Concrete—”Liquid Stone”—has made possible many innovations in architecture. Yet concrete is no Space Age wunderkind; it has its roots in antiquity. Concrete, albeit of a kind inferior to the modern product, was used by the Romans in the construction of the Pantheon, which has endured since the time of the Emperor Hadrian.
While the Roman concrete structures endured, concrete technology languished after the fall of Rome. The seventeenth century is still the Dark Ages so far as concrete is concerned. But the up-timers will bring about a “Concrete Renaissance” in short order.
What Is Concrete?
Concrete is a composite material, made by combining an aggregate (a hard particulate material) and a cement (a matrix forming material) with enough water to cause the cement to set and bind the aggregate together. The binding is the result of the chemical reaction of the cement with the water. The aggregate is a combination of fine aggregate (sand) and coarse aggregate (e.g., gravel).
Mortar is a paste-like mixture of sand, a binder (e.g., cement) and water. It doesn’t contain coarse aggregate, but of course the mortar is used to bind together cut stones or bricks, and to fill in gaps between them. Those stones and bricks are much larger than the coarse aggregate of concrete.
Concrete and Cement in Canon
We know that when Grantville made its involuntary journey into seventeenth century Germany, some concrete construction came along for the ride. Mark Huston’s “Gearhead” (Grantville Gazette, Volume 9) mentions “a pair of concrete bridges.” There is a concrete floor in the building which Chad Jenkins has converted to a shop for washboard manufacture, see Rittgers, “Von Grantville” (Grantville Gazette, Volume 7). There is also a concrete floor at the farm where Harmon Manning suffered his ultimately fatal fall, see Ewing, “An Invisible War” (Grantville Gazette, Volume 2). Pam Miller has a concrete porch, see Vance, “Protected Species” (Grantville Gazette, Volume 13). The high school has a concrete “awning” over the entrance, see Flint, 1632, Chapter 11. And there is at least a concrete slab in the Grantville city jail, see Weber, “The Company Men” (Grantville Gazette, Volume 2).
Concrete was used in the displaced West Virginia mine featured in Mark Huston’s “Twenty-eight Men” (Grantville Gazette, Volume 10), in a wall separating the working and non-working sections of the mine. The wall was built out of concrete blocks, and thus, even if the wall was assembled after the Ring of Fire, the blocks themselves may have been cast up-time.
Since the Ring of Fire, there has been some new concrete work. Sometime before March 1632, Delia Higgins sold the remaining dolls in her collection, and used the proceeds for two projects. The first was building a warehouse. Her intent was to build a concrete warehouse, a “work of art”, with “the best combination of up-time and down-time construction techniques possible.” Gorg Huff, “Other People’s Money” (Grantville Gazette, Volume 3)(timeframe March-October 1632). What she got was, “if not exactly a work of art,” a structure which “was functional, and very large.” It was built with “fairly standard down-time construction techniques, with concrete pillars added for support.”
In the process of trying to persuade the high school chemistry teacher, Alexandra Selluci, to help with the warehouse project, Delia got talked into becoming the “sugar grandma” for the Grantville High Tech Center’s “brand new concrete research program, complete with structural engineering courses where the teachers were half a chapter ahead of the students, or sometimes half a chapter behind.” In Delia’s opinion, “the kids that had gone into concrete were phenomenal. They were about four to one down-timer to up-timer, about average for the high school. They wanted to build things. Great big things, dams, skyscrapers, and roads, and were willing to work at it.”
Later in OPM, Delia reveals that she saw the warehouse as a stepping stone to a grander project, the Higgins Hotel. “The concrete program at the school was developing a group of young people who could make structural concrete, and form it into structures that would support tremendous weight. Hiring Michel Kappel was done both to get a down-time builder familiar with up-time building techniques, and as favor for Karl Schmidt. Claus Maurer was a master builder with more experience than Herr Kappel, but again, part of the reason for hiring him was to get him familiar with the available up-time tech. It wasn’t her fault that they had fought with each other and with the teachers at the tech center and Carl over at Kelly Construction. Besides, materials were so expensive that the cheapest halfway decent material was quarried granite from the ring wall.”
By December, 1631, the high schoolers are working on “some sort of concrete project,” and mortar is available. See Huff and Goodlett, “Birdie’s Village” (1634: The Ram Rebellion). In Cooper, “Stretching Out, Part One: Second Starts” (Grantville Gazette, Volume 11), there is a passing reference to an equally mysterious “concrete project” which is apparently looking for venture capital in July 1633.
By July, 1633, the conservatory at the new hospital has “cement paths” (people often confuse cement and concrete, cement is a component of concrete). See Ewing, “An Invisible War” (Grantville Gazette, Volume 2). I can’t help but wonder whether the hospital itself, a three story building completed a year earlier, is of concrete construction.
The Higgins Hotel is at least partially built as of summer 1633, see Cooper, “Stretching Out Part 1” (Grantville Gazette, Volume 11) and “The Chase” (Ring of Fire II), but the stories don’t say whether it used concrete.
Somewhat inconsistently, in the Friends’ “Burgers, Fries, and Beer” (Grantville Gazette, Volume 7), set in January 1634, Julio Sanabria wonders where he would get the cement, fire clay, and lime he needs in order to put his masonry tools to use. You can’t make concrete without cement, and concrete was already being made.
The Grid lists a concrete company, started by William Roberts and his brother Ronald Chapman. Roberts is a managerial type and Chapman worked “for a company in Fairmont as a foreman of a team that built pre-fab metal buildings.” The relationship of this company to the high school concrete research lab is unclear. It is possible that the company existed only on paper.
Concrete and Cement Knowhow
Grantville is in do-it-yourselfer territory. There are also going to be a fair number of “how-to” manuals (some even read by their owners), as well as homeowners with hands-on experience making (and repairing, especially those who didn’t read the manuals) concrete flatwork (floors, driveways, patios, porches, walkways), foundations, walls and even outdoor furniture. But no pink concrete flamingos, I hope.
In addition, Grantville had at least two general contractors before the RoF, Happy Acres and Home Center (Grid). I doubt that either of them has built a skyscraper, and there isn’t much concrete construction in Mannington, but chances are reasonable that they have employees who have worked with concrete.
There were also construction technology courses offered at the Marion County Technical Center. I have no idea which courses were offered in 1999-2000, but the current catalog includes “Basic Masonry and Landscaping,” “Foundations and Framing,” “Fundamentals of Building Construction,” “Masonry and Plumbing,” and “Construction Systems.”
The up-timers with college engineering degrees are most likely to have attended either West Virginia University or Fairmont State. WVU offers a degree in civil engineering, with undergraduate elective courses in Civil Engineering Materials (CE310), Concrete and Aggregates (CE412), Construction Methods (CE413), Construction Engineering (CE414), Advanced Concrete Materials (CE416), and Reinforced Concrete Design (CE 462). While there is no guarantee that any particular civil engineer in Grantville has taken any of these courses, it is certainly possible.
In any event, there are going to be up-timers who know how to estimate how much concrete is needed for a job, prepare the forms to receive the concrete, put down the steel reinforcements, monitor the pour, and cure the concrete. A smaller number (those who didn’t just use ready-mix) will know how to proportion concrete, that is, decide the proper ratios of cement, aggregates and water.
What is less certain is that the up-timers will know, firsthand, how to make cement. Cement is usually bought ready-made and even a building contractor needn’t have firsthand knowledge of cement manufacture. If any up-timers do, it is probably because they worked in a cement plant outside Grantville. An EPA ranking of Portland cement plants, by size, listed Capitol Cement, in Martinsburg, in 29th place. That was the only West Virginia listing. (EPA) However, I have found a 1976 reference to the Marquette Cement Manufacturing Co. plant near Fairmont (PSC).
Grantville, according to canon, has pretty much every encyclopedia you can imagine. “Not just the great one, the 1911 Britannica, which they guarded so carefully, but all of them—the later Britannica editions, the World Book and Americana, Columbia, and Funk and Wagnalls, old and new, large and small.” (Flint and DeMarce, 1635: The Bavarian Crisis, Chap. 5). My understanding is that the public library has the Encyclopedia Americana, and both the modern and the 1911 editions of the Encyclopedia Britannica. The high school has the World Book, and the junior high, the Collier’s. There’s also a nearly complete ninth edition of the Britannica, and, I suspect, several CDROM-based encyclopedias, most likely Compton’s and Encarta. Besides having articles on concrete and cement, these encyclopedias have related tidbits scattered across their many volumes, which a sufficiently diligent researcher can uncover.
Grantville is modeled on Mannington, West Virginia, and I have checked the high school and public library holdings for more specific works. North Marion High School has the Time-Life Masonry (1977). It and the public library have Kicklighter’s Modern Masonry: Brick, Block, Stone (1977).
There may also be other relevant books. For example, the high school has Trachtenberg’s Brooklyn Bridge: Fact and Symbol (1979), and Stevens’ Dam: An American Adventure, and concrete was used in their construction.
There is no easy way of determining what books might be in private (home and work) libraries, or at the Voc-Ed Center. Bear in mind that any engineer almost certainly has kept all of his or her college engineering textbooks. Even a retired engineer would hesitate to part with them.
There is also a documented relationship between the North Marion High School of Mannington, WV and LaFarge, an Ohio cement company. LaFarge gave the high school a $300,000 atomic absorption spectrophotometer in 1997 (Zeller). Surely the student research projects developed using the AAS would have included ones dealing with cement. And perhaps the school got some cement technology texts along with the AAS.
Down-timers’ Cement and Concrete Technology
Previously, I said that Roman concrete was inferior to modern concrete, and I should explain why. First, it had a compressive strength of only 2800-3000 psi (RomanConcrete.com; Spratt), comparable to the “low end” of the strength range of modern concrete. Secondly, it was not reinforced.
Many sources state that concrete technology was “lost” in the Middle Ages. But I very deliberately used the term “languished” in the introduction. The Normans used concrete in the construction of parts of Reading Abbey (1130), the White Tower of London, and other structures. (Davidovits, May, Ferguson). But Ferguson comments that “concrete in the hands of the Normans was a total failure,” and lists a dozen Norman concrete towers which fell down.
Mukerji asserts that the Roman formula for “hydraulic cement” (that is, one which hardens in contact with water) wasn’t lost, but rather remained “tacit knowledge” among masons and military engineers, at least in limited areas, so that it was known in, e.g., seventeenth century France. Idorn (38) says that use of hydraulic cement was monopolized by the authorities; e.g., Christian IV of Denmark imported trass (from Dutch merchants) for making hydraulic mortars for his palaces and castles.
In 1568, the French architect Philibrt de L’Orme taught preparing a mortar from burnt quicklime, river sand, pebbles and water, with the pebbles being “of all sizes.” (Jackson 23).
A crude form of concrete was apparently used as ballast between the frames in the galleon Nuestra Senora de Atocha (1620)(Crisman).
The Advantages of Concrete
Concrete competes as a building material with steel, wood, and brick, and as a road pavement with asphalt. Concrete has numerous advantages as a structural material.
On-site Fabricability. The 1911 Encyclopedia Britannica (1911EB) says that concrete has “the immense advantage over natural stone that it can be easily molded while wet to any desired shape or size.” It has similar advantages over steel and wood. Steel can be cast only at a high temperature and wood not at all. Steel can be bent but only through persistent application of great force, and wood can be bent only gingerly and slowly, to avoid breakage.
Convenience. “Its constituents can be obtained in almost any part of the world, and its manufacture is extremely simple.” (1911EB).
Compressive strength. Like natural stone, it possesses great resistance to compression; its compressive strength is usually 4,000-15,000 pounds per square inch (psi), or higher, of cross-sectional area. (Levy/Down, 279). (In 2000, concrete with a strength of 8,000 psi or higher was considered “high strength”; Nilson 50.) The compressive strength of wood, parallel to the grain, is comparable, perhaps 6,000-7,500 psi, but perpendicular to the grain, wood is much weaker, perhaps 450-1050 (Green). As for natural stones, granites and marbles are stronger (up to 30,000 psi), and soft limestone weaker (700 psi)(Cowan 105).
Strength-to-Cost. The figure-of-merit (two-thirds power of strength, divided by cost per unit volume) is 80 for concrete, 60 reinforced concrete, 80 wood, 45 brick and stone and only 21 steel (Ashby 100).
Stiffness-to-Cost. The figure of merit (half power of Young’s modulus of elasticity divided by unit cost) is 40 for concrete, 20 reinforced concrete and brick, 15 wood and stone, and only 3 steel.
Stiffness-to-Weight. For columns which fail by buckling, the figure of merit is the half power of Young’s modulus, divided by the density. Steel is 59, but concrete is almost as good, 49 (Gordon 321). So by making the columns just a little thicker, you can use reinforced concrete instead of steel, saving perhaps 99% in steel consumption.
Fire resistance. Concrete itself is non-combustible, and has a thermal conductivity about 5% that of steel (PCA). However, it should be noted that with reinforced concrete, the reinforcing steel becomes ductile at high temperature, and since it presumably is there to provide tensile strength, the result may, ultimately, be structural failure.
Biological and chemical resistance. Wood rots, and is attacked by termites (or, at sea, teredo worms). Steel corrodes. Concrete isn’t vulnerable to these threats, but it can be attacked by acids, sulfates and chlorides (from deicing salts). Special concretes are used for construction in the vicinity of high-sulfate soil (or groundwater). And of course the reinforcement in reinforced concrete can corrode if corrosive agents can reach it.
Temperature stability. A concrete wall or floor will absorb heat during the day and re-radiate it at night. That’s true of any material, but concrete has a greater “thermal mass” than wood (HousingZone). (An interesting variation is a panel with a lightweight thermal insulating material sandwiched between layers of concrete.)
Soundproofing. The airborne sound insulation of a concrete first floor is 9-22 dB higher than that of a timber floor (“Going Up”).
Disadvantages of Concrete
Low Tensile strength. Unfortunately, concrete’s tensile strength (resistance to being pulled apart) is only about 10% of its compressive strength (Twelvetrees 41); Gordon (44) quotes a value of 600 psi (whereas commercial mild steel is 60,000 and high tensile engineering steel is 225,000). Because plain (unreinforced) concrete is strong in compression and weak in tension, it can be used in columns, arches and domes, but not in beams (horizontal structural members).
Low Compressive Strength-to-Weight Ratio. If we divide the compressive strength by the density (~2), we get values of about 2000-7500 psi/unit weight for unreinforced concrete. For steel, despite its greater density (~7.5), we get values of 4800-8000 (Twelvetrees 31).
Flexural strength. Concrete is not good at resisting failure from bending; its flexural strength is perhaps 12-20% compressive strength. (Cadman).
Brittleness. Concrete is also brittle, that is, once cracked it is easily fractured. The “fracture toughness” of concrete is 0.2-1.4, compared to 0.7-0.8 for soda lime glass and 50 for steel (Matt Gordon).
Shrinkage and Expansion. Freshly laid concrete shrinks as a result of the chemical reactions between its ingredients, the evaporation of water from the concrete, and the rising of air voids to its surface. Once hardened, concrete expands and shrinks in response to changes in temperature and moisture levels. Of course, all these dimensional changes stress the concrete, perhaps causing cracks.
Fortunately, if concrete is reinforced with a material, like steel, which is strong in tension, and ductile, it becomes an all-purpose structural material.
We start by reviewing the ingredients of concrete: cement, aggregate and water. Cement itself is a complex material. Once we know what goes into both concrete and cement, we can consider how concrete is mixed, laid, cured and tested.
Cement, in essence, is a binding agent. It can be used in mortar or in concrete. Cements are traditionally classified as being either hydraulic (those which, at least after setting, are resistant to water) or non-hydraulic (those which must be kept dry). The term “hydraulic” also has come to imply that when first mixed with water to make concrete, the cement reacts chemically with the water, forming hydrates which help bind the aggregate. (These hydrates are themselves insoluble in water, thus conferring the water resistance.)
The pozzolanic cements result from the mixture of a source of calcium (usually lime) and a source of silica (possibly also containing alumina). The lime is derived by heating chalk or limestone (calcium carbonate); the process is called calcination. The lime may be either the highly reactive quicklime (calcium oxide) or the somewhat less reactive slaked (hydrated) lime (calcium hydroxide), the latter being obtained by reacting quicklime with water.
Portland cements likewise are derived from a mixture of calcium- and silica-rich materials, but this mixture is subjected to a further calcination at a high temperature.
The natural cements are prepared from a source material which naturally contains both lime and silica. Hence, no mixing step is needed. Like Portland cements, it is calcinated, but at a lower temperature than that typical in Portland cement production. (Eckel, 151).
High-alumina cements are made from limestone and low-silica bauxite; they were invented in 1908.
In the twentieth century, the commercially dominant hydraulic cement was “Portland cement.” However, we will first discuss the older pozzolanic cements.
A pozzolan is a source of silica (silicon dioxide) which can react with lime to form a cement. A material can contain silica but not be useful as a pozzolan. For example, most sands contain silica, yet are unreactive. Moore says that this is because they have a tightly bound structure which frustrates the reaction. (Sands are used in concrete, but as aggregate.)
The ancient Roman cement, which is pozzolanic in nature, is described in Marcus Vitruvius Pollio’s De Architectura, Book II. This classic text became available to Europeans, in Latin, Italian and German printed translations, in the fifteenth and sixteenth centuries. There is also Sir Henry Wotton’s The Elements of Architecture (1624), which is derived from Vitruvius.
The first known post-Roman use of a pozzolanic cement was in Italy. The Venetians used the “black lime of Abetone” in the fifteenth century, and the Roman pozzolana was used by Fra Giocondo in the mortar of the pier of the Pont de Notre Dame in Paris (1499). (Giocondo published an edition of Vitruvius in 1511.)
Vitruvius refers to “rubble work,” with stones mortared together. In chapters 4-5, Vitruvius says that one may mix sand (a silica source) with lime to make mortar. He recommends use of three parts sand to one of lime when the sand is from a pit, and a two to one ratio if the sand is from sea or river. According to Moore, Vitruvius’ “pit sand” is actually volcanic ash, specifically, pozzolana.
Pozzolana. The eponymous pozzolana is a volcanic ash discovered at Pozzoli, near Vesuvius, but also found elsewhere in Italy (including near Rome). The 1911 Encyclopedia Britannica article on “cements” gives compositions for both Neapolitan (27.8% soluble silica, 5.68% lime) and Roman (32.64%; 4.06%) “Pozzuolana.”
Vitruvius, in chapter 6, says: “There is a species of sand which, naturally, possesses extraordinary qualities. It is found about Baiae and the territory in the neighborhood of Mount Vesuvius; if mixed with lime and rubble, it hardens as well under water as in ordinary buildings.” This “sand” is obviously a pozzolanic ash. Herring points out that this pozzolan could react with lime because it was already calcined by the volcano.
Santorin earth. This is really a volcanic tuff, which blankets the Greek island of Santorini (Thera). It is about 64% silica and 3.5% lime (USBM). It was used in ancient Greek mortar (Lea 3) and, millenia later, it was still exported for use in making pozzolanic cement (1911EB “Santorin”).
Pottery shards were ground up, in antiquity, to produce a pozzolan. Pottery is made by heating (calcinating) clay, and clay is rich in silicate minerals. Brick could be recycled in a similar way. Vitruvius says that if mortar is made from river or sea sand, it is improved by addition of one-third part of ground potsherds.
A modern clay-derived pozzolan is metakaolin. It is obtained by calcinating the clay mineral kaolinite, an aluminosilicate clay mineral. Metakaolin is one of the most reactive pozzolans.
Pumice is a very light, highly porous igneous rock, with a silica content of 60-75%. In 1911, pumice was chiefly obtained for commercial use from the Lipari Islands north of eastern Sicily, and especially from Monte Pelato and Monte Chirica. The Lipari Islands have exported pumice since antiquity, and Canneto is the center of the pumice trade.
Trass, a Germanic pozzolan, is a tuff (rock derived from volcanic ash) found in the Eifel, a volcanic region of Germany lying between the Rhine and Moselle rivers (1911EB). The article on “Trass” specifically mentions the Brohl and Nette valleys, and the town of Andernach. Eckel (635) says that trass occurs along the Rhine, from Koln to Coblenz, and that the towns of Brohl, Kruft, Plaidt and Andernach near Coblenz are significant players in the trass industry. 1911EB characterizes it as 19% soluble, 50% insoluble silica. It is lacking in lime so it is less reactive than pozzolana. Nonetheless, the Romans recognized its resemblance to the Vesuvian material.
(Johnson, 387). In 1837, trass sold for $5.225 per cubic meter, whereas common sand cost $0.85. (Treussart 90).
Extinct volcanoes can also be found in the Vogelsberg (west of Fulda), the Roehn (east of Fulda), the Lausitz (north of Dresden), and in the Eschwege at the Werra, east of Kassel. (MB).
Kieselguhr (diatomaceous earth, diatomite), which is derived from the silica skeletons of fossil diatoms, is over 80% silica. In 1911, it was not an economical material for cement making, because it was in demand as an absorbent for the nitroglycerin in dynamite. The 1911EB mentions deposits of diatomite in Richmond, Virginia, in Aberdeenshire (between Logie Coldstone and Dinnet), in Wales (Llyn Arenig Bach), and on Skye. It is in fact found in Germany (e.g., Obrehole), but I don’t know whether it was a known substance (e.g., for filtering beer) insofar as the down-timers are concerned.
Ground granulated blast furnace slag (GGBFS) is a product of steelmaking (1911EB). Slag cements were first used in 1774, in mortar (Prusinski).
While USE Steel in Grantville will no doubt be happy to sell its slag, King warns that the slag “requires a fair amount of processing to become a useful pozzolan.” GGBFS is produced by rapidly quenching (cooling) molten iron blast furnace slag by immersing it in water or blowing air over it, in a “granulator,” and then grinding it. The GGBFS is then combined with lime to make slag cement. (It should be noted that slag, processed differently, can be used to make an aggregate.)
Coal fly ash. When coal is burnt, it leaves behind both bottom ash and fly ash, the latter being the particles which are carried up into the smokestack. Fly ash was first used in a pozzolanic cement in the construction of the Hoover Dam (1929).
The silica content of the fly ash is dependent on the type of coal; 20-60% for bituminous, 40-60% for sub-bituminous, and 15-45% for lignite. The ash also contains lime; 1-12% for bituminous; 5-30% for sub-bituminous, and 15-40% for lignite. Fly ash is classified as being either Type C (calcium-rich) or Type F (calcium-poor). The type C ash is more reactive than the type F ash, and is even self-cementing. (“Fly Ash,” Wikipedia).
Fly ash particles have diameters of 1-100 microns. The particles with sizes under 10 microns are the most pozzolanically active, and ASTM limits the concentration of particles larger than 45 microns to 38%. The particle size distribution varies depending on the coal deposit and also on the plant design and operating parameters.
By way of a bonus, since fly ash particles are almost perfect spheres, they act like microscopic ball bearings, improving the workability and pumpability of the concrete in which they are used (Copeland).
Grantville has a coal-burning power plant which may already be equipped with devices for filtering out fly ash to minimize air pollution. By the beginning of 1634: The Baltic War, there is a coal gas plant in Magdeburg. They are connected by rail and water, and the fly ash can therefore be shipped to any point along the line which is convenient for cement and concrete manufacture.
How much the up-timers know about this utility of fly ash? It is not mentioned in 1911EB. The Encyclopedia Americana notes that it can be removed from the smokestack gas by electrostatic precipitators (“Power, Electric”) but doesn’t mention its significance for cement-making.
On the other hand, the Allegheny Power Company (which presumably owns the Grantville power plant) reported to the SEC that its subsidiaries sold 131,000 tons of fly ash (and 168,000 tons of bottom ash) in 1996, and that the uses of the ash included “cement replacement.”
So I am sure that at least the power plant manager, Bill Porter, knows about this possibility.
Talmy, USP 5521132 gives the composition of the fly ash from the Rivesville Power Plant, which was the model for the Grantville plant. It is 58.79% silica, 27.91% alumina, 8.41% iron oxide, and only 1.20% lime. Its LOI (loss on ignition), a measure of the unburnt carbon on the particles, is 28.3%. That’s high, so it will have to be burnt off. ASTM C618 requires that the LOI be no more than 6% (King 5).
Rice husk ash. Traditionally, rice was milled just to remove the chaff (outer husk), leaving brown rice. The brown rice may be further milled to remove the bran (inner husk), leaving white rice. If the husks are burnt, about 20% of the husk weight remains as ash, and this ash is about 95% silica, and constitutes a highly reactive pozzolan (Allen, King). The difficulty in preparing the ash is burning the husk at a temperature low enough so that the silica doesn’t form inactive crystals while burning it long enough to ensure that all the cellulose is consumed.
Americans don’t think of rice as a European crop, but it was brought to Spain and Portugal by the Moors (“Rice,” Wikipedia), and has been grown in Italy at least since the fifteenth century. Lombardy was the first major Italian producing region. By 1644, there was rice production in the Veneto, and in the nineteenth century canal construction made it possible to grow rice in the Piedmont. (Seed). Rice can also be grown in France and Greece.
Silica fume. Once the semiconductor industry is reestablished, there will be the possibility of using silica fume (0.1 micron silica particles, a byproduct of silicon production) as a high-activity pozzolan. Silica fume is expensive and difficult to work with, so it will probably be relegated to the same niche market it enjoys now (concrete with compressive strength exceeding 15,000 psi and with high chloride resistance). (King 7; SFA).
“Portland cement” was patented by Joseph Aspdin in 1824, and improved by his son William in 1843. Aspdin’s cement was made by heating together finely ground limestone and clay. He cooled the resulting “clinker” and pulverized it. This powder could be stored until it was ready to be activated by addition of water.
The Aspdins used too low a temperature (probably lower than 1400 deg. C) to achieve a true Portland cement. (Blezard 8). (EB11 specifies a “clinkering” temperature of 1500 deg. C (2732 deg. F) which is in accord with modern practice.)
There is a good description of the modern American cement-making process in “Cement,” Encyclopedia Americana. Sources of lime (e.g., limestone, chalk, marl, marble, shells), and of silica (sand, sandstone, clay, slag, ash) and alumina (clay, shale, bauxite) are quarried and crushed, then mixed together and ground up some more. The grinding can be done wet (that is, in a water slurry) or dry. Wet grinding yields a more homogeneous blend, but the powder has to stay in the kiln longer. (Camp)
This “rawmix” flows into a continuously operated, inclined, rotating kiln. EA says that this is typically 300-400 feet long, inclined at one half inch to the foot, and rotated at 30-90 revolutions per hour. The kiln is hottest at the discharge end.
The material takes 2-4 hours to pass through the kiln, and reaches a temperature of 2600-2800 deg. F. First water is driven out, and then the carbonates decompose into oxides. Ultimately, some of the material liquefies, and the lime (calcium oxide) reacts with the silica to form calcium silicates, notably dicalcium silicate (belite) and tricalcium silicate (alite).
Shale and clay often have a high aluminum and iron content. Alumina (aluminum oxide) serves as a flux, that is, it reduces the melting point so that more of the charge is liquefied at the peak kiln processing temperature. Thanks to the flux, liquid appears at about 2400 deg. F, but even at the peak temperature, only 20-30% of the charge is in the liquid phase. When the charge is cooled, the alumina is converted into tricalcium aluminate.
The EA “Concrete” article explains that tricalcium aluminate “produces a very high heat of hydration” and “has poor durability because it reacts with sulfate alkalis found in soil and water.” Overly high aluminate levels may be reduced by adding iron ore to the kiln.
Iron oxides also act as fluxes, and they react with aluminate to form tetracalcium aluminoferrate. This iron compound is responsible for the grey color of the cement; if cement is made from low-iron materials, it will be white in color. On the other hand, the iron improves the resistance of the concrete to sulfate water.
Tricalcium aluminate forms because a source (e.g., bauxite) of aluminum oxide (alumina) is added to the kiln when making Portland cement. It is provided to reduce the melting point of the composition so it is liquid at the peak kiln processing temperature, thereby favoring formation of alite and belite. Unfortunately, it has undesirable properties, Like tricalcium aluminate, the tetracalcium aluminoferrate acts as a flux.
The products of the calcination reaction in the kiln are black hard nodules with diameters of one-quarter to one inch diameter, called “clinker” because they make a clinking noise in the kiln. These are mixed with 4-5% gypsum (hydrous calcium sulfate). EA states that the purpose of the gypsum is to slow down the “setting” of cement, since otherwise a Portland cement concrete mix might set, and become unworkable, before pouring was complete.
On average, every thousand tons of cement requires roughly 1511 tons of various oxides (1315 tons calcium oxide, 71 tons silica, 108 tons alumina, 17 tons ferric oxide), and 53 tons gypsum. (Van Oss 22). To get those oxides probably means processing up to twice the weight in raw rock.
Vertical Kiln Development
Of course, the post-RoF cement industry is going to begin more humbly than with the monster rotary kilns described in EA. The first cement kilns were intermittent, vertical kilns. Such kilns are “old” technology, already used in pottery, lime and brick making, and so there will be a rapid adaptation.
The simplest kiln design is a pit kiln, in which the fire is allowed to burn downward. Unfortunately, most of the heat is wasted, because it escapes upward.
An improved design is a simple shaft kiln; this involved digging a horizontal tunnel into the side of a hill, and a vertical shaft down to meet it. An arch of limestone is built at the junction. The rawmix is piled on top of the arch, and the fuel goes below it (“separated feed”). The fuel is lit and the fire burns upward. (Lazell 24-30; Eckel 409-19).
In both pit and shaft kilns, earth acts as the insulator. The dome kiln is the free-standing equivalent, made of brick or perhaps brick-lined metal. The interior was egg- or bottle-shaped, with the top portion serving as a chimney. The arch was replaced with a grating, and the fuel (preferably coke, but sometimes firewood) and rawmix was piled above the grating in many alternating layers (“mixed feed”). A typical dome kiln was 15-20 feet high and perhaps six feet in diameter.
Normal operation was discontinuous. The kiln would be loaded with perhaps 50 tons slurry and 12 tons coke. It will take two days to fire up, two or three days to burn through, and additional time for cooling down, drawing out the clinker, and reloading the kiln. Dome kilns produced perhaps thirty tons clinker per batch, and one batch per week (EB11). According to Eckel, production is 0.5-1 ton clinker per cubic meter of burning space, and 23-30 pounds of fuel are needed per 100 pounds clinker.
Intermittent operation is wasteful of energy, since the kiln must be cooled down and then reheated for the next batch. But the new arrangement made it theoretically possible to operate the shaft or dome kiln continuously. One worker could (cautiously) collect clinker which has fallen through the grating, while another added new layers at the top. We then have a “running kiln.”
In practice, the clinker tended to hang up, forcing a cool-down (Redgrave 158). Also, it was difficult to maintain a consistent burn in running lime kilns (Johnson) and I suspect that the same problem would have carried over to cement kilns. Lipowitz (32) said in 1868, “many attempts to establish a kiln on the perpetual system have been devised, but hitherto the desideratum of a perfectly unexceptionable running kiln is still unattained.”
Chamber kilns were adapted from brickworks, and the basic concept was that excess heat from one chamber was transferred to another. They thus achieved a substantial fuel savings. Chamber kilns are an old technology, but they reached their pinnacle in 1858, when Hoffman invented the “continuous” chamber (ring) kiln, briefly described by EB11.
The first vertical kilns capable of sustained operation appeared in the 1880s. These were larger than dome kilns (the Coplay Cement Company’s nine Schofer kilns, operated in Delaware 1893-1904, were ninety feet tall), with separate drawing, burning and loading floors for the workers, and multiple ports and chutes through which to regulate the supply of rawmix and fuel. They probably had better linings, too. However, my sources are maddeningly vague about just how they avoided the problems of the old “running” dome kilns.
The new kilns differed in terms of where exactly the fuel and rawmix were added, the fuel used (coke or small coal), and where the interior narrowed and widened. EB11 diagrams the Dietzsch type, in which the shaft is staggered to create a horizontal ledge to which the fuel was added. A pair was usually built back-to-back. The upper vertical shaft contains the unburnt rawmix and the lower shaft is the burning zone.
EB11 also mentions the Schneider kiln, which had a single vertical path. The Schofer (Aalborg) kiln was similar. Eckel says it produced 10-15 tons clinker daily, consuming 280 pounds coal per ton product.
Rotary Kiln Development
It took roughly ten years (1885-1895) to achieve a truly practical rotary kiln. There were “many practical difficulties” and “an immense amount of expensive experimenting” (Sabin 23; ER Chap. 20; Brown 39-42; Redgrave 167-176).
One problem with the early rotary kilns is that they were simply too short, Ransome’s being twenty-six feet long and Navarro’s, forty feet. Consequently, there was a lot of underburnt clinker, and also much heat was wasted. While our heroes will know that the modern rotary kilns are hundreds of feet long, without foreknowledge of the problems of the pioneers, the first post-RoF rotary kilns are likely to be short prototypes, of underwhelming performance.
Secondly, there were various problems with the kiln lining, both spalling of the lining and balling of clinker upon it.
And finally there were the issues of finding the right fuel and minimizing fuel consumption. The first fuel experimented with was gas. Next came a “jet of burning petroleum,” because it allowed precise control of the temperature of the kiln. Indeed, a chemist asserted at the time that the “rotary kiln can be successfully operated only in localities where crude oil is abundant and cheap.” (Prentice) However, in most places oil was expensive, and the rotary kiln didn’t really catch on until it was adapted (1895) to use blown pulverized coal.
The great advantages of the rotary kiln, once perfected, were its low labor cost (20-30% that of continuous shaft kilns) and high production rate (over double) (ER 188). Its bugbear was fuel consumption.
For several decades, the standard dry-process kiln was sixty feet long and six feet diameter, and the wet-process cousin could be up to eighty feet. The dry-process kiln produced 160-180 barrels (each 376 pounds) of clinker daily, consuming 110-150 pounds coal per barrel. (Eckel 424; Sabin says 175-250 barrels for 95-120 pounds coal/barrel.) The wet process was even more wasteful of fuel (ER 17).
It took Edison from 1899 to 1902 (Vanderbilt) to build the first “long” (150 foot) kiln, despite his study of the “short” kiln technology. The Edison kiln tube, nine feet in diameter, was made from ten-foot sections of cast iron, bolted together. It was suspended on fifteen rollers, rotated by an electric motor, and there were ten thousand bearings, lubricated with an automatic oiling system. The kiln had a pitch of eighteen inches and powdered coal was forced in by pressurized air to create a forty foot combustion zone at the lower end. Only two men were needed per shift.
Edison shocked the industry by producing 350-375 barrels daily, while consuming only 65 pounds coal per barrel (Eckel 424). Edison ultimately increased production to 1100 barrels/day (Vanderbilt 185). The “light bulbs came on,” and by 1918 there were kilns over 200 feet long.
In the standard sixty footer, the combustion zone, in which clinker was formed, was near the lower end, and about ten feet long, and the heated gases which rose from it had only the upper forty feet in which to decompose the rawmix. Much of the heat of the gases was wasted, and the processing path was so short that the rock still retained much of its carbon dioxide, releasing it when it reached the combustion zone. (Dyer)
By increasing the length, the cylinder could be fed faster, tilted more steeply and rotated more quickly, and still burn the stone properly, without the produced carbon dioxide interfering with the combustion. Eckel (429) estimates that the output (barrels/day) will be between one-eight and one-twelfth of the product of the length (feet) and the square of the internal diameter (feet) at the discharge end.
In re-inventing the rotary kiln, our heroes will need to solve problems relating to forming the giant metal cylinder, developing a proper lining (early 20c usage was alumina brick), providing the mechanisms for turning the cylinder, and assuring proper heating.
Daily Production (barrels)
US Labor Cost
Hoffman (per chamber)
Rotary, wet 60′
(Sabin 25; E:Eckel 411, 424)
In general, the fuel-efficient vertical kilns long remained popular in Europe, where labor was cheap and fuel was expensive. In America, where the reverse was true, by 1900, the rotary kiln accounted for 90% of production.
The clinker is ground down, to a very fine powder. How fine? EB11 says, enough so most passes through 0.005 inch sieve holes. In modern practice, to an average of ten microns, which makes baby’s talcum powder seem coarse in comparison. Cement plants will need to be able to measure particle size in order to ensure a consistently high quality product.
Grinding of clinker was done originally with millstones (EB11). However, by the late nineteenth century, ball mills were available. In a ball mill, the material to be ground enters one end of a rotating cylinder, and leaves at the other. The cylinder contains balls made of a hard material, such as steel, rock or ceramic, and the clinker is ground down by friction and impact. The mill can be powered by animals, wind, water or electricity.
Cement powder must be kept dry until use since cement reacts with water (“sets”). Cement also gradually loses strength when stored.
There were five basic types of Portland cement in use in 2000:
Type I: General purpose.
Type II: Moderate sulfate resistance.
Type III: High early strength (gains strength faster than type I, which allows earlier removal of the forms which the concrete is poured into. According to Arnold 32, type III cures about twice as fast as Type I)
Type IV: Low heat of hydration (for use in massive structures, like dams, where it is difficult for the heat to escape because of a low surface-to-volume ratio)
Type V: High sulfate resistance.
There were also “air-entrained” variants of types I-III (IA-IIIA) so you could say that there are actually eight basic types.
Domestic production in 2000 was over 90% types I/II, with the balance split primarily between types III and V. Type IV was less than 1% of production. (USGS).
Type III is used mostly for precast concrete manufacture (so the molds can be reused more quickly) and for emergency repairs. Commercially, types IV and V have been largely superseded by Portland-pozzolan blended cements (see below). What that all means is that, as a practical matter, the up-timers probably have experience only with types I and II.
The typical chemical composition of the five types of Portland cement is given in the Encyclopedia Americana “Cement” article. The chart lists the proportions of tricalcium silicate, dicalcium silicate, tricalcium aluminate, tetracalcium aluminoferrate, calcium sulfate, magnesium oxide, and free calcium oxide for each. EA notes that it isn’t desirable for free calcium oxide to exceed 2-3%.
While this may make it seem that the chemical composition is critical, the ASTM specifications are “not very strict since cements with different chemical compounds can have similar physical behavior” (Camp). But Camp also says that “high quality cements require adequate and uniform raw materials.” How do we reconcile these two statements?
What it comes down to is that the physical properties of the cement (and corresponding concrete) have to be predictable (so that, say, the concrete in the third floor is just as strong as the concrete in the lower floors). So once you determine that a particular combination of raw cement-making materials makes a physically desirable cement, you want to make sure that all of the cement produced will continue to manifest those desirable properties.
It is also worth noting that both chemical and physical properties of Portland cement have changed over the years, with post-1930 cements being more finely ground and containing more belite than older cements. The concrete made with post-1930 cements strengthens more quickly (so construction is faster) but is less durable (Mehta).
The quarried stone can vary from day to day in its content of the various oxides. Quality control—measuring the chemical composition of the stone and the clinker, and adjusting the proportions of limestone, clay etc. appropriately—began in the 1870s (Blezard 17).
In modern practice, the chemical composition of the rawmix is tightly controlled—within 0.1% or better! This accuracy is achieved by hourly X-ray fluorescence or, every three minutes, gamma neutron activation analysis. (“Portland Cement,” Wikipedia).
Obviously, we are not going to achieve that kind of control in the early post-RoF period. Nor is such control critical, as long as one periodically tests the efficacy of the cement. After all, nineteenth century rawmix certainly wasn’t monitored by X-ray fluorescence. EB11 says that “the silica may range from 19 to 27%, the alumina and ferric oxide jointly from 7 to 14%, the lime from 60 to 67%.”
I would imagine that there would be an attempt to perform daily chemical analyses on the rawmix, and that these analyses would be correlated with tests of the cement emerging at the other end of the kiln. The latter will include both chemical tests (see below for the desired constituents of different types of cement) and physical ones, i.e., using the cement to make a concrete and then testing the concrete for strength. In addition, there would probably be at least daily chemical testing on the raw materials (e.g., limestone). Even this crude quality control is going to be new to the seventeenth century.
Grantville chemists will have to dredge up their old quantitative chemical analysis course textbooks and figure out how to assay for silica, alumina, iron oxide, lime, magnesia, etc. (Waterbury, Appendix III). Once they have an oxide analysis, the proportions of the four main reaction products can be predicted using what is called the Bogue calculation. Classical chemical analysis can require hours or even a few days to complete (Blezard 22), so adjustments will be sluggish by modern standards.
So-called “natural cement” is a cement, similar to Portland cement, which was produced by calcination of a naturally occurring mixture of lime and clay, such as the “dolostone” (magnesium-rich limestone) of Rosendale, New York. It was used to make concrete for the Brooklyn Bridge and Grand Central station in New York City. EA “Cement” distinguishes between “hydraulic limes” (with a silica-alumina content of 10-20%) and “natural cements” (20-35%).
1911EB acknowledges the existence of natural-cement deposits near Chittenango, in Madison County, New York, as well as in Kentucky (Louisville), Indiana (Clark county), Illinois, Oregon (Rogue river), Pennsylvania (Williamsport, Lycoming County). EA mentions Rosendale and Fayettesville, NY.
Natural cements do occur in Europe, notably in Belgium (Tournai district) and England (septarian nodules found in southern England) (Eckel1905, 214-8). However, this will have to be discovered the hard way.
According to EA, natural cements can be made by processing the rock in “small, upright, wood-burning kilns . . . fired for about a week”, and then grinding the resulting “clinker” between millstones, using waterpower. Firing temperatures are similar to those of lime kilns (1000-1200°F) (ER 21).
In any event, users of “natural cement” were at the mercy of nature; they had to be content with the particular mixture of lime and clay which the formation produced, whether it gave the cement and concrete the correct characteristics or not. The clinker must be sorted, and the under and over burnt material tossed away. Losses will probably be on the order of 25%. (Reid 9). The strength of natural cement is perhaps half that of Portland cement (Mills 41).
High Alumina Cement
This is made by heating a mixture of limestone and aluminum ore (bauxite). According to EA “Cement,” it’s resistant to sulfates and chlorides, and it hardens faster than Portland cement (and hence is useful for emergency road repair). Hot, moist conditions can cause it to suffer a catastrophic change in its microstructure—which is why the Europeans ban its use in structural concrete (Camp 6).
For sources of bauxite, see Cooper, “Aluminum: Will O’ The Wisp?” (Grantville Gazette, Volume 8).
In the twentieth century, this referred to a combination of Portland cement with a pozzolanic cement. EB11 says that adding trass can increase strength.
Blended cements were first used in underground and underwater structures because they imparted increased durability. Later, they were used in massive concrete works because of their reduced heat evolution (compare type IV Portland cement) and in weather-exposed concrete to reduce expansion (and cracking) as a result of the alkali-aggregate reaction. Ultimately, they came into general use. (Lea 424-5).
In general, the blended cements have a lower short-term (28 day) strength and a higher long-term (1-5 years) strength. (Lea 436, 473).
The aggregates form about 60-80% of the concrete (Ahrens 17). You may think of the aggregates are the “bones” of concrete, and the cement as the “sinews.” Most of the weight of concrete, and also most of the structural strength, is attributable to the aggregate.
A concrete mix will usually include both fine (under one-quarter inch) and coarse aggregates. The maximum size for the aggregate is usually an inch or two. (Roman concrete used much larger pieces, see Cowan 120.) The purpose of the fine aggregate is to fill in the spaces between the larger chunks, and it’s usually sand. (Arnold 27)
The proper shape of the aggregate pieces is going to be a matter of study and debate. EB11 “Concrete” teaches that “spherical pebbles are to be avoided,” and that the grains of sand should be of “an angular shape.” However, modern concrete technologists are of the opinion that “the ideal aggregate would be spherical and smooth.” (Camp, Chap. 6).
In the nineteenth century, aggregates were haphazardly collected from the pit or crusher. (Bauer 54). But ideally, the coarse aggregate is not a single size but rather exhibits a spectrum of sizes. An example of a desired grading curve, Fuller’s “ideal curve” (1907) specifies that the fraction of aggregate smaller than a particular size is the square root of the ratio of that size to the maximum size of the aggregate.
Use of a well-graded aggregate permits more economical use of cement, but this is balanced somewhat by the greater amount of work which may be involved in achieving that degree of grading for the available aggregate. A sieve analysis is used to determine whether the aggregate approximates the desired grading curve.
Normal density aggregates include crushed limestone, sand, river gravel, and crushed recycled concrete. The density of concrete made from such aggregates would be perhaps 150 pounds/cubic foot.
Low density aggregates might be expanded clay, shale, slate, slag, or crushed recycled brick. They bring concrete density down to 100-130 pounds/cubic foot (Ahrens 23).
Very low density aggregates include expanded mica, vermiculite, perlite, pumice, and glass or ceramic spheres. Concrete density can be as low as 20 pounds/cubic foot, which is lighter than water. The concretes used in the annual “concrete boat” races make considerable use of these aggregates.
Sand and gravel. Sand and gravel may be deposited by glaciers, water, or wind. In general, wind deposits (dunes and loess) are not useful in concrete-making.
If you look at a map of Europe during the last ice age, you can see that the terminus of the continental glacier was along a gentle NE to SW curve, falling a little north of Hamburg and Dresden. In southern Germany, there are likely to be deposits left by alpine glaciers. However, since these glaciers were advancing during the seventeenth century (remember the Little Ice Age), I am not sure that they will be accessible.
Water deposits can be left by rivers or the sea, and in general river sand is more useful. The problem with sea sand is that it encrusted with sea salts. Twentieth century facilities can desalinate sea sand but this isn’t likely to be a practical option for our protagonists.
You can find river sand (or gravel), not only in the beds and banks of existing waterways, but also in ancient channels. In the mountains, gravel may be found in alluvial fans.
Julio Sanabria may have exaggerated a bit when he assumed that “sand is everywhere,” but there shouldn’t be a problem finding some sand in the vicinity of Grantville and Magdeburg.
Crushed stone. In the United States (1989), the preferred rocks for crushing are, in descending order, limestone, then granite, traprock, dolomite, sandstone/quartz/quartzite, and marble, and finally volcanic cinder/scoria, slate, and marl (USGS). Quarrying is likely to occur where the rock outcrops or at least the soil overburden is thin. It might also be done as a byproduct of roadbuilding.
The choice of stone is dictated by a combination of its engineering properties, availability and cost. The rocks vary in mechanical strength, density, durability, chemical stability, surface characteristics, content of undesirable impurities, and the suitability of the shape of the crushed fragments (Waddell 2.6-.19).
Air-cooled blast furnace slag. This can be obtained wherever pig iron is produced. Slag concrete is actually stronger than that made from gravel.(Ramachandran).
Bottom Ash. This is similar to fly ash, but unfortunately higher in alkalis and sulfates. Ramachandran suggests that it be used as a lightweight aggregate in concrete block production.
Red Mud. This is a waste product of the production of alumina from bauxite. It can be formed into balls and then fired, like clay pottery, to produce an aggregate. A fairly high temperature (1260-1310 deg. C) is required to melt it.
Sawdust. Sawdust can be used as an aggregate (as in “woodcrete”), but sawdust concrete has reduced strength and, if the sawdust content is high, is flammable. The best wood sources are spruce and Norway pine. (Ramachandran).
Waste glass. Glass refuse can be used as a lightweight aggregate, but it reduces strength, and also renders the concrete susceptible to certain chemical attacks.
The key is to use water which doesn’t contain significant levels of impurities that adversely alter the properties of the concrete. For example, seawater has a high content of dissolved salts, including sulfates and chlorides, which can cause a variety of problems.
According to EA “Concrete,” if the water is good enough to drink, it is good enough to make concrete. However, drinkable water isn’t strictly necessary.
In the seventeenth century, water quality cannot be taken for granted. The ASTM standard method of deciding whether water is acceptable for concrete mixing is “if the setting time does not differ by more than 30 minutes and the strength is not reduced by more than 20% when compared with a sample [made] using distilled water.” (Camp).
In cold weather, it can be advantageous to use hot water, thereby speeding up the setting of the concrete (Arnold 32).
The Romans experimented with animal fat, milk and blood as additives (UIUC).
Air. It is possible to entrain air into concrete, either when mixing the concrete, or by use of an air-entrained cement. Why is this good, when concrete layers are taught a variety of techniques to prevent voids? The difference is that these “good” air bubbles are small (one to three thousandth of an inch), numerous (400-600 billion per cubic yard), and well-distributed. The air content may be 3-7%, rather than the usual 1-2%.
Air-entrained concrete is more resistant to freeze-thaw cycles. In ordinary concrete, when temperatures drop below freezing, the residual water expands, and stresses the concrete, possibly cracking it. But in air-entrained concrete, all those air spaces can compress if the concrete nearby is stressed. Air-entrained concrete is also more workable. The mix must be adjusted to avoid loss of compressive strength.
Air-entrainment was developed in the 1930s. It is achieved by adding a surfactant to the concrete. The conventional air-entrainment agents include “Vinsol resin” (the “petroleum-hydrocarbon, insoluble fraction of a coal-tar, hydrocarbon extract of pine wood”) and “Darex-AEA (“a triethanolamine salt of a sulfonated hydrocarbon”)(Bauer 47). Vinsol resin is mentioned in the CRC Handbook of Mechanical Engineering, a reference book reasonably likely to have been in Grantville.
No doubt people will think about using ordinary household detergents, but use of the wrong one could reduce strength and not achieve the desired effect. So any candidate surfactant has to be tested for its effect on the concrete (CCN).
Accelerants. Agents, e.g., calcium chloride (CRC), can be added to reduce the setting time.
Retardants. Other agents, e.g. sugar ( CRC), can be used to cause the concrete to set more slowly. They are typically used during hot weather.
Water-reducers. These reduce the amount of water needed for the mix to have the desired level of workability. Decreasing the water:cement ratio increases strength.
Superplasticizers are “second-generation” water-reducers; they were introduced in the 1980s. They include sulfonated melamine (or napthalene) formaldehyde condensates and ligonsulfonates (CRC).
Pigments. You can imagine an early architect telling a client, “you can have any color concrete wall you like, as long as it’s grey.” In the late twentieth century we had the choices of blue (cobalt oxide), brown (iron oxide), buff (another iron oxide), green (chromium oxide), and red (yet other iron oxide). Pigments can be expensive, and tend to weaken concrete, so they are used sparingly. (Chen 30).
Plain concrete weights 150-160 pounds/cubic foot. Lightweight concrete weighs 35-120 pounds/cubic foot. There are two basic methods of lightening concrete. Either you use a lightweight aggregate (see above), or you add a foaming agent (e.g. aluminum powder) to put gas bubbles into the concrete. In our time, concrete using a lightweight aggregate costs 30-50% more than ordinary concrete. (Merrill).
Structural lightweight concrete, made with expanded shale or clay aggregate, has a strength of 2,500-6,000 psi. It was used in constructing the 52 story One Shell Plaza in Houston. Intermediate lightweight concrete, made with pumice, scoria or herculite aggregate, has a strength of 1000-2500 psi. Extra-lightweight concrete, made with perlite, vermiculite and polystyrene bead aggregate, has a strength of only 100-1000 psi. (Ali).
Pumice and scoria are considered volcanic glass. Concrete made with them weighs 90-100 pounds/cubic foot. (Lewel).
Perlite is a volcanic rock which expands dramatically when heated, somewhat like popped popcorn. Encyclopedia Americana says that it is found in New Mexico, Greece, Hungary and Italy. Concrete made with expanded perlite weights 35-75 pounds/cubic foot.
Vermiculite is a clayey mineral (think “kitty litter”) which, in the crude state, has a density about twice that of water. When heated, it “expands explosively,” perhaps 20-30 fold. Encyclopedia Americana notes that it is mined in Montana (near Libby), North Carolina, South Carolina and Wyoming. It is doubtful that any up-time information is available on where it can be found in Europe. Concrete made with expanded vermiculite weights 35-75 pounds/cubic foot.
High density aggregates, such as barite, limonite, magnetite and steel balls, have been used to increase the strength of a concrete structure, especially fortifications.
Steel. Reinforcement of concrete was proposed by Joseph-Louis Lambot in 1848. Reinforcing concrete with steel means that you take advantages of the strengths of both materials. The steel provides tensile strength, while the concrete provides compressive strength and also protects the steel from the environment. It can be provided in the form of individual bars, welded wire fabric, and cable (strands twisted together).
The steel is placed where it will carry the tensile loads of the structure. This is possible only because the concrete adheres to the steel, so tensions are transferred from one material to the other. The force of adhesion is dependent on the surface area of the reinforcing bar (“rebar”), and the bar must be long enough so that the strength of the bond is greater than the strength of the bar itself.
Most materials expand when it is hot and contract when it is cold. The dimensional change with temperature is measured by the “coefficient of expansion.” If the coefficients for steel and concrete were dissimilar, then changes in temperature would disrupt the adhesion needed for the composite material to behave as a single unit. However, they are in fact quite similar, Twelvetrees (47) giving values of 0.0000066 and 0.0000055, per degree Fahrenheit, respectively. Still, a change of temperature will cause some stress; the steel bears it better because it is about 100 times as elastic as concrete.
Besides resisting tensile stress, the rebar also helps distribute strain throughout the concrete, and hence reduces the chance of rupture at a point of concentration.
Obviously, our ability to use steel-reinforced concrete is dependent to some degree on steel production. However, it is equally clear that it is more economical of our precious steel production capacity to build with steel-reinforced concrete than with steel alone. For a reinforced concrete beam, if the relative cross-sectional area of the rebar was 1% (“Reinforced Concrete,” Wikipedia), the rebar might add just 3.6% to the weight of the beam. That 1% corresponds to 132 pounds of steel of per cubic yard concrete. (Taylor 536).
During the first decade post-RoF, I expect that reinforcement practice will be similar to that of the early twentieth century. Figure 50-200 pounds of steel reinforcement per cubic yard of concrete, depending on the use. (Taylor, 14-32).
The advantages of reinforcement aren’t limited to poured concrete. Precast blocks usually have two or three vertical holes; steel bars may be placed through the holes for increased strength.
Prestressed concrete. This is a variation on ordinary reinforced concrete. The idea is that a tensile stress applied to steel tendons generates a compressive stress in the surrounding concrete. The tendons are placed so that the compressive force offsets tensile forces which are imposed on the concrete by the overall structure in service. In other words, prestressing permits the concrete to match its strength (resistance to compression) against what would otherwise be a foe to which it is especially vulnerable (a tensile structural load). (Waddell, Chap. 41). Prestressed concrete is used to make the floors of many high-rise buildings.
In pre-tensioning, the “tendons” are tensed in a “stressing bed” by hydraulic jacks. The ends are anchored by reinforced concrete or structural steel abutments which extend deep into the ground. The concrete is poured into the bed and, once it cures to a desired compressive strength, the tendons are cut loose at the ends, which causes the transfer of the stress from the steel to the concrete (“detensioning”). Typically, pre-tensioning is done when the concrete members are manufactured in a central casting yard for transport to building sites.
In post-tensioning, the concrete is cast so that it contains ducts (by using thin-walled steel forms), and, once the concrete has gained sufficient strength, the “tendons” are run through the ducts and tensed and grouted. Post-tensioning is more likely to be carried out on-site.
It is fairly common for prestressed concrete to also be precast.
Cast or Wrought Iron. Traditional cast iron has a tensile strength of 10,000-20,000 psi, and wrought iron, 20,000-40,000 psi (Gordon 44). While inferior to steel in strength, and quicker to rust, they can be used in reinforcements if steel supplies are inadequate.
Non-Ferrous Metals. The common ones (copper, zinc, brass, bronze, aluminum) won’t work. While their strength is at least equal to that of cast iron, their coefficient of thermal expansion is significantly higher than that of concrete. There’s also the problem of corrosion by caustic alkalis in the concrete. Tin has a good coefficient, but is weaker than cast iron.
Wood Reinforced Concrete. Mass production of steel was unknown in the seventeenth century prior to the Ring of Fire. Hence, the steel industry in the early post-RoF years has a lot of catching up to do in order to be on par with what it was in the mid-nineteenth century, when steel-reinforced concrete was invented. This led me to wonder whether one might, as an expedient, use large wooden beams in place of steel bars as reinforcements for the less demanding concrete structures. While the tensile strength of wood is inferior to that of iron, it is still far superior to that of concrete—at least if the tensile forces act along the grain of the wood.
That said, it doesn’t seem likely that wooden rebar (as opposed to wood fibers, see below) is practical. With regard to posts, Radford (154) said “no form of wooden reinforcement, either on the surface or within the post, can be recommended. If on the surface, the wood will decay; and if a wooden core is used, it will in all probability swell by the absorption of moisture, and crack the post.” So the problem of moisture must be addressed by applying some kind of waterproof coating to the wood.
Another issue is whether the concrete will adhere to the wood, which is critical for the transfer of tensile stress from the concrete to the reinforcement. Cobleigh says, ” A wooden reinforcement in the center of a concrete fence post is worse than useless. It does not make a bond with the concrete, and thus weakens, instead of strengthens, the post. Of course, the same is true of wooden reinforcement of any concrete work.”
Finally, there is the issue of whether the wood and concrete would expand or contract the same amount if the temperature changed. From what I can tell, the coefficient of thermal expansion of wood is about half that of concrete (Luebkeman).
Bamboo can be used as rebar, but it isn’t bound well by cement, and the bamboo must be treated so that water absorption doesn’t become a problem. (Swamy, 141, 157).
Fiber Reinforced Concrete. “And Pharaoh commanded the same day the taskmasters of the people, and their officers, saying, ‘Ye shall no more give the people straw to make brick, as heretofore: let them go and gather straw for themselves.” (Exodus 5:6). Clearly, fibers have been used to reinforce brittle materials since antiquity. For example, straw and horsehair were baked into mud bricks (Mohr 113). Cement and concrete have been reinforced with a variety of fibers, including steel, wood, asbestos, glass, textile, plastic, and carbon.
Fibers tend to be used in two different ways. First, they can be directly incorporated into the cement of conventionally reinforced concrete to reduce local cracking. Secondly, they can be incorporated into a matrix of some kind, and the resulting composite fabricated into rebar used in lieu of steel reinforcement, to increase tensile strength (while adding less weight).
Direct incorporation of glass or hemp fibers can actually reduce tensile strength (Materschlager). Steel fibers do increase tensile strength by 30-40%, and flexural strength (resistance to first cracking) by 50-150% (Frank). Of course, the tensile strength of unreinforced concrete is abyssmal, so 30-40% isn’t much of an improvement.
In the case of direct incorporation, it is critical that the cement adhere to the fibers, or the fibers create weak spots. It may be possible to overcome adhesion problems with suitable coatings.
Straw was used to reinforce bricks in ancient Mesopotamia. The Finns added asbestos fibers to clay pots as early as 2500 BC. They were added to cement in 1898, but given the health concerns with asbestos, this history isn’t likely to be repeated in the new time line.
Steel fiber reinforcement has been studied since the 1950s. Steel fibers are usually 0.5-2.5 inches long, and are added at a concentration of 0.25-2% by volume. (R&T).
Plastic (acrylic, nylon, polyester, polypropylene, rayon) fibers are also popular, but of course to make plastics, we need a variety of reagents and organic chemical feedstocks. So plastic fibers are going to be a relatively late introduction to the new time line concrete industry.
Wood fibers are the subject of current experimentation. They are added to cement shingles to increase ductility. Cement shingles have advantages in areas susceptible to wildfires. (Muhollem). Wood fibers have also been used in fiber-cement sidings (Mohr).
There are other natural fibers, too. The vegetable fibers include bast (flax, hemp, jute, kenaf, akwara, bamboo), leaf (sisal, henequen, pineapple, banana, elephant grass), and seed or fruit fibers (cotton, kapok, coconut husk “coir”). There are also animal fibers like wool. Obviously, the tropical fibers aren’t going to be readily available in seventeenth century Germany, at least at a low enough cost.
Direct incorporation of glass fibers is possible only if alkali-resistant glass is used; ordinary glass fibers don’t tolerate the highly alkaline environment of concrete.
Combination of fibers of different lengths or compositions can have synergistic effects (Banthia).
Composite Rebar-Reinforced Concrete
It is possible to fabricate rebar out of a composite material, a fiber-reinforced plastic (“FRP”). The purpose of the plastic matrix is to protect the fiber from abrasion and chemical attack, and to transfer loads to it. Hence it has to be able to bind to both the cement and the fiber.
The properties of FRPs are strongly influenced by the length, diameter, arrangement and composition of the fibers, and the composition of the matrix.
Matrix. The FRP has to be readily fabricatable into rods, which means the plastic (resin) is usually either thermosetting or thermoplastic (which can be recast). It is desirable that the rods be bendable on-site. The resins in commercial use are synthetic; the thermosets include polyester, epoxy, and phenolics, and the thermoplastics, polycarbonate, polysulfone, and polphenylene oxide. In order to explore these possibilities, we first have to reconstruct the plastics industry to the point at which we have fine control over the mechanical and chemical properties of the plastic.
There are a number of natural resins which might be tested as potential matrixes. But Humphreys warns that natural resins “generally lack the processing and performance characteristics sought after in a matrix resin.” (For desired characteristics of matrix resins, see Hale 49.)
Casein, a milk protein, can be precipitated from milk with heat and acid, and hardened with formaldehyde to make a semi-synthetic plastic. According to canon, it was made by the winter of 1631-32 (see Offord, “Bootstrapping,” Grantville Gazette, Volume 11, and DeMarce, “Songs and Ballads,” Grantville Gazette, Volume 14). It is probably too soft, and too vulnerable to water absorption and biological degradation to be a good FRP matrix.
Most of the research I have seen on natural resins has focused on their combination with natural fibers to make a biodegradable material. Biodegradability is great for bottles, but not what one wants in a skyscraper, bridge or dam.
Glass (“GFRP”) and Carbon (“CFRP”) composite rebars are commercially available nowadays. There is some information about glass and carbon fibers available in the Encyclopedia Americana.
Glass Fibers. A glass fiber-reinforced plastic (“GFRP”) is often called, somewhat misleadingly, “fiberglass”. The latter term actually refers to the “bundle” (cloth, tape, etc.) of glass fibers used to make the GFRP.
A GFRP rod has a higher tensile strength than one made of steel, and it is lightweight and corrosion-resistant. However, its mechanical properties are different enough from those of steel (they vary by direction, and it isn’t as stiff) that direct substitution of GFRP bars for steel bars is not possible. (CPPI).
Glass fibers are intriguing, because there is a down-time glass industry—we aren’t starting from scratch. Coarse glass fibers have been used for decoration since antiquity (Cooper, “In Vitro Veritas” (Grantville Gazette, Volume 5), and they are mentioned in Antonio Neri’s L’Arte Vetraria (Florence, 1612).
When glass is drawn into a fiber, it deforms, wiping out most of the weakening surface defects. The tensile strength of glass fibers with a diameter of 1/2000th inch is perhaps ten times that of bulk glass, and comparable to high tensile engineering steel. Glass fibers of those dimensions were first drawn by Griffith in 1920.
However, he was just testing the strength of individual fibers, not trying to produce them en masse. To make glass fibers (preferably, 5-20 micron diameter) on a commercial scale, we need to be able to produce a homogeneous glass (to control viscosity during drawing, and so the final product will have consistent mechanical properties), and we need glass drawing machinery which can reliably produce fibers of the correct dimensions. The Encyclopedia Americana “Fiberglass” article says to draw the glass through a platinum orifice plate, but platinum is not available immediately after RoF (it was considered a “waste” metal).
Moreover, the new fibers tend to stick to (and weaken) each other. Hence you also need to integrate, into the drawing operation, treatment of the fiber with a protective film (and in turn we need to identify a suitable chemical for that purpose). (Gordon, 74-76, 183-4). And once we have the glass fibers, we still have to find the right matrix.
I would have expected it to take some years to solve these problems. However, canon (Huff and Goodlett, “The Monster,” Grantville Gazette, Volume 12) says that sometime between June and November 1633, “fiberglass” is available, albeit at a high price. Georg Markgraf ‘s second airplane, the “Jupiter,” has an GFRP skin “made of a composite of fiberglass and resin.” Perhaps the “fiberglass” was hand-drawn?
Unlike the composite rebars discussed here, Georg’s aircraft body doesn’t have to have a lot of strength. It is “semi-monocoque” construction, meaning only part of the stress is borne by the body itself.
Carbon Fibers. Alternatively, it is possible to carbonize certain artificial and natural fibers. The first carbon fibers were made, by Edison (1879), from cotton or bamboo, for use as incandescent light bulb filaments. They were “extremely brittle” (Lee, II147).
High-performance carbon fibers were first made commercially available in 1959, at a price of over $500/pound (Jacobs 544). The usual starting materials were polacrylonitrile or rayon. Of course, we have to make these synthetic fibers before we can carbonize them.
The most interesting potential source for carbon fibers is pitch, which can be derived from oil or coal. Fibers can be pulled, like taffy, from the pitch, and these carbonized. In fact, if a high enough temperature is used, they can be converted to a particular form of carbon, graphite, with especially desirable properties (ACS; U. Kentucky).
Bear in mind that carbon fibers were leading-edge materials even in 2000 (Hegde; ACS), so even entertaining the notion of carbon fiber-reinforcement is going to give some readers apoplexy. But at least there is no doubt that the raw materials are available. The trick is going to be working out the manufacturing technology.
Hale 47 quotes 1998 fiber prices per pound as follows: E-glass ($1), S-glass ($5), aramid (e.g., Kevlar®) ($15-50), standard graphite ($17-35), high strength ($40), high stiffness ($65), and ultra-high stiffness ($275-650).
The great advantage of FRPs over steel is their high ratio of tensile strength to weight. This is especially important for vehicles, many parts of which can be made completely out of composites.
Unfortunately, FRPs suffer from a relatively high price and relatively low stiffness. Even with today’s production technology, these composites are much more expensive than steel ($3-4/pound for GFRP, and more for CFRP, compared to $0.32 for epoxy-coated steel. (Purdue).
In part, the high prices are because GFRPs and CFRPs are still produced in relatively low volumes (MIT). Arguably, they will have a better chance in the new time line, because steel itself is a specialty product in the seventeenth century. On the other hand, steel will probably be needed for the very machinery used to manufacture GFRPs and CFRPs. And the availability of concrete as a building material should reduce the demand for wood, which could in turn bring down the price of iron and steel (because in 1630, charcoal was over three-quarters of the cost of smelting iron; Sass, 162).
Moreover, civil engineers need a material which not only has high tensile strength, but which is also stiff. GFRPs are quite inferior to steel in terms of stiffness, and CFRPs can only be made stiff with substantial difficulty and expense. CFRPs also fail at a lower strain than steel. (Humphreys).
Hence, I think it may be more than a decade before we see substantial building use of GFRPs or CFRPs. However, their weight advantages are of particular moment to the aircraft industry, and that industry is likely to fund research which will eventually benefit builders, too.
Macro-Defect Free Cement
It doubtful that anyone in Grantville will know about it, and even less likely that they will know how to make it, but in the 1980s it was shown that one could essentially eliminate pores by use of the combination of a Portland (calcium silicate) or better, a calcium aluminate cement, a water-soluble polymer (e.g. polyvinyl alcohol-acetate copolymer), water and glycerine. Compressive and flexural strength increase ten-fold, toughness more so, and stiffness doubles. (Ghosh, 352-3).
Ferrocement. This material is similar to reinforced concrete, but there is no coarse aggregate in the concrete (so it is really reinforced mortar), and the reinforcement is closely spaced layers of small-diameter wire mesh and bars. The wire was originally steel, but other materials (e.g., bamboo) have been used. (Haussler; Robles-Austriaco). A crude form, with a single mesh layer, was invented by Lambot in 1855 (Jackson 28). Experimentation revealed that the weight of steel should be about 27 to 37 pounds per cubic foot.
“Micro-Reinforced Concrete” has been touted for blast resistance (Hoffman; Hauser; Chusid; Excendinc). From what I can tell, it is just ferrocement “on steroids”—supplemental cementitious materials, superplasticizers used to reduce the water-cement ratio, etc. I suspect there has also been some optimization of the mesh.
Seacrete. This was a highly experimental, concrete-like material as of the time of writing. Calcium carbonate, from seawater, is electrodeposited on a wire mesh which carries the necessary electrical current (Hibbert, USP 4246075). Seacrete is potentially as strong as concrete. There are a few catches. One is to get high strength, you have to use low currents so the rate of deposit is slow—you have to grow seacrete for a year or more to get a strength of 8000 psi. Another is that all the electricity needed is expensive (see “Seacrete,” Wikipedia). And third, there is no moving Seacrete, you grow it where you want the wall (or whatever) to be. So its use is likely to be limited to underwater or coastal structures.
Pykrete. This is a composite material made of 14% sawdust (or wood pulp) and 86% water. It is frozen to produce a concrete-like material. In World War II, there was a proposal to use it to make aircraft carriers (Project Habbakuk). A sixty foot experimental structure was constructed and floated at Patricia Lake, Alberta.
Concrete Cloth. This is a recent (2003) development, a dry concrete mix-impregnated fabric, usually bonded to the outer surface of an inflatable plastic (PVC). The theory is that the plastic is inflated, the concrete is hydrated, and the concrete cloth hardens into a dome shape, for use as an emergency shelter. (ConcreteCanvas)
Polymer concrete. A concrete-like material in which part or all of the cement is replaced with a polymer (e.g., polyester or vinyl ester resin, or latex). The polymer inhibits water absorption. It is often used for emergency repairs because it can obtain useable strength in minutes. Unfortunately, the polymer is expensive (and not immediately available post-RoF).
In Part II of this article, I will explain how concrete is produced, laid and tested, and what sort of structures it could be used to build.