycle cylinders on AA found some of the faster Chembond catalyst, but I did not get his supplier.

I really liked the resin-binding material that I bought, but the toxicity is more than I want to deal with.
But the resin-binder I was using had a resin, a hardener, and a catalyst, and the resin and hardener were always mixed on a fixed basis, with the catalyst being used to vary set time.
I had a set time dialed in to about 10 minutes, which allowed almost continuous mold making (mix, ram, set, pop off the snapflask, repeat).

You should consider a thin application of furnace cement on the exterior of the joint to prevent runout, and use weights to keep the mold halves together.

Edit:
And if you use the Chembond ss material, don't leave the pattern in the mold beyond the strip time, else your pattern may be glued permanently in the mold. And waxing the pattern each time is a must to be able to get the pattern out of ss-bound sand.

Edit02:
Per the Chembond data sheet below, there is a trade-off between mold hot strength and shakeout (I assume they mean "shakeout strength"). The Chembond 4905 binder has a hot strength of 2 (with 1 being the best), and a shakeout of 4, so I interpret that to mean that a Chembond 4905 mold is quite strong, but requires more work to break up and clean off after the casting work is complete.
I would much rather have a mold that is plenty strong so I can easily handle it without damage.
I have used ss sand with too much ss, and that is no good either because it hardens like granite, and is almost impossible to break up after casting.

 

Here is the cut sheet on the Chembond stuff.

They call this stuff a "no-bake" solution, so it seems a waste of time to bake a no-bake mold.
I would just flame the interior lightly with a propane torch, or at least try that approach, which I think is what the manufacturer recommends.

The information also states that you can use additives to the sand with Chembond binder.

Edit:
There is a lot of information in the pdf file, including information about alcohol-based coatings, etc.
SS bound sand can be used for both cores and molds. I always vent cores through the center and out the top of the mold.
The instructions mention that the mold is still hardening after you strip the pattern, and with resin-binder I found that you need to lay the molds on a very flat surface for about 10 minutes after strip time else the mold will not quite be hard enough and will distort, even though the mold feels hard after the set time.