A Bog Iron Ore Analog: (also cross posted to Hammered Out Bits)
A post this morning on Early Iron has lead me to jump the gun a wee bit, as I am working up a full report on this work for later in the month.
The team here (mainly Gus Gissing) have been working towards creating a 'bog iron ore analog'. One continual problem we have endured is wide variation in ore sources and qualities. Nature of the ore in turn effects the fine details of smelter design, and determines the characteristics of the blooms produced. With continual changes in ore type, it has been quite difficult to establish a predictable production from individual smelts.
Based on Gus's earlier work on 'Mars soil analogs', we took a look at what iron oxides were commercially available. Pottery supply companies sell a number of iron oxides as red pigments for glazes. Under the name 'Spanish Red' the material is is listed at 96.5 % Fe2O3 and 2 % silica. (I'm trying to trace down what the balance is). It is also available in a finer version through places that mix paints or sell base colours for printers inks, but is considerably more expensive in that form. It is also used to colour concrete, but I have not been able to track down some place that sells it in the kind of sizes we require. I expect the concrete pigments might be even cheaper.
The 'Spanish Red' is the cheapest of the straight Fe2O3 powders. It comes in 25 kg (50 lb) bags for about $50 CDN. The stuff we can get here in Ontario comes in a grit that is a bit finer than flour. As other iron experimenters have found, straight from the bag it is just too fine to work inside the kind of air blast our various smelter designs are producing. To get around that problem we are creating an artificial 'bog iron ore analog' (christened DARC DIRT ONE ).
The starting point is 80% by weight of the Fe2O3. We also were concerned that straight out of the bag it might prove too 'dry', so to get some extra slag production we are adding 10 % by weight of fine white silica sand. This is sold around here at Home Depot as 'decorator sand'. I suspect a plain beach sand would work fine too. The last component is 10% of plain white flour. This is basically inert, and serves as a binder. The powders are mixed by shaking in a large container with a (tight!) lid, then enough water is added to make a paste. The paste is then spread out to dry, our suggestion is in a layer about 1 /4 inch / 5 mm thick. The dry paste is then broken up with your hands to pieces roughly 'pea to peanut'. For the small batch tests I just spread the paste on a cookie sheet and set it under the wood stove for a couple of days.
From the looks of our test batches (one lb / 500 gm scale) I think you will want to run the broken material over a fine screen, say a window bug screen on a frame. Any of the dust can just be added into the next batch. The larger pieces are ready for your smelt.
This mix compares chemically and mechanically pretty closely to the samples of natural bog ore I dug in northern Newfoundland. (I say this knowing quite well that 'bog ore' can look a lot different from region to region!) At this point that assessment is based on the mark one eye ball, we have not done any detailed testing of the material we have come up with yet. One potentially huge advantage to the use of dry oxide powders exists for those involved in detailed archaeological reconstructions. Other small amounts of oxide powders (many also available at pottery supply) can be made to the base mix to simulate the specific ore contents from a given geographical location.
Gus (through his business Harder Gissing Machining) has donated enough of the oxide for two production batches. I picked up the the materials over the weekend. Todays work is to mix up a full production batch of DARC Dirt 1 in preparation for a full smelting test. As we are completely buried here under snow, with temperatures below freezing, I will not be able to just spread the past on a plastic sheet and let it air dry as I would hope to. I bought a bunch of thin aluminum foil 'oven liner' pans and will spread the paste on those and stack them in the kitchen oven to dry. Figure (hope!) a couple of hours at 200 F will do the trick. My aim is to prepare 25 kg (50 lbs) of the analog to use in a smelt next week with Skip wen I am down at Smeltfest at Lee's place in Virgina.
So expect a full smelt report on the use of the analog later in the month.
The post on Early Iron had mentioned using ground magnetite. This is sold at the pottery supply as 'black iron oxide'. The chemical form of this is Fe3O4. It is also a very fine power, and is actually half the cost of the red oxide. We intend to also make a full scale test of this source material, using a similar mix to paste, dry and crush. The chemistry of the black oxide is a bit different, which may or may not influence a smelt.
Using the Spanish Red, the cost per smelt (a bit over 25 kg / 50 lb 'ore') will work out to about $55 CDN - roughly $1 per pound. The huge advantage is that this material is easily available and will be standard smelt to smelt.
Labels: iron smelting