The Voyage of the Iron Dragon

by Robert Kroese

  With any luck, they wouldn’t have need of a sizeable defense force again. The island was ideal for defensive purposes, and the local Carib tribes were unlikely to mount a large-scale attack for an island they apparently hadn’t considered worth settling. If Gabe had been successful in destroying the Viking army, it would be years before Harald had the resources to challenge them again—and he’d have to find them first. Only a select few in Europe knew they had fled to the Caribbean, and with the exception of Gabe, they were all now at Camp Aldrin.

  Months went by without any news from Europe. This was not unexpected: with access to the personnel and resources they needed on this side of the Atlantic, their ships no longer traveled east of Bermuda. O’Brien was tempted to send a ship to Höfn to find out for certain what had happened there, but it was too risky. If the ship were intercepted by an agent of the Pope or Harald, the crew might lead them back to Antillia. Maybe it was better not knowing, in any case. O’Brien assumed that none of those who had stayed behind had survived, but it was possible that they had overestimated the danger, or that Gabe had figured out a way to save Svartalfheim. Stupid, stubborn Gabe and his insistence on sending all their guns along with the evacuees. Maybe the attack had come but Gabe and a few others had survived and were even now living out the last of their days at the abandoned facility. Even if they were, O’Brien would never see them again: Gabe knew better than to follow them to the Caribbean.

  The one big concern, from a security perspective, was Camp Orville. A ship that had been expected to return before the evacuation had not arrived in time. The coxswain of that ship was Fritjof, who had been entrusted with a message for Aengus: since Svartalfheim was going to be abandoned, no further shipments of lumber or pitch would be needed there. All ships would henceforth be sent south to Camp Collins, their stopover point in Bermuda. O’Brien had suspected that Fritjof, delayed by bad weather, had opted to head directly to Camp Collins himself rather than return to Iceland, but their men in Bermuda said no ships had arrived from Camp Orville. As the weeks went by with no word from Aengus, speculation arose that Orville and Wilbur had at last fallen victim to LOKI—probably in the form of rebellious natives. In the spring, O’Brien sent a ship to Nova Scotia, but the crew returned a month later, saying that both camps had been abandoned and the main lodge at Orville had burned down. The coxswain, who had been to Orville three times before, was unable to locate any Mi’kmaq he recognized, and the crew, fearing for their safety, boarded their ship to return to Antillia.

  All the satellite locations besides those in North America and Bermuda had been shut down at the same time as Svartalfheim. Sensitive personnel were put aboard ships bound for Höfn or Bermuda. Those whose services were no longer needed were given a month’s pay in silver and released. Rumors of the Dvergar would undoubtedly spread, but none of the ex-employees knew enough about Pleiades to be cause any problems.

  Meeting the dietary needs of the nearly 2,000 people at Camp Aldrin was the biggest challenge the Committee faced initially. Antillia was too small for hunting or farming, and the ships hadn’t had the space to bring live farm animals anyway. They brought with them enough provisions to last a few weeks after reaching their destination, mostly wheat flour and dried fish and meats. For the foreseeable future, they would be heavily dependent on Camp Erhardt, in Bermuda, where several large farms had been established. Longer term, O’Brien hoped to be able to trade for food with the local Indian tribes.

  Reyes’s health responded well to the warm climate. She fared better than expected during the three-week voyage to the Caribbean and came out of her coma six weeks after their arrival. Now fifty-six years old, she remained weak for several months, lacking the endurance and concentration for more than the most basic tasks. O’Brien remained in charge of the program, but Reyes was included in meetings whenever possible. Her recovery buoyed the flagging spirits of the community after the loss of their homes and so many of their men at Svartalfheim.

  After dormitories and a few other essential buildings were constructed, the main priority at Camp Aldrin was to establish shortwave communication with the other facilities. The engineers had already perfected the process of making vacuum tubes, but the project of setting up radios at the satellite facilities had been delayed by the evacuation. They had successfully transported and moved their vacuum tube fabrication facilities, and the engineers set about building transceivers and amplifiers. A year after their arrival at Antillia, they tested their first long-distance shortwave communications between Camp Aldrin and Camp Hughes, nearly 2,000 miles away.

  Word had been sent to Camp Hughes of the abandonment of Svartalfheim; oil shipments now went directly from the platform in the Gulf to Antillia. A refinery was set up at Camp Aldrin, where the petroleum was purified into diesel, fuel oil, gasoline, kerosene, lubricants and asphalt. Petroleum byproducts were also used to make plastics, rubber and vinyl. Two new wells were drilled at Camp Hughes, serviced by the same offshore platform, to meet the increased demand. Oil was no longer needed for heating, but it was used for virtually everything else: smelting iron, running machinery, fueling airplanes and rockets, and powering generators. Without a ready supply of geothermal energy, even the process of splitting seawater into hydrogen and oxygen had to be powered by a diesel generator. And since both liquid oxygen and liquid hydrogen had to be constantly refrigerated, several barrels of oil a day would be needed just to keep stored propellants from boiling away in the Caribbean heat.

  Despite the loss of Orville and Wilbur, Pleiades had no shortage of ships: it had taken nearly three hundred vessels—a collection of knars, karves and snekkjas—to get all the evacuees and supplies to the Caribbean. Once shelters had been constructed on Antillia, many of these ships were repurposed for exploration and resource acquisition. Over the next year, two new camps were established: an iron mine in Trinidad and a bauxite mine in Jamaica for the production of aluminum. The iron ore and bauxite were processed at the mine site; finished iron and aluminum bars were shipped to Antillia, where they would be smelted into tools, machine components and building materials. A small factory was set up at Camp Aldrin to turn sand into glass. Other than food, most of the other resources they would need—such as wool, copper, titanium and gold—had been brought in sufficient quantities from Iceland.

  Construction of buildings at Camp Aldrin proceeded rapidly. There was little usable lumber on the islands in the area, so most of the early buildings were constructed of lumber pulled from snekkjas that were no longer needed for raiding. Later buildings were built with concrete made from cement produced at a plant near the iron mine in Jamaica. Once they’d built a steel fabrication facility, larger buildings were constructed of steel frames and corrugated steel sheets. A medium-sized hurricane would tear the flimsy steel buildings apart, but the odds were against one hitting Antillia within the next few decades, and the buildings weren’t intended to last longer than that. Most of the buildings were clustered at the east side of the island; the launch pad for the Iron Dragon would be constructed at the far west end. An asphalt road would eventually connect the settlement to the launch site. Alongside the road would be an asphalt-covered concrete runway for testing airplanes. Before any of this could be done, though, earth-moving machines had to be built (or reassembled, since many components had been brought from Camp Armstrong) to raise and level the uneven, rocky ground.

  Reyes gradually improved over the next two years, and O’Brien gladly handed the reins of Pleiades back to her. With her recovery came a change in mood among the Eidejelans. Building a rocket capable of reaching orbit no longer seemed like an impossible dream. Gone too was the secrecy of the past decades: engineers now spoke openly of the project, as there was no one around for 3,000 miles but primitive tribes of people who didn’t even speak their language. Reyes still insisted on calling the project Pleiades, but every teacher and construction worker on Antillia knew the name Iron Dragon.

  At the same time, their hard-fought successes thus far rein
forced in their minds the stakes of their project and the possibility of failure. On some level, everyone involved in Pleiades—except perhaps Alma—had adopted a sort of defensive myopia, refusing to think about the fate of their project. There were so many ways it could fail, falling victim to external enemies, internal dissent or simply some technical problem that proved to be insoluble in medieval Europe, that they’d been forced to focus on whatever challenge they were currently facing. They had pressed on because as long as their goal didn’t seem absolutely impossible, they felt an obligation to keep going.

  Now, though, the major obstacles had fallen: they had assembled and trained a team of engineers, located the material resources they needed, and built the tools and machines to fabricate the rocket. They’d built vacuum tubes and turbopumps and designed and flown airplanes and small rockets. And they’d done it all right under the noses of the Cho-ta’an and the powers-that-be in Europe. If they failed now, it would be because of some minor oversight or fluke accident—one of the millions of little things that could doom a rocket launch.

  Theoretically, they had several hundred years in which to keep trying, but the Committee was in agreement that the project would be unlikely to survive the death of the two remaining spacemen. In particular, the project was highly dependent on the loyalty the Eidejelans felt for Reyes. The Committee had seen how close Pleiades had come to falling apart when Reyes was in a coma; if it hadn’t been for the imminent threat of invasion to unite the Eidejelans, Gabe and O’Brien wouldn’t have been able to keep the project going. There had been grumblings for years among some of the engineers that they’d be better off using their technological knowledge to overtly change history rather than acceding to the inevitability of a long and destructive war with an alien race of which none of them had any firsthand knowledge, and the new openness at Camp Aldrin only fanned the flames. Their proximity to an untamed continent didn’t help: there was nothing preventing a rebellious contingent from boarding a karve and heading to Cuba or Florida and setting themselves up as gods among the natives.

  As always, their most insidious enemy was LOKI: historically, no evidence had ever been found of an ancient iron mine in Jamaica or a gold mine in Trinidad, which mean that unless there had been a global conspiracy to obscure the truth, those facilities had at some point been destroyed. When and how that would happen was anybody’s guess. That Reyes and O’Brien were able to spend their time worrying about such abstract concerns was, of course, a symptom of their own success: Harald’s army was no longer a threat, and it had been so long since they’d seen any sign of the Cho-ta’an that Reyes allowed herself the hope that they might actually be left in peace to finish the project they had begun so many years ago.

  Chapter Forty-five

  Rebuilding the aeronautics program took until their fifth year at Camp Aldrin. All the engineers had made the voyage from Svartalfheim; the difficult part was constructing new buildings and rebuilding all the machinery they’d left behind. A runway nearly a mile and a half long was constructed between Camp Aldrin and the launch site. New propellant tanks were constructed and the turbopumps were reassembled. Most of the airplane prototypes they’d built over the past few years had to be sunk in the Atlantic, as there was nowhere in the Caribbean to land them. The exception was the P-51 Mustang, which O’Brien couldn’t bear to destroy. It had been disassembled and its components parceled out among several knars.

  Initial testing of rockets had begun in and around Iceland several years earlier. These were mostly solid-fuel rockets no taller than a person. They were unmanned and never made it beyond the stratosphere. Very few were recovered; most were lost in the North Sea. During the last two years before preparations for evacuation began, they had begun testing significantly larger rockets using liquid hydrogen and oxygen and propellants, with mixed success. Most failed to launch, usually due to a problem with the pumps. A few exploded on the launch pad. Two made it into the stratosphere, but the Eidejelans had yet to get an object into orbit. For that, they needed to build a multi-stage rocket, and that feat remained beyond their abilities while they were at Svartalfheim.

  Alma’s engineers had successfully built kerosene-powered turbopumps capable of liquefying hydrogen and oxygen for storage on the ground, but building pumps powerful enough to overcome the pressure in the combustion chamber of a rocket—and compact and light enough to be carried by that rocket—proved more difficult. The pumps needed to spin at nearly 37,000 rpm, injecting fuel into the combustion chamber at a rate of a thousand pounds per second. Building such pumps would be an impressive engineering feat even in the twentieth century, and trying to do it with medieval technology was unthinkable. Still, Alma’s team persisted, fabricating tools capable of ever-increasing levels of precision, confident that given enough time, the problem could be solved.

  Developing a reliable ignition system proved to be another thorny problem. One solution would have been to switch to a hypergolic fuel system for the second stage. Hypergolic propellants, such as the combination of dinitrogen tetroxide and hydrazine, spontaneously combusted upon being mixed together, eliminating the need for a separate ignition system. Later versions of the Titan II rockets had used hypergolic fuels, simplifying their design. The liquid oxygen/liquid hydrogen combination of propellants had been selected primarily because of the ease with which it could be produced from seawater, and it was now beginning to look like that may have been a mistake. Hypergolic propellants had their own problems, however: they had a lower specific impulse than liquid hydrogen/liquid oxygen, meaning that more fuel would be required for the launch. Hypergolic fuels were also more corrosive and toxic, requiring more effort and expense to be put into storing and handling them safely. In the end, Reyes recommended that Alma’s team stick with their original choice.

  One problem that was well-known in relation to rockets using liquid propellants in general and the Titan rockets in particular, known as “pogo oscillation,” was a positive feedback cycle in the fuel system that could potentially tear the rocket apart. This cycle could occur for a number of reasons, such as a surge in engine pressure increasing back pressure against the fuel coming into the engine, reducing engine pressure, causing more fuel to come in and increasing engine pressure again. Pogo oscillation was, ironically, such a well-documented problem that although it presented a challenge, solving it—primarily through the implementation of dampening mechanisms—proved relatively straightforward.

  Another difficult but relatively straightforward problem was making the propellant tanks strong enough to hold gases pressurized to 5,000 psi but not so heavy that they impeded the rocket’s journey to space. The tanks, as well as most of the structural components of the rocket, were made of aluminum, because it was lighter than steel. The tanks were fabricated from finely machined aluminum panels joined with rivets and then welded together. The goal was to build a rocket that had a mass-to-fuel ratio of less than one to ten, meaning that every ten pounds of fuel required less than one pound of structure to carry it. Alma’s engineers achieved this by using an aluminum-magnesium alloy, as well as developing a process for fabricating sheet metal plates with a variation in thickness of less than a hundredth of an inch, to ensure that the optimal amount of metal was used for the tanks.

  No matter how light the rocket was, however, there was no getting around the need for a two-stage launch vehicle. Carrying empty fuel tanks into orbit was unnecessary and would require more fuel than those tanks could carry. This meant solving the problem of stage separation, which combined the problems of timed ignition and propellant pumping, while adding some additional complications. After the first stage separated and before the second stage ignited, the rocket would momentarily be in freefall, meaning that the second stage fuel and oxidizer would be weightless. To settle the propellants at the base of the tank so they could be fed into the combustion chamber, small solid-propellant rocket thrusters on the second stage would be fired. Propellant would then be pumped into the combustion chamber a
nd ignited, pushing the second stage into orbit. Additionally, small thrusters on the third stage would fire to slow it down, separating it from the second stage. The process would be repeated when the second stage fell away and the Gemini vehicle continued into orbit.

  Most of the other problems that remained to be solved related to the lack of miniaturized electronic components. Each of the four IDL-issued cuffs contained a computer that by far surpassed the entire processing power of NASA in the 1960s, but using them on the Iron Dragon meant connecting them to other components. The twenty-third century optical data transfer connection used by the cuffs was well-documented, but constructing a connection that would allow the cuffs to talk to other components was a project several orders of magnitude more complex than building a Titan II rocket. One of Alma’s engineers had experimented with programming the cuffs to produce audio signals of different frequencies which could be amplified and then translated into binary code by a machine that used filaments of varying resonant frequencies to receive the input. The binary code was then transferred to punch cards, which could be read by a person with sufficient training—or, theoretically, by another machine. The process proved so slow and cumbersome, though, that Alma deemed it would be more sensible to build computing machines from scratch rather than crippling the cuffs by tying them to devices with less than a billionth of their capacity.

  Vacuum tubes had quickly become ubiquitous at Camp Aldrin. As the need for thousands of finely machined components arose, machine operators found it impossible to meet demand without the ability to program certain repetitive behaviors into the machines. Even at Camp Armstrong, engineers had begun to equip some machines with basic mechanical logic circuits so that, for example, a bolt-making machine would cut a bolt to exactly three eighths of an inch, drop the bolt in a bucket, and then start on the next bolt without a machine operator having to intervene. As parts became more complex and more numerous, and demand for precision increased, programmable machines employing vacuum tubes were added to the logic circuits.