Every day we walked into the city and dug into basements and shelters to get the corpses out, as a sanitary measure. When we went into them, a typical shelter, an ordinary basement usually, looked like a streetcar full of people who’d simultaneously had heart failure. Just people sitting there in their chairs, all dead.
—Kurt Vonnegut, on removing corpses from bomb shelters in Dresden during World War II
Nine people died during the most massive explosion of George Washington Rains’s powder works. Eight of them, seven men and one boy, were inside the building containing the powder when it blew. According to Rains, the tragic explosion likely occurred because one of the men, after lighting a cigarette, callously flicked his extinguished but still-hot match onto the powder-dusted ground. Smoking in the literal powder keg was forbidden, but the usual supervisor was off work that day. The stray granules littering the floor ignited, one at a time, passing the flame like water flowing over the ground, and the deflagration radiated outward to reach the large volumes of powder in storage. The smoke plume from the resulting explosion could be seen for miles. The eight people inside the building were, not surprisingly, “reduced mainly to small fragments and dispersed.”
However, Rains’s brief description of the sentinel, the lone man standing guard outside the building, presents the key to understanding blast trauma: The sentinel was found dead but intact. “His body was not otherwise disturbed,” according to Rains; he had not been hit by shrapnel, and more important he seems to have been found dropped to the ground where he had stood.
Author Kurt Vonnegut more famously described the same phenomenon. He spoke in an interview about his time in Germany during World War II, right after the firebombings that decimated Dresden. His job had been to excavate the bomb shelters and basements to remove the rotting corpses before the entire city started to stink of human putrefaction. The people he found had usually died without moving, without any signs of struggle, and were often still seated peacefully in their chairs. They were not outwardly wounded; they were not blown wildly across the room.
There are multiple ways for victims to die in a firebombing, and Vonnegut’s cases cannot be retroactively declared to have all occurred solely because of one single cause. However, they share the same key descriptors as the sentinel in Rains’s factory blast: undisturbed, no external injuries, simply dead where he stood. To a blast researcher, this scenario sets off all the mental alarms. It starts our heads screaming that we should at least suspect what is called by our field a “primary blast injury.”
Medically speaking, the injuries from an explosion are neatly categorized into one of four tidy bins. A blast victim can receive only one type, or they can receive a sloppily assorted grab bag of trauma containing any delightfully painful mixture of the four. Helpfully, the injury types are numbered for easy reference: primary, secondary, tertiary, and quaternary.
Quaternary trauma is a sort of “other” category. It covers the grotesque and myriad injuries that can occur as the result of a blast, but do not necessarily occur consistently or predictably from every type of blast. The umbrella of quaternary trauma covers adventures like burns, radiation sickness, and infectious biological goodies that may have been intentionally dispersed by the explosion.
Tertiary trauma is the injury type that most people expect. It is the classic scene from the action movie: Our sweaty, grease-covered hero flips on the wrong light switch and a massive red-and-yellow fireball propels him wildly flailing across the theater screen. When he stands up, he grimaces. He is unswayed in his determination to save the helpless yet busty heartthrob with the long flowing hair, but first he needs to stretch his lower back so the audience can appreciate how much pain he is shaking off in order to continue this mission. His lower-back pain is a tertiary injury: an injury incurred from “whole body translation,” which is engineer-speak to describe someone getting physically thrown by the blast.
Secondary injuries are unfortunately an overwhelmingly common injury type, especially with the rise to prominence of improvised explosive devices (IEDs) in the combat zones in Iraq and Afghanistan. This injury type is the result of objects getting thrown and hitting a person because of the blast. The objects are often shrapnel from the charge casing, or sometimes nails and ball bearings included solely for the purpose of inflicting more harm. Secondary injuries frequently take the forms of trauma to the limbs, cuts deep enough to reach the skeleton, and amputations.
These three injury types—secondary, tertiary, and quaternary—are logical, meaning that they make obvious sense and even people with zero blast experience can predict that they are expected possibilities. In contrast, primary blast injuries—the kind likely incurred by Rains’s sentinel, and possibly incurred by the victims in the Dresden bomb shelters—are an impressive, strange, and horrifying fluke produced by the bizarre physics of explosions mixed together with human frailty. Primary injuries result from the pressures produced by an explosion, usually—but not always—because of a shock wave.
To understand how a shock wave maims, first it is crucial to understand how a shock wave is born. Normally, sound moves like billiard balls on a massive, smooth felt table. First, a noisy event occurs, like an impact or a vocal cord vibrating into movement. A gas molecule in close proximity to the action gets pushed away: This is the cue hitting the cue ball. The cue ball travels outward until it hits the 4-ball, another gas molecule. Clunk. They impact, and the cue ball transfers some of its energy to the 4. Both balls now move, slightly slower and in an outward direction, until they impact other balls, hitting their next closest neighbors. The overall wave front of the motion moves forward, but each individual ball travels only slightly across the table. The motion gets passed outward, expanding and slowing just a bit with each collision as the leading edge of movement travels across the table.
Sound travels outward, each molecule of material transferring energy to the next, growing in reach but decaying in strength as it moves. Eventually it hits an ear and gets heard, or a wall and echoes back toward the source. It moves the same way in water as it does in gas, except faster, because the molecules start out closer together in the denser liquid.
A shock wave occurs when the pool cue is placed in the hands of the most furious, most irate patron in the hall. He is a high explosive. Apoplectic and red-faced, the explosive burns quickly. In fact, the burn front moves through the entirety of the explosive much faster than normal sound. Therefore, the entire reaction happens too fast for the gaseous products created by the burn to expand outward in a normal way. The material is burned and gone before the balls can travel outward on their own, too quickly for them to thwack their neighbors at their natural speed. The whole charge has reacted, is consumed, becomes a tiny, compressed, superheated ball of hyperpressurized gas before the 4-ball ever gets the message. The resulting gases expand all at once, together, suddenly, violently, and the pool cue is shoved, hurtled, rammed down the length of the table, picking up ball after ball and adding them to the front of the wall of molecules moving forward, picking them up faster than they can move on their own.
This is how a shock wave develops. The molecules accumulated at the wave front are densely packed, shoved together tightly by the gas urgently expanding behind them. They are so densely packed that each molecule can reach its neighbor more quickly than it could in a normal situation, and so this unique wave moves faster than the speed of normal sound.
The molecules downstream get hit without warning. In its purest form, the shock wave goes straight from zero to its maximum pressure in an instant; on a graph it is a vertical line followed by a sloping decay back down. If it were a car it would go from 0 to 60 in exactly zero seconds.
When they reach high enough pressures, these waves can disintegrate everything in their paths. The substantive fabric of objects gets jerked into motion fast, too fast, by the instantaneous rise of the shock, and they break apart into chaos like a fragile porcelain teacup getting hit by a rapidly moving concrete floor.
Most of the human body handles mild to moderate levels of shock surprisingly well. Severe pressures will cause tissue disruption, which is a polite phrase that describes a horrifying concept. However, the lower-pressure shock waves can travel through most of our anatomy without harm. These waves can move straight through water without much chaos and disruption, and human bodies are, after all, mostly water. It’s the gas pockets inside certain organs that cause the real drama.
In the chest wall, which is mostly water, sound moves at roughly 1,540 meters per second. In a gas pocket, which is basically air, it moves at roughly 343 meters per second. Therefore, waves moving through the body that hit a gas pocket are forced to slow down at the interface by about 80 percent. And as they are forced to slow, that energy must get transferred somewhere.
To make a bad situation even worse, the human lungs are basically two lumps of bubbly fluid. The bubbles are the alveoli containing the gas that gets moved in and out by the lungs during breathing, and the blood vessels and tissues surrounding the alveoli are curvy lines of fast water surrounding a network of pockets of slow air. Together, the structures make a maze for the pressure waves. As the pressure waves navigate the circuitous pathways of the labyrinth, moving quickly through water lines but changing direction and bouncing off walls of air, they are forced to slow down even more. Way down, much slower than they would if they could travel straight through a simple pocket of normal, uncomplicated air. They slow down to a measly 30 meters per second. The phenomenon of this extreme slowdown in the water-air maze is called “the hot chocolate effect” after the delightful frothy bubbles of the delicious beverage.
If the shock wave is like a runaway semi-truck traveling through the watery tissue of the chest wall, racing headlong at over 1,540 meters per second down a mountain, then the lungs are the gravel pit of a runaway truck ramp. The truck is slowed to 30 meters per second, less than 2 percent of its prior speed—it has no choice—but its energy must go somewhere. The gravel flies outward, everywhere, a rampant hailstorm of dispersed energy and momentum. So too respond the delicate tissues that form the walls of the lungs. They rupture and shred, and blood sprays into the alveoli, filling the precious gas pockets needed for breathing. This process is called spalling.
Gas pockets in the intestines can cause a similar problem, leading to bruising and tearing of the intestinal tract, but the intestines have a higher threshold for injury and are less commonly damaged than the lungs. The same is true of some of the smaller bones in the skull, particularly the ones that form the fragile archways around the sinus cavities. These bones will on occasion show spider webs of fracture from primary blast, but they are sufficiently difficult to injure that these patterns are typically only seen in autopsy reports.
The delicate tissues of the brain can also experience forces from the blast, which can cause traumatic brain injuries without ever disrupting a perfectly undamaged skull. These injuries occur via a different and far more complex mechanism that is still the subject of active research. Critically, the brain remains intact after a primary blast injury, and the only potential sign of trauma is a faint inkblot of blood that may be spread across its surface.
Fatalities from primary blast occur at lower pressures than the pressure levels required to translate a human body. To rephrase that in plain English: A person will die, choked with blood, from a shock wave that was far too weak to move them.
During World War II, the bombings of German cities led to civilians dying in exactly this way. It was common for blast victims to be found dead, looking externally peaceful, with no other signs of trauma except the occasional trace of bloody foam around their noses and mouths. These victims often had no damage to their muscles or bones.
A shock wave will have a 100 percent fatality rate if it reaches a peak overpressure level of 350 kilopascals and lasts for 7 milliseconds (7⁄1000 of a second) before it drops back down to zero. When a blast hits a typical adult, it will be impacting approximately 1 square meter of surface area. The 350 kilopascals of pressure multiplied by 1 square meter of area gives a resulting total force of 350,000 Newtons. This force is equal to 492 Rachels, the narcissistic unit of force I made based on my weight while standing still, and the pressure is equal to the water pressure at a depth of 25 meters below the surface of the ocean. However, it lasts for a mere fraction of a moment.
Using a quick back-of-the-envelope calculation, engineer-style, this shock wave could move a 72 kg person a maximum of 0.2 meters—less than the length of an average adult’s foot. That is assuming no friction, and no resistance to motion of any kind. Therefore this wave, even though it would be universally lethal, provides barely enough impetus to knock someone over.
Recreational scuba divers plunge to depths below 25 meters every day without harm; the pressure level alone is not the problem. But in a blast the traveling shock wave rises quickly from zero to maximum pressure, and it also impacts one side of the body rather than simultaneously immersing it from all sides. The shock wave jerks the blood and tissues into motion, even though it is a small amount of motion, before they can adapt.
The sentinel falls where he stood. The citizens, huddled in their bunker, die seated in their chairs. Autopsies frequently show no skeletal damage from the primary trauma, no broken bones or fractured skulls, just blood in their lungs, and possibly in their intestines or on the surfaces of their brains. They are unmoved. They are un-translated. Our action hero, thrown across the room by the winds of an explosion, is not standing up and dashing off after a mild back injury; he is dead from blast lung, 100 percent of the time.*
Each major war comes with advances in military technology, and fresh young soldiers almost always become the first victims. During the Civil War, the Rains brothers brought mayhem with their inventions of land mines and underwater torpedoes. By World War I, machine guns and high explosives were deployed to the trenches, rendering cavalry charges obsolete and introducing the first widespread patterns of blast trauma. World War II brought submarine wolf packs roaming the seas and atomic bombs, and along with them came innumerable medically novel in-water and nuclear-blast victims. In Iraq and Afghanistan, the rise in popularity of IEDs caused so many blast-induced traumatic brain injuries that these injuries became known as the “signature wound” of the conflicts.
During peace, militaries seek to prepare for the next war by developing innovative, more effective ways to kill. Then, when the war starts, soldiers are shipped home with different injury types than in the previous wars. Medical scientists and researchers, who take a long time to produce answers because of the tedious, exacting nature of scientific research, struggle to keep up with the technological advances in weaponry.
Blast casualties are uncommon in the civilian world, especially when compared to other traumatic events like car crashes, so the way the field moves forward is usually because a war starts. Soldiers come home wounded or dead. The country wants to know why, and decides it is finally willing to fund the science. To explain it, and to please oh please hopefully stop it, people like me head to the lab to once again take up our Sisyphean race against the development of new weapons. When the conflicts are over and the dead seem to rest in peace, we are the ones who can be found sifting through the wreckage, still trying to figure out what really happened.
During the violence in the 1970s in Northern Ireland, soldiers and civilians alike were routinely hit by blasts. Two British military scientists studying the casualty reports began to notice a disturbing pattern.
As British military physicians they treated only the British soldiers, so they saw one combat narrative repeat itself over and over again. A gelignite charge was placed. This soft, moldable charge carried the chemical heritage of black powder, as it used both potassium nitrate and wood pulp in its reaction. The explosive charge, originally manufactured for the innocent purpose of blasting open a commercial mine, was instead detonated in a public place. Its shock wave propagated outward into the lungs and brains of nearby human beings, along with jagged flying missiles.
The physicians, Graham Cooper and Susana Mellor, began to track the patterns of the cases. By 1984, after fourteen years of slowly accumulating macabre forensic reports, they had reached a surprising conclusion: Those wearing the rubber foam–based body armor of the British military may have been less likely to die from the secondary trauma of shrapnel and blast debris, but they unexpectedly seemed more likely to die from the primary trauma of blast lung. One hundred and forty of the soldiers in their case reports died without a single sign of any injury beside blood in their lungs.
The trend seemed counterintuitive because the life-saving vests provided an additional barrier between the soldiers and the shock waves, a line of extra material that should have guarded the lungs from the blast. Logically, it seemed that any additional barrier around the soldiers should reduce the risk of blast lung. It took another several years, but eventually Cooper was able to set up an experiment to explain the anomaly.
The burst of an overinflated Mylar party balloon releases a small but legitimate shock wave. The air inside the balloon is pressurized slightly compared to the outside, squeezed together by the balloon’s strong walls, and when the gas is released suddenly it travels outward in the same manner as the gases created by high explosives. The mini shock wave gets heard as a pop.
Volcanic eruptions also suddenly release pressurized gas, and thereby create shock waves. The explosive eruption of the volcano Krakatoa in 1883 produced a shock so powerful that it was heard almost 3,000 miles away and circumnavigated the globe seven times.
High explosives are not the only way to make a shock wave. They work by creating a ball of high-pressure gas, but their rapidly burning chemical fury is not the only way to create a ball of high-pressure gas.
Scientists take advantage of this convenient fact by using devices called shock tubes. These strong metal tubes have one closed, sealed end that gets pressurized, called the driver. The driver is separated from the rest of the tube by a replaceable plastic membrane. This membrane, usually Mylar, just like a fancy party balloon, pops when the driver is sufficiently pressurized. The pressure releases and propagates down the rest of the length of the tube, forming into a beautiful, smooth, and predictable shock wave without most of the mess and danger of high explosives.
Cooper mounted his hefty shock tube to the ceiling. He let the long metal tube hang from the rafters, aiming its lethal maw downward into a gigantic tub of water. He secured one disk of material on the surface of the water at a time, either foam rubber, Kevlar, or thin sheets of copper, and he repeatedly triggered the shock tube. With each building-shaking pop he measured the pressure waveform that propagated into the water.
The water represented the chest wall, and Cooper’s elegant experiment showed that the foam rubber used in the British body armor of the time was actually amplifying the shock waves being transmitted into the bodies of the soldiers. When he blasted the water through a disk of foam rubber, the peak pressure levels in the water were about 25 percent higher than when he blasted the water alone with no rubber on top. The material of the body armor increased the overall amount of energy transmitted into the simulated “chest wall” by 230 percent. Normally a large portion of the shock wave would reflect directly off a soldier’s chest, but a fluke of the bubble-filled rubber softened that reflection and allowed more to carry forward into their bodies instead. Cooper’s experiment proved that the suspicious pattern of trauma was real, that the soldiers wearing body armor were at higher risk of primary blast injuries. He and Mellor had been right.
Cooper not only proved that the effects of blast could propagate straight through solid materials, but he showed that the wrong choice of material, the wrong configuration between a soldier and his environment, could in fact amplify the fatal potential of an explosion.
Modern-day body armor is designed with this effect in mind, and the Kevlar used in most bulletproof vests also massively reduces the amount of blast pressure that can get transmitted into the lungs of the wearer. It is actually because of this protection that traumatic brain injuries have become so common in Iraq and Afghanistan; the Kevlar protects the lungs of the wearer, so they survive a blast that should have killed them from primary lung trauma. However, their brains are still vulnerable. In a cosmic twist of fate, Kevlar, now being used to protect people from explosions, was invented by chemist Stephanie Kwolek while she was working at DuPont, a company founded by explosions.
A shock wave, by definition, skyrockets in an instant from zero pressure to its maximum pressure. The rate of increase is literally infinity. Once you can measure the rate of increase because it occurs over some known length of time, then the wave is no longer technically a shock. Instead, it is simply a pressure wave.
Pressure waves are less likely to cause damage because they allow their targets that little grace period of a few extra milliseconds’ time to adapt to the coming impact. The material might be able to flex and respond without failing, like Silly Putty, which can stretch when pulled slowly instead of tearing when jerked rapidly. Structures are more likely to survive, but human beings, unable to adapt to these rapid time scales, can still be injured and killed by sharp-rising pressure waves.
When the US Navy performs ship shock testing with high explosives, nobody is allowed to loiter on the deck of the ship being tested; it floats alone and unoccupied for safety reasons. The shock has too much potential to tear and destroy the vessel. It is too risky. However, when the testing is performed with air guns, which create fast-rising pressure waves that aren’t quite shocks, curious engineers and scientists are sometimes allowed to stand on the deck to feel the ping of the high-pressure wave bouncing off the hull.
The data suggest that 2 milliseconds is the cutoff when predicting risk to humans. In other words, any pressure wave that reaches its maximum in less than 2 milliseconds can be considered to maim and kill like a shock wave. Scientists established this limit using an influx of blast-research cash that came with the escalating tensions of the Cold War.
Our lungs are fragile. We bleed easily. A “not quite a shock” wave is still enough to destroy us.
At the tail end of the Gulf War, a military blast researcher triggered a controlled detonation in his laboratory. The charge he detonated was a small lump of C4. It sat on the floor in the center of an enclosed room, a room designed to mimic the inside of a US Army tank. Armor-penetrating missiles had been causing blast-like trauma to American soldiers inside tanks during the Gulf War, and the goal of this experiment was to understand how. Coincidentally, the room was also roughly the size of the inside of the Hunley, albeit a different shape.
The researchers wanted to see how the shock waves behaved when they were trapped. They already knew that the pressures would bounce wildly off the walls, ceiling, and floor, like a sound echoing in an empty room, and that each time the echoes intersected with one another they would combine and amplify. They knew that, like a whisper reverberating loudly through an empty cave, the blast waves, when trapped, would rebel against the walls and build on their own reflections to grow in magnitude. But what they needed to know was this: After all that echoing, would they still kill?
Sheep were positioned at various locations around the room, suspended from webbed harnesses. They were alive but anesthetized by veterinarians into a pain-free unconsciousness. After the detonation of the charge, a military blast researcher named Johnson and his colleagues examined the externally uninjured sheep for signs of lung trauma. They discovered that the patterns of trauma were much worse than what would have occurred if the same charge had been detonated in an open field, because the reflections had massively increased the lethal potential of the blast.
The shapes of the waveforms that were measured inside the room were crucial: The waveforms were no longer polite and tidy shock waves with an infinite rise followed by a smooth, sloping decay. Instead they were jagged monstrosities, with one or two large peaks somewhere inside the long time span before the pressures finally dissipated. The rise times of the big peaks were under the 2-millisecond cutoff. The unfortunate injuries of the unconscious sheep showed that the jagged pressure waveforms were still inarguably lethal, that they were still essentially comparable to shock waves for the purposes of causing and predicting human injuries.
Only a few years before the 1995 discovery of the Hunley, Johnson and his colleagues reached a determination that would prove crucial to understanding the fate of her crew. They determined that the risk of blast injury was dependent not just on the first wave passing into an enclosed space, but on the maximum pressure level that was achieved at any time while the blast wave ping-ponged around within the echoing volume.
These conclusions were again consistent with the case reports from World War II. In one particular incident, a shell detonated just outside a train station bunker. Many people took shelter in the tunnel that led down to the station, hoping its walls would shield them from the effects of the bombs. These unfortunates were found dead exactly where they had last been in their final moments, still standing up and clutching one another for comfort. They had been protected by the tunnel walls from any shrapnel, from any secondary injuries to their limbs or bones, but nonetheless the shock wave had reverberated down the narrow passageway. For many of them, the only trace of injury was a slight dust-covered froth at the mouth.
Johnson’s conclusions would again be reaffirmed in the most unfortunate possible way only a few years later, in a separate conflict, when the same pattern was revealed in the victims of terrorist bombings in Israel. Bombs set off inside buses resulted in over six times the fatality rate when compared to similar bombings that used nearly identical charges, but instead occurred outside in open, unconfined spaces. Furthermore, every single one of the bus victims died because of pulmonary failure. Even though the bus victims had experienced far less secondary wounding than the open-air victims from shrapnel and projectiles, the reflections inside the enclosed space had nevertheless sealed their fates.
Iraq and Afghanistan too brought new blast scenarios. American vehicles were frequently subjected to roadside bombs. Naturally, this scenario also warranted scientific testing, led by a scientist named Howard Champion.
Seventeen kilograms of C4 is a little cube of nondescript gray putty, roughly 25 centimeters per side, and unless you understand its power it is unexciting to look at. The particular pile that was sitting on the ground for an experiment a few years before 2009 had a target: an armored vehicle. As the clock ticked down to the moment of detonation, the rugged armored military vehicle sat quietly a mere 3 meters from the pile.
Vehicles and inanimate structures do not have lungs; they cannot experience spalling of blood. They are destroyed differently than human beings. Structures are far more resistant to shock than the fragile tissues of the human body, and it takes roughly a thousand times higher pressures to “kill” them.
The pressure gauges positioned immediately outside the armored vehicle measured a long-lasting blast with a maximum pressure of a blistering 760 kilopascals, more than double the limit where every person exposed would have died of fatal lung trauma. However, unsurprisingly, the armored vehicle, in contrast to a frail human body, was fine.
Even though the vehicle itself was fine, the squiggly lines produced by the gauges inside the armored vehicle were stunning. These lines showed a sharp-rising pressure waveform, a waveform that was not quite a shock wave but was still fast-rising enough to kill, with a long and jagged tail. The peak pressure inside was only 48 kilopascals, giving theoretical occupants a mere 4 percent chance of serious blast lung injury. Low odds, but the squiggly line still proved that some transmission was in fact occurring.
The armor of the vehicle protected its occupants against shrapnel, and against almost all of the blast. But like Cooper’s body armor, some pressure still got through, and then, like the blasts inside Johnson’s enclosed room, the penetrating waves reflected around the inside of the vehicle to increase the chances of trauma.
The day after my adviser, Dale, wandered into my office to declare, “What about the Hunley?” he and I looked at the drawings of the massive beer keg of black powder, the photos of the spar with its cartoonishly peeled-back copper shards, and the stationary skeletons seated at their battle stations. We knew about the hallmarks of blast trauma, about the body armor of Cooper, the case reports of World War II, the small rooms blasted by Johnson, the armored-vehicle transmission data published by Champion. We suspected that—for the first time in our field—the decades of experiments by our scientific ancestors had all intersected to explain a case of fatal trauma that had occurred inside a protective metal structure.
The original plan had been to build a computational model using the US Navy’s proprietary software called DYSMAS, the program I had already been using to gauge the blast exposure levels of World War II soldiers swimming in the ocean. However, the black powder was proving difficult and fickle. The scientific literature about black powder was all over the place, describing performances that varied wildly based on the powder’s composition, its confinement and exact setup, and a million other variables. Unlike standardized high explosives like TNT and C4, which had well-known and polite equations to describe their behavior in every situation, the physics to describe black powder seemed to make most blast scientists simply give up in frustration.
We finally decided the computer model was a no-go. The black powder introduced too many variables. With its slow burn rate, the results produced by any computer model could change dramatically based on the construction of the theoretical charge casing. And after the powder exploded, the casing burst, and the bomb released its pressure into the water, black powder’s unusually slow rise time to peak pressure might have strange and unpredictable effects on the wall of the submarine that could be difficult to model accurately.
The DYSMAS code wasn’t designed to model propagation through surfaces. It was designed to examine the walls of a structure, see what they would be exposed to, and determine whether they would fail. Modeling the trampoline-like way they could flex and push a secondary whoosh of air into the inside of the submarine was something that had never been done with that model, and had never been tested to work.
To test our theory, we would need to do live experiments. Or, in the words of Dale . . .
“Build a model of the submarine, Rachel, and blast it.”
During the Civil War, engineers conducted their own explosive experiments with black powder in order to gradually evolve the technology. For my own tests I would be able to lean upon decades of work by other researchers, but in the 1860s the largely undeveloped state of the field meant that novice blast experts in all areas were eager to develop the first reliable underwater mines, and many of them lost their lives and body parts in the process.
One ambitious such inventor was engineer Francis Kemp. He thought his experimental “water rocket” might come in handy against the Union forces prowling Lake Pontchartrain, menacing his Confederate home city of New Orleans. The recent secession had created a need for defenses to guard the port city, and like the citizens building fish boats, he too wanted to weaponize the water. The torpedo was his rudimentary experiment in blast science, his attempt to optimize an explosive technology he thought could help. The shore he had selected for the test was near Bayou St. John, the site of the tests of the submarines.
As Kemp prepared the charge, the sensitive black powder somehow accidentally received the minor jolt it needed to start the chemical reaction. Bent over to work on the device, Kemp’s face and head absorbed the full force of the blast, including shrapnel from the casing. He was expected to die, with “part of his head shattered to pieces,” but after a long month he mostly recovered . . . minus the loss of one eye.
Demure Chicago lady Mrs. E. H. Baker was hoping to observe a more successful one of these underwater blast tests. She picked up her skirts and gingerly stepped down from the carriage. She paused to wait for her friend to join her, then together the two women sauntered with their picnic basket toward the festive crowd gathered along the shore of the James River, south of Richmond, Virginia. Mrs. Baker’s dear childhood friend happened to be married to a high-ranking Confederate officer named Atwater, and he had invited the ladies to join him for the day.
As they approached the jovial crowd of Confederate officers and their accompanying women, Mrs. Baker noticed a dilapidated old scow bobbing silently in the middle of the river: the target of the day’s test. She waited with apprehension while an innocent-looking green float bobbed nonchalantly toward the unsuspecting target. The float paused, then gradually reversed its direction and retreated. The crowd waited in anticipation.
A massive gush of water hurtled the scow into the air. The scow burst upward while its hull exploded like fireworks. Mrs. Baker felt the ground tremble and heard the muffled rumble of the underwater blast. The crowd roared in delight as tiny airborne fragments of boat plunked slowly down into the river, and Atwater could not contain his elated surprise at the destructive power of the new weapon.
Mrs. Baker, in contrast, was doing her best to hide her sorrow. The green float was attached to a small model of a submarine that was being prepared for use in waters farther south. The personnel within the submarine had destroyed the boat by attaching a torpedo, retreating to a safe distance, and triggering the charge. The test was a display to show the officers on the shore how the full-sized vessel would strike terror into the Union sailors and help break the blockades. Baker understood the devastation such a submarine could wreak on Union men.
She gleaned as much information as she could from Atwater, wrote it down, stayed with him and her friend a few more days to be polite, and then returned to Chicago. There, she promptly delivered her notes to her boss: famed spymaster Allan Pinkerton. Visiting her childhood friend had been a ruse. Pinkerton, an ardent abolitionist and supporter of the Union cause, sent the information about the submarine directly to Union general George McClellan.
The North knew that the South was building submarines, and they soon began guarding their warships against such attacks. It seems plausible that at least some of the warnings they received were from Allan Pinkerton and his network of spies. The major problem with Mrs. Baker’s story, its fatal flaw, is that there is no other evidence that Mrs. Baker existed. Dozens of modern books retell her story, but they all trace back to the single source of Allan Pinkerton’s book Spy of the Rebellion. Baker and her husband do not appear in any US Census records for the state of Illinois. She is not in any of Allan Pinkerton’s corporate records, except for references to the story in Spy. She does not appear in his private listings of the spies he employed during the Civil War, and she does not appear in any of the extensive personal notebooks he kept cataloging all his known spies, employees, criminal contacts, and informants. None of his known female spies used the alias Baker.
Pinkerton writes affectionately about Atwater. The text of Spy contains several paragraphs about how Atwater never really wanted to be part of the secession, how he despised slavery, and how he secretly hoped the North would win the war. From Pinkerton, a vocal and vehement abolitionist, the words are unexpected and exaggeratedly heavy praise that read like an attempt to exonerate a Confederate officer . . . possibly because the officer had been sending him information about submarine tests in secret.
Pinkerton was famously unwilling to release the names of his operatives and informants. At one point the US government insisted that he would only be paid once he submitted the spies’ complete names, but Pinkerton still refused.
“The names of any employees (or operatives, as I usually style them) should only be known to myself,” he wrote. He explained further, “Already some of my Force have suffered death at the hands of the Rebels.”
Pinkerton’s detailed billing documents, which list spies by the initials of their real names, show no trace of Baker. Unfortunately, the detailed billing documents that Pinkerton submitted to the US government for his Civil War espionage have only been preserved starting with the spring after the Baker story was said to have taken place. However, they show no trace of a trip to Richmond. The spy who incurred each charge is listed by their initials, and Baker’s initials are not listed for any other missions.
Only two Atwaters seem to have reached the rank of captain in the Confederate Army, and only one of them would have been an officer of sufficient rank and near Richmond at the time of the test, which was the fall of 1861. His full name was James W. Atwater, and he seems to be the Captain Atwater of Pinkerton’s story.
Atwater was married, but his wife died just before Pinkerton claims that Mrs. E. H. Baker traveled to visit her. Capt. James W. Atwater was also “a man of strong convictions” who “hated all manner of shams,” so much so that the phrases made it into his obituary. Given that the submarines were considered “infernal,” it is possible Atwater considered them a “sham” as well.
Pinkerton believed in protecting his sources above all else. Some of Pinkerton’s information about the submarine is accurate. It is possible Pinkerton fabricated the entire story, using bits and pieces from the newspaper articles about the 1864 Hunley attack to fill in the details to publish his book in 1883, but none of these early publications contain details about the loop of wire or the number of men in the prototype vessels.
It wasn’t long before the experimental boats and spar torpedoes were taken into action. After a failed early attempt, a Confederate officer, Lt. William Glassell, found himself glowering at a Union marshal with an angry smirk on his face. He had been fished out of the water and arrested. Now he sat in prison, silent, wearing head-to-toe gray to show his resilient Confederate pride, and refusing to answer any further questions.
He had been the leader of a tiny crew of four, wedged inside a long cigar-shaped boat of a class that was starting to be referred to as “Davids.” The small boats, stealthy, sneaky, infernal, were designed to devastate the Goliath Union warships prowling outside the Rebel harbors. His boat was made of wood, just under 2 meters in diameter, a rough cylinder with pointy tapered ends. This boat, also named David, couldn’t submerge—a fraction of the wooden hull always stuck up above the waterline, the rest below, with a massive smokestack jutting upward to provide air to fuel the hungry engines inside.
Glassell and his crew had maneuvered their David close to the USS New Ironsides that night late in 1863, close enough to stab it with the 70-pound black powder torpedo on their spar. The big ship somehow survived, even though it would need to limp back to safe harbor for repairs. The David, however, was not so lucky; the frothy plume of seawater from her own blast had swamped her smokestack, partially flooded her interior, and extinguished the fires of the engines needed to propel the little boat home. She was stuck adrift, a sitting duck waiting to be hit by the small-arms fire of the crew of the New Ironsides.
Glassell and two of his crew had abandoned ship, preferring to take their chances in the open water rather than risk drowning in the partially flooded, narrow wooden tube. The fourth man, unable to swim, stayed put. Glassell and one other crewman were plucked out of the water by the sailors of the New Ironsides, put into shackles, and taken north for interrogation. The other two David crew escaped, got the engines restarted, and made it safely back to shore.
Glassell told his story to the US Navy interrogators in full, then staunchly refused to answer more questions. He sat in prison, spurning any clothing that wasn’t Confederate gray, waiting to hear the Union government’s decision regarding what would be his “final disposition.”
The almost-victorious David proved to the Confederate officers in Charleston that a combat victory by a little cigar-shaped boat was possible. More stealth was needed, they concluded, from a boat that could fully submerge. A boat like the Hunley.
The Hunley herself had been tested in the rivers outside Mobile, Alabama, as Gen. Franklin Buchanan watched, and as other experimenters were testing their torpedoes and boats in Virginia, Louisiana, and other locations. The Hunley had been using a torpedo towed on a line at the time, but these tests quickly revealed that not only could the little submarines sink large ships with their torpedoes but also that the placement and configuration of the torpedo was key. Placing the torpedo beneath, rather than beside, the target increased the amount of damage it would do to the enemy vessel. Torpedoes that exploded above the waterline could cause some destruction, but much of the force would vent harmlessly out into the open air. When the torpedoes were triggered underwater and beneath the ships, the bombs would bite more massive chunks out of the hulls. The David had a spar that could be angled downward slightly but would have been more difficult to fit completely beneath an enemy hull, so the blast was less destructive than it could have been.
The science of the 1800s could not explain this phenomenon, but modern physics has mathematically solved the problem. The explosion of a bomb underwater causes a bubble of gas to form and burst outward, made up of the gaseous products released by the chemical reaction. As the bubble collapses, a rapid jet of water follows it and smashes against the hull of the target ship. The combined forces of both the pressure wave and the water jet wreak havoc on the enemy structure. The bubble then travels upward toward the surface of the water, repeatedly collapsing and expanding as it moves. Submarines in World War II would later learn the destructive power of these pulsating gas bubbles, as boats unfortunate enough to pass directly over mines were easily shredded by their motion.
Positioning a torpedo beneath a ship combined the destruction of the pressure waveform with the destruction of the gas bubble. The experimenters of the Confederacy learned the technique through trial and error.
The high density of the water also means that the blast pressures can propagate away from the bomb and toward other objects in the water, like a waiting submarine, more efficiently and with less loss. If a target is 5 meters away, a blast that occurs underwater transmits pressure levels that are more than twenty times higher to that target than if the same blast had occurred in air.
Objects near the surface are less vulnerable, shielded from the full potential of the blast by their proximity to the air. Deeper objects, positioned farther down in the water and therefore farther away from the protective effect of the surface, will receive the full, unmitigated wrath of the blast.
Hunley, McClintock, and Watson tested their boat by setting off a torpedo that was a full 400 yards away in one of the rivers of Mobile. Kemp’s face felt the wrath of his much smaller bomb, but in the air where the pressure waves were less destructive. “Mrs. Baker’s” submarine also used a long line to pull the trigger, waiting to set off the charge until the submarine was much farther away. Glassell and his David crew used a smaller bomb than the rest of the boats, a bomb that could not submerge as fully. The men of the David were also positioned inside a vessel made of wood with a thick hull that sat farther out of the water, and was therefore protected by proximity to the surface.
The torpedo of the Hunley carried a much larger charge than the almost-successful David. Her spar sank farther down into the water, with a bend at the end to allow the black powder torpedo to be positioned completely beneath the hull of the Housatonic. The Hunley’s thin hull was metal, a material more eager to transmit pressure waves than wood. She was sunk farther down into the water, nestling her into a cocoon of the ocean, a cocoon that surrounded her in a saltwater medium that could more efficiently transmit the full force of any blast.
I explained all of these design features out loud as I sat in a conference room, patiently clicking through one slide at a time during my preliminary exam. With each slide, the members of my advising committee riddled me with a gauntlet of scientific questions about blasts, injuries, and physics. The five men—four professors from Duke, including my adviser, and one US Navy PhD underwater scientist from my base in Florida—were the panel of judges who would one day determine whether I would be deemed worthy of graduating with my PhD. To get my degree I would have to be proclaimed “done” by this committee. The preliminary exam was a rite of passage to determine whether or not I was worthy to continue my work and use it to aim for graduation. Fail the exam twice, and you are asked to leave Duke.
So much of science comes down to measuring the correct squiggly line on a computer screen. Without the benediction of this committee, I would not get the chance to move forward with this project and experimentally measure the squiggly line of pressure inside the boat. And without that measurement, my theory would have to stay an opinion.