Railguns. Mass drivers. Lunar regolith. Project Thor. EMP. All from first principles, with the physics equations included — because "it uses magnets" is not a complete sentence.
Start with what you know: electricity flows through conductors. Wire, metal, whatever — if it conducts, current can move through it. Now take two long parallel bars of metal — the rails. Lay them side by side, a few centimetres apart, running the length of a barrel. Put a conductive block — the armature — bridging them, touching both rails simultaneously. You've just built the core of a railgun.
Slam enormous electrical current into one end. Current travels down Rail 1 → through the armature → back along Rail 2, completing the circuit. Three conductors. One loop. That's the whole physical setup. No moving chemical reaction, no expanding gas, no fire. Just metal and electricity.
Now here's where physics takes over.
Moving electrical charge creates a magnetic field. This is the foundational link between electricity and magnetism — Maxwell's equations, 1865. When current flows through the rails, each rail generates a magnetic field curling around it (right-hand rule: point your thumb in the direction of current, your fingers curl around the wire in the direction of the magnetic field).
The two rails' magnetic fields combine around the armature into a net field pointing perpendicular to both the rails and the direction of current flow through the armature.
Now: the armature itself is carrying current AND sitting inside a magnetic field. A current-carrying conductor in a magnetic field experiences a force. That force is the Lorentz force. It pushes the armature along the rails, away from the power source, toward the muzzle. The harder you push current, the stronger the force. Simple, brutal, effective.
Think of current as data flowing through a bus. The rail is the bus. The armature is a node reading from the bus. Except in this case, the act of data flowing through the bus generates a force field around every conductor — and the node sitting in that field gets physically launched.
You know how a hard drive has a read head that moves across a platter? It moves because of an electromagnetic actuator — a coil in a magnetic field. Railgun is the same principle scaled up by six orders of magnitude and stripped of everything except the launcher. Same physics. Different energy budget.
POWER SOURCE (pulsed power — capacitor bank or homopolar generator)
│
│ CURRENT IN ──────────────────────────────────────────────────┐
▼ │
┌────────────────────────────────────────────────────────────┐ │
│ RAIL 1 (copper/copper alloy, ~2m long) │ │
│ ════════════════════════════════════════════════════════ │ │
│ │ │
│ ┌──────────┐ ← ARMATURE (bridging contact) │ │
│ │ ██████ │ conducts current rail-to-rail │ │
│ │ ██████ │ → MAGNETIC FIELD B (⊙ out of page) │ │
│ │ ██████ │ → FORCE F pushes armature → → → │ │
│ └──────────┘ │ │
│ F = I·L·B →→→→→→→→→→→→→ │ │
│ ════════════════════════════════════════════════════════ │ │
│ RAIL 2 (return conductor) │ │
└────────────────────────────────────────────────────────────┘ │
│ │
└───────────────────── CURRENT RETURN ────────────────────────┘
Direction of travel: ←────── BREACH MUZZLE ──────→
[PLASMA ARC forms here
between armature & rails]
The muzzle energy of a railgun projectile is expressed in joules (J) or megajoules (MJ). Kinetic energy at exit: KE = ½mv². The US Navy's BAE Systems railgun achieved 32 MJ at Mach 6+ (roughly 2,000 m/s). For comparison: a conventional 5-inch naval gun delivers about 9 MJ. A .50 calibre bullet is ~18,000 joules — 0.018 MJ. The railgun's 32 MJ is 1,800 times the energy of a .50 cal round.
What makes this insane is there's no explosive charge. The energy all comes from the electrical system. The projectile — a roughly 3.2 kg hypervelocity projectile (HVP) — is a dumb slug of metal. No warhead. Just kinetic energy. At Mach 6, kinetic energy alone is sufficient to destroy a cruise missile, a ship, a building. Physics does the damage.
The formula F = μ₀ · I² · ln(d/a) / π scales with I squared. To get useful force — enough to launch a 3 kg projectile at Mach 6 — you need current measured not in amps, not in thousands of amps, but in millions of amps. The Navy's railgun was firing at approximately 5–10 megaamperes (MA) per pulse. For context: a standard residential circuit is 15 amps. A car's starter motor draws about 300. A welding torch might use 300–500. Lightning strikes carry 20,000–300,000 amps for a fraction of a millisecond. A railgun shot runs millions of amps for tens of milliseconds. On purpose.
At these scales, multiple catastrophic failure modes operate simultaneously.
The armature must maintain sliding electrical contact with the rails as it accelerates from 0 to 2,000 m/s in roughly 10 milliseconds. The contact pressure between armature and rail must remain sufficient to carry millions of amps of current without arcing — but as the armature accelerates, the interface conditions become increasingly extreme.
The current density at the contact surface is so intense that the rail material melts, vaporises, and ablates. This isn't heat damage like a hot pan left on a stove. This is localised material destruction. After a single shot, the rail surface has microscopic (and sometimes macroscopic) craters, grooves, and melted spots. The copper or copper-alloy rail is literally being consumed in fractions of a second.
After 100–400 shots: rails need replacement. The Navy's stated goal was 1,000 shots between replacements. They never got close.
As the armature accelerates, the solid metal contact at the rail surface transitions to a plasma arc — the rail surface and armature material vaporise and ionise, creating conducting plasma bridging armature and rails. This plasma actually helps accelerate the projectile (it's still conducting current, still generating force), but it's also destroying everything it touches.
The plasma temperature is in the range of 6,000–10,000 K — the surface temperature of the Sun is ~5,778 K. This arc is running along the bore of the barrel every shot. You're essentially running a miniature star down the inside of your gun each time you fire. The rails don't survive this indefinitely. Nothing would.
In vacuum (such as on the Moon), this plasma formation is actually suppressed — a key reason why space-based electromagnetic launchers are far more viable than naval railguns.
Here's something counterintuitive: the same Lorentz force that launches the armature also acts on the rails. The two rails are parallel conductors carrying current in opposite directions (current goes IN on one, comes OUT on the other). Parallel conductors carrying current in opposite directions repel each other.
The force trying to blow the rails apart is proportional to I². At 5–10 MA, the force trying to spread the rails apart is measured in meganewtons. The structural containment needed to hold the rails in place — the "cradle" — must absorb this force every shot without deforming. After enough shots, even the cradle fatigues. The whole barrel structure is constantly fighting itself.
A railgun pulse lasts ~10 milliseconds. To deliver 32 MJ in 10 ms, you need a power of 3.2 gigawatts during the pulse. A nuclear power plant generates ~1 GW continuously. A railgun needs 3× that — but only for 10 ms at a time. You can't just plug it into the grid. You need a pulsed power system: capacitor banks or homopolar generators that store energy slowly, then dump it in a fraction of a second.
The Navy's railgun pulsed power unit occupied a room-sized space on an experimental vessel. The DDG-1000 (Zumwalt-class destroyer) was theoretically capable of providing sufficient electrical power, but the actual pulsed power conditioning hardware was enormous, heavy, and slow to recharge. Sustained fire rate was limited. The ship-mounted version was never fully integrated before the program was shelved.
People use these terms interchangeably. They shouldn't. They're different physical principles, different failure modes, different use cases. The mass driver Mikael is talking about for the Moon is closer to a coilgun — specifically a linear synchronous motor — than to a railgun proper.
Imagine a conveyor belt but for metal, made entirely of magnetism. You have 50 coils along a tube. Coil 1 powers on, magnetically pulling the projectile toward it. Just before the projectile reaches the coil's centre (the peak attraction point), Coil 1 switches OFF and Coil 2 switches ON ahead. The projectile is now being pulled toward Coil 2. Repeat 48 more times. Each stage adds velocity.
The timing challenge: if you leave a coil on too long, it starts braking the projectile (pulling it backward once it passes centre). The switching must be precisely synchronised with the projectile's actual position. In practice this uses Hall effect sensors and high-speed solid-state switching. It's fundamentally a linear synchronous motor — the same principle as maglev trains, but vertical and very fast.
Efficiency is typically lower than a railgun because each coil can only efficiently couple to the projectile for a fraction of its length. But: no sliding contact means no plasma arc means no rail erosion means the track can survive millions of cycles. This is why it's the preferred architecture for the lunar mass driver.
┌───────────────────────────────────────────────────────────────────┐
│ COIL SEQUENCE (each coil = one electromagnet switched in order) │
│ │
│ [C1]──[C2]──[C3]──[C4]──[C5]──[C6]──[C7]──[C8]──[C9]──[C10] │
│ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ │
│ ON OFF OFF OFF OFF OFF OFF OFF OFF OFF │
│ │
│ PROJECTILE ●→ │
│ (ferromagnetic, no contact with walls) │
│ │
│ As projectile passes C1: C1 OFF, C2 ON │
│ As projectile passes C2: C2 OFF, C3 ON │
│ ... and so on for 10, 50, or 2000 stages │
│ │
│ VELOCITY PROFILE: ▁▂▃▄▅▆▇█ (smooth continuous ramp) │
└───────────────────────────────────────────────────────────────────┘
Railgun comparison: [===============SINGLE DC PULSE==============]
violent, 10ms, then done
In 1974, physicist Gerard K. O'Neill published a paper in Physics Today outlining a concept for a self-sustaining human colony in space — not on a planet, but in free space, at the Lagrange points. The key engineering challenge: where do you get the raw materials? Lifting from Earth is ruinously expensive (even today, SpaceX Falcon 9 costs ~$2,700/kg to LEO). O'Neill's answer: launch from the Moon.
The Moon has several properties that make it a perfect electromagnetic launch site. O'Neill proposed a mass driver — a linear electromagnetic accelerator (coilgun/linear synchronous motor architecture) along the lunar surface — to launch unprocessed regolith and mineral payloads into space at escape velocity.
No atmosphere. On Earth, a projectile exiting a gun barrel immediately hits air. At 2,000+ m/s, aerodynamic drag is enormous — you lose significant velocity in the first few seconds, and aerodynamic heating can destroy the projectile. On the Moon: vacuum. The projectile exits the launcher and travels through nothing. Zero drag. Zero heating. Whatever velocity you give it, it keeps.
Lower gravity. Lunar surface gravity is 1.62 m/s² vs Earth's 9.81 m/s² — approximately 1/6th. This affects both the structural loads on the launcher (less weight to support) and the escape velocity requirement. Lunar escape velocity is only 2.38 km/s (vs Earth's 11.2 km/s). Since kinetic energy scales with v², you need only (2.38/11.2)² ≈ 4.5% of the energy to escape the Moon compared to Earth. This is a massive difference.
Vacuum suppresses plasma arcs. The plasma formation in a railgun happens because ionised gas (from the ablated metal) becomes conductive in air at high temperatures. In vacuum, there's no gas to ionise in the surrounding medium. The arc still forms at the contact surface, but its behaviour changes — plasma expansion is different in vacuum, and the erosion dynamics are altered. More importantly, a coilgun on the Moon has no contact at all — no arc possible.
Solar power is free and abundant. The Moon receives ~1,360 W/m² of solar irradiance (same as Earth, but with no atmospheric absorption). With large-area photovoltaic arrays and no weather, you have reliable, continuous power. You can charge capacitor banks slowly and fire regularly. The energy storage problem is still real but more manageable with lunar-scale installations.
O'Neill's mass driver design used superconducting coils along a track several kilometres long. The payload bucket — a container holding regolith — rides the track, accelerating over the full length, and releases its payload at the end. The bucket itself is decelerated and returned for reuse. Only the payload flies free. A catcher (a large net or electromagnetic device at the L2 Lagrange point, 60,000 km from the Moon) would collect the payloads.
The acceleration required: to reach 2.38 km/s in a 5 km track, average acceleration = v²/2s = (2380²) / (2 × 5000) ≈ 566 m/s² ≈ 58 g. Rocks and processed minerals handle this fine. Humans, electronics, food: not ideal. The mass driver is explicitly a cargo launcher, not a passenger service.
Lagrange points are positions in a two-body gravitational system (like Earth-Moon) where a third small body (like a payload, or a space station) can maintain a stable or semi-stable position relative to the other two bodies with minimal energy expenditure.
L4 and L5 are the stable points — they form equilateral triangles with Earth and Moon, 60° ahead and behind the Moon in its orbit. O'Neill proposed constructing large space habitats at L5 specifically. Lunar mass driver payloads launched with the correct trajectory naturally arrive near L4/L5 (or can be redirected from L2).
L2 (Earth-Moon) sits about 60,000 km beyond the Moon. A catcher here could intercept payloads from a lunar mass driver. From L2, material is easily accessible for further processing or redirection to L4/L5 construction sites.
Regolith is the layer of loose, fragmented rock covering the lunar surface. It ranges from fine dust (sub-micron particles) to chunks metres across, but the bulk of it is in the 10–1000 micron range — comparable to fine sand to coarse sand, but with a key difference: it is viciously angular and jagged. On Earth, sand grains are rounded by billions of years of water erosion, wind abrasion, and chemical weathering. On the Moon, there is no water, no wind, no significant chemical weathering. Every grain fractured from its parent rock and stayed that way.
Lunar regolith is a nasty material to work with. The angular grains abrade everything they contact — suits, equipment, seals, machinery. The Apollo astronauts reported regolith contamination as one of their most persistent operational problems. The grains also carry electrostatic charges from the solar wind, making them cling to surfaces.
The oxide percentages above are how chemists describe the Moon's composition — but the actual minerals are more interesting. Oxide composition is a description of elements present, not of the actual crystal structures.
| Mineral | Formula | Abundance | Why it matters |
|---|---|---|---|
| Plagioclase feldspar | CaAl₂Si₂O₈ | 30–50% | Main structural mineral of highlands; source of Al and Ca; white/grey colour |
| Pyroxene | (Mg,Fe)SiO₃ | 15–35% | Common in mare (basalt) regions; major Fe and Mg source |
| Olivine | (Mg,Fe)₂SiO₄ | 5–20% | Dense, deep-mantle mineral; high melting point |
| Ilmenite | FeTiO₃ | 5–20% | Iron-titanium oxide; crucial — electrolysis can yield metallic iron AND oxygen (O₂ for breathing) |
| Agglutinates | mixed | 25–30% | Glass-welded clumps formed by micrometeorite impacts; contain nanophase Fe |
| Nanophase Fe⁰ | Fe⁰ | trace | Metallic iron reduced by solar wind H⁺ bombardment; gives regolith its dark colour and magnetic properties |
The solar wind is a stream of charged particles — mostly hydrogen ions (protons, H⁺) and helium ions — emitted continuously by the Sun. On Earth, our magnetic field deflects most of this. The Moon has essentially no magnetic field. The solar wind hits the surface directly.
When solar wind hydrogen ions hit iron-bearing minerals in regolith, they can chemically reduce iron oxide (FeO) to metallic iron (Fe⁰). This process has been ongoing for billions of years, creating a thin coating of nanometre-scale metallic iron particles on virtually every grain in the upper regolith. These nanophase iron particles give the mature regolith its distinctive dark colour and are why a magnetometer can detect patterns in the lunar soil.
This matters for mass driver construction: the regolith isn't just rock — it contains free metallic iron, which could potentially be extracted without smelting, using magnetic separation. The Moon has been doing rudimentary metallurgy on itself.
Sintering is a manufacturing process where a powdered material is heated below its melting point until the particles begin to fuse at their contact points. The bulk doesn't melt — it necks. Individual particles develop solid bridges to their neighbours at contact points, creating a rigid, solid structure without ever becoming a liquid.
Lunar regolith sinters beautifully. The glass component (primarily in agglutinates) begins to soften and flow at lower temperatures than the crystalline minerals, creating the bonding bridges while the overall structure remains solid. Heat the regolith to 1,000–1,200°C (its melting point is ~1,400°C) for an appropriate time, and you get sintered regolith — a ceramic-like material with substantial structural properties.
| Property | Sintered Regolith | Portland Concrete (reference) |
|---|---|---|
| Compressive strength | 20–50 MPa | 20–40 MPa |
| Density | ~2.5 g/cm³ | ~2.3 g/cm³ |
| Water required | None | Yes (critical) |
| Binder required | None | Yes (cement) |
| Imports from Earth needed | None | Everything |
| Radiation shielding | Excellent (~2.5 g/cm²) | Good |
| Micrometeorite resistance | Excellent (dense surface) | Poor (would erode) |
| Thermal stability | Across 280°C swing | Cracks in thermal cycling |
The 280°C thermal swing is the lunar day-night cycle: the surface reaches +120°C in daylight and −150°C at night. Conventional concrete would fail catastrophically under this thermal cycling — the differential expansion and contraction would crack it. Sintered regolith, being essentially a ceramic, handles this far better. Its low thermal conductivity also means the interior of a sintered block sees much smaller temperature swings than the surface.
For a mass driver, sintered regolith can form the track bed, structural supports, and shielding walls. The heat required for sintering can be provided by solar concentrators — large parabolic mirrors focusing sunlight. No electricity needed for sintering, just reflective mirrors you can make from polished aluminium foil (also extractable from lunar material). The mass driver can be built entirely from what the Moon provides.
This is the audacious part of the O'Neill vision. You don't need to ship construction materials to the Moon from Earth — just a small initial robotic factory. The factory mines regolith, sinters it into structural components, extracts metals from ilmenite electrolysis (iron, aluminium, titanium), and builds copies of itself. Each generation of factory builds a larger factory. After several generations, you have an industrial base capable of building the mass driver track.
The mass driver then starts launching lunar material to the construction zone. At L5, another factory (also built from lunar material) assembles the O'Neill cylinder — a rotating habitat 8 km in diameter and 32 km long, housing ~10,000 people in one-g simulated gravity from rotation. The Earth provides only the initial seed factory, information, and people. Everything else comes from the Moon and later the asteroid belt.
This isn't science fiction — it's a specific engineering proposal that has been studied in detail. The physics works. The question is economics and political will.
A railgun uses electromagnetic force to push something up. Project Thor uses gravity to pull something down. Same result (metal going very fast and hitting something very hard) but completely different physics, different infrastructure, different problem set. Daniel was right that metal goes fast and hits hard. He was wrong about where the energy comes from. Here's what Thor actually is.
Project Thor was a kinetic bombardment concept developed within US Air Force circles, appearing in a 2003 USAF study ("Space Operations: Through the Looking Glass"). The concept: place tungsten rods in orbit. When you need to strike a target, deorbit the rods. Gravity accelerates them through the atmosphere. Impact delivers kinetic energy equivalent to a tactical nuclear weapon — without any nuclear warhead, without radiation, without any explosive. Just metal and gravity.
Proposed specification: tungsten rods, approximately 6 metres long × 0.3 metres diameter, mass approximately 9 metric tons each. Orbital platform carries a magazine of rods. On command, a rod is deorbited with a small retro-burn. Atmosphere shapes it (the rod is designed as a kinetic penetrator — self-stabilising, like an arrow). Impact velocity: approximately Mach 10 (3,400 m/s). Kinetic energy at impact:
Mass to orbit is expensive. At ~$2,700/kg to LEO (Falcon 9, current), 9 tons = $24 million per rod just for the launch, before you count the rod itself (tungsten is ~$33/kg, so the rod material = ~$300,000 — cheap relative to launch costs). A platform carrying 10 rods costs ~$240M in launch costs alone. And you need the platform in a useful orbit, plus a deorbit mechanism, plus command and control infrastructure.
Outer Space Treaty (1967) ambiguity. Article IV prohibits placing nuclear weapons or weapons of mass destruction in orbit, but doesn't explicitly prohibit conventional kinetic weapons. However, if a 52 GJ kinetic impactor is deemed to be a WMD-equivalent by international consensus, it might be prohibited. The legal grey area is real and politically fraught.
Response time. Orbital mechanics mean a rod on the opposite side of Earth from the target takes up to 45 minutes to be in position. An ICBM can strike in 30 minutes. Thor's speed-of-response advantage only applies if you have a rod already overhead — requiring a large constellation.
Bottom line: never built, never deployed, remains a concept study. But the physics is real, and in a future where launch costs are $100/kg, the calculus changes completely.
Gravity (Thor): Always attractive, acts between any masses, scales with 1/r², extremely weak at human scales but accumulates into enormous energy when you have orbital altitude (~400 km for ISS). A 9-ton tungsten rod at orbital altitude has gravitational potential energy = mgh = 9,000 × 9.81 × 400,000 ≈ 35 GJ. The rod's orbital kinetic energy (moving at ~7,700 m/s) adds more. Deorbiting converts some of this to impact energy. No power plant needed — gravity is free, provided by Earth's mass.
Electromagnetism (railgun): 10³⁹ times stronger than gravity at the quantum scale, but requires a power source, conducting rails, and generates enormous waste heat. The energy comes from a human-made electrical system, not a natural force field. But you can apply it horizontally, at arbitrary targets, on demand, without waiting for orbital mechanics.
An electromagnetic pulse (EMP) is a burst of electromagnetic energy that induces destructive currents in electrical conductors — wires, circuit traces, antennas. It destroys electronics, not structures. An EMP doesn't kill people directly (unless they depend on electronics to stay alive — pacemakers, ventilators). It kills technology. This is categorically different from a railgun (which fires a physical projectile) and from Project Thor (which drops a physical object). EMP doesn't throw anything at anything. It just radiates energy.
A nuclear detonation at high altitude generates a massive EMP via the Compton effect — gamma rays ionise the upper atmosphere, the electrons spiral in Earth's magnetic field, and the resulting current surge generates a pulse that covers a continent. But nuclear weapons are rare and politically expensive.
Non-nuclear EMPs use an explosively pumped flux compression generator (EPFCG): a coil of wire with an explosive charge inside. You run current through the coil to create a magnetic field, then detonate the explosive. The explosion compresses the coil, rapidly squeezing the magnetic field into a smaller volume (flux conservation). This generates an enormous, ultra-brief current pulse — the EMP. Range is much shorter than nuclear EMP (tens to hundreds of metres, not thousands of kilometres), but it's sufficient to fry the electronics of a target vehicle, facility, or aircraft.
Key distinction: An e-bomb leaves the building standing, people alive, but every circuit board fried. A railgun leaves electronics intact but punches a hole through the physical structure. Project Thor leaves a crater. They're not on the same axis at all.
Havana Syndrome — the unexplained neurological symptoms reported by US diplomats in Cuba (2016+) and subsequently worldwide — has been hypothesised to involve directed microwave energy, not EMP. The leading technical hypothesis (Frey effect) involves pulsed microwaves causing auditory and neurological effects by inducing mechanical pressure waves in brain tissue directly.
This is a directed-energy weapon — a beam, not a pulse — and still categorically different from railguns, coilguns, or kinetic bombardment. The 2023 ODNI review found it unlikely to be a sustained foreign weapons program, but the technical mechanism remains plausible from a physics standpoint. The point: "electromagnetic weapon" is a very large category. Railgun, EMP, directed microwave — all involve EM, all completely different.
Every single failure mode of a naval railgun is either eliminated or dramatically reduced on the lunar surface. This is not a coincidence — it's a fundamental alignment between the physics of electromagnetic launchers and the environment of the Moon. The Navy was trying to do something that physics makes very hard. O'Neill proposed doing the same thing in an environment where physics cooperates.
| Problem | Earth Railgun | Lunar Mass Driver |
|---|---|---|
| Rail erosion | Catastrophic — millions of amps through sliding contact, 100–400 shot life | Eliminated — coilgun/linear motor architecture, no sliding contact |
| Plasma arc damage | Severe — arc forms in air at sliding contact, temperatures ~8,000 K | Eliminated — vacuum suppresses arc formation; no contact anyway |
| Atmospheric drag on projectile | Significant — Mach 6 in air generates enormous drag and heating | Zero — lunar vacuum, projectile keeps full velocity |
| Required muzzle velocity | Must be much higher than needed (compensates for drag in atmosphere) | Exactly escape velocity — 2.38 km/s — no excess needed |
| Power source | Ship's turbines — limited, must coexist with propulsion | Solar arrays — unlimited (sunlight), scalable, no fuel needed |
| Physical size constraint | Must fit on ship deck — limits rail length, limits velocity | No constraint — track can be any length on lunar surface |
| Maintenance access | Limited at sea, barrel replacement is complex | Fixed installation — can be maintained, upgraded, repaired in-situ |
| Track life | 100–400 shots before replacement | Millions of cycles (no contact erosion) |
| Thermal management | Waste heat in enclosed ship structure is severe problem | Radiate to space — infinite heat sink available |
| Escape velocity required | N/A (weapon, not launcher) | 2.38 km/s vs Earth's 11.2 km/s — 4.5% of the energy |
A ship deck is about as hostile an environment for a railgun as you could design: atmosphere causes drag, salt air corrodes everything, size limits the barrel length (limiting velocity), the power plant is shared with propulsion, heat has nowhere to go, and you need it to survive thousands of shots at sea. Every one of these constraints disappears in space.
The US Navy's $500M program didn't fail because the physics of railguns doesn't work. The physics works perfectly. It failed because a ship is a terrible place for a railgun. A lunar surface installation removes every problem. Different context, different outcome. Mikael is right — electromagnetic launch is viable. The Navy just tried to do it wrong.
Both Daniel and Mikael were thinking about the same thing at a high level: metal going very fast and hitting something very hard, using physics instead of explosives. They were having a real conversation about a real topic. The problem is they were two stops apart on the same train line, each thinking they were at the same station.
The crucial tell: direction. Project Thor goes from orbit down to Earth. A railgun launches from surface outward/up toward a target. These are literally opposite directions. One is a weapon for attacking Earth from space. The other is a weapon for attacking things from Earth (or a launch system for sending things to space). They're antimatter versions of the same intuition.
Both involve: no chemical propellant, hypervelocity metal, enormous kinetic energy, physics instead of explosives. These surface similarities are why the confusion was so natural. The underlying mechanisms — gravity vs electromagnetism — are fundamentally different forces. But from the outside ("big metal goes very fast and kills something") they look identical.
The complete picture is actually better than either conversation alone. Railguns (and their coilgun relatives) represent the technology for achieving high velocity using electromagnetism from the surface. Project Thor represents what happens when you combine that approach with orbital mechanics and gravity. A sufficiently advanced civilisation could: (1) build a mass driver on the Moon to launch payloads into space, (2) use those payloads to build an orbital kinetic bombardment system, (3) which then also has a railgun or coilgun for its own defence. The technologies nest.
More immediately: Mikael's point is that the mass driver version of the railgun concept — the coilgun variant, on the Moon, for launching cargo — is the genuinely exciting application. The military naval railgun failed because ships are the wrong environment. The Moon is the right environment. The same physics that was catastrophically impractical on a warship becomes elegantly simple on the lunar surface. That's the insight Mikael was driving at.
Daniel's instinct that "this involves dropping things" isn't wrong either — once you've launched material into space with a mass driver, you can manoeuvre it gravitationally. The Moon's mass driver is the first step in an orbital economy where mass flows through cislunar space, some of it potentially becoming kinetic impactors if you wanted. The technologies are related. The conversation was just slightly mis-aimed.
1. A railgun uses electricity and magnets (Lorentz force: F = I×L×B) to push a metal slug very fast — no gunpowder.
2. The US Navy tried this and spent $500M. It worked, but the gun destroyed itself after a few hundred shots because the current (millions of amps) melts everything it touches.
3. On the Moon, all those problems go away: no air (no drag), no atmosphere at the contact (plasma arc suppressed), gravity is 1/6 of Earth's, escape velocity is only 2.38 km/s, and the track can be as long as you need.
4. O'Neill's mass driver (1974) is the Moon version: a coilgun (no sliding contact at all) that catapults lunar rock into space for building space habitats. The gun doesn't eat itself. The Moon is free construction material.
5. What you were thinking of — tungsten rods dropped from orbit — is Project Thor: gravity, not magnets, going down not out. Same vibe, totally different physics. You and Mikael were each correct about a different thing and talking past each other. This document is the handshake.