Why the Droids Walked Through: A Velocity-Selective Barrier Model of the Gungan Field-Generator Shield, Reconstructed from the Surviving Records of the Naboo Engagement of 32 BBY
Abstract
At the engagement on the grass plains of Naboo in 32 BBY, the field-generator shield of the Gungan host turned tank and artillery fire and blaster bolts, yet columns of battle droids walked through it at marching pace. We reconstruct the barrier from the copies of the occupation command's engagement log released to the Republic, which give 12 formation crossings (816 droids, 1.6 to 4.6 metres per standard second, m/s) and 41 rounds that ended at the field (38 to 780 m/s), together with 9 returns reporting bolts turned. No record gives the shield's mechanism, so we offer one of our own: a thin viscoelastic skin of plasma that yields to slow intrusion and stiffens under fast. Complete separation of the two groups fixes only an interval. The transition speed lies above 4.6 m/s (4.0 to 5.3 allowing 15% error in dead-reckoned speed) and below 38 m/s (33 to 44), a factor of about 8, and, counting each formation as one outcome, the data allow a gate as gentle as an exponent of 1.6. In the model, shifting the transition upward by a given factor thickens the skin by that factor and relaxes the required stiffness by its inverse. Slow lobbed ordnance, a stalled body inside the skin, a droid carrying its own field and the plasma supply are the weak points. Five tests would narrow the interval and could overturn the skin hypothesis.
1. Introduction
On the day of the engagement on the grass plains of Naboo, in 32 BBY, the Gungan host raised a dome of energy over its ranks, powered by generators that great beasts had carried into the field. Tank and artillery fire from the invading force struck the dome and was stopped, and the bolts of the escorting droids were turned aside. Then the first ranks of battle droids, marching at the pace of a column on a parade ground, walked through the surface of the dome and on toward the Gungan line.
The shield stayed up while they crossed it, and no entry in the log records the droids slowing at its surface.
This paper, written in 30 BBY, two standard years after the engagement, asks what kind of barrier admits a body walking at a couple of metres per standard second and refuses a shell. A deflector engineer's first instinct is to treat the dome as a failure, a hole in a field that ought to have held, a flaw to be explained away by a weak generator, a tired beast or a lapse in the supply, and not a property anyone could have chosen on purpose. We think the record points the other way, because a shield that passed slow droids and stopped fast ordnance did so consistently, over a long front, on the day in question, and a barrier that behaves like that has a design.
The published literature supplies pieces. Bench work on thin repulsor-field sheets has reported that their resistive stress depends on the rate of deformation (Brandreth, 31 BBY), and energy absorbed per unit area in layered deflector barriers has been measured in the same period (Barrowdene, 31 BBY). The personal emitters carried by some heavy droids have been surveyed for coverage and bolt rejection (Ashkelan, 30 BBY), and the order of battle at Naboo has been reconstructed from Republic and Naboo sources (Bastable, 30 BBY). No account joins these to the Gungan generators, and no source states how those generators work.
We therefore supply a mechanism of our own and propose a velocity-selective barrier, estimate from the records what speed threshold they imply, and say what the threshold costs a designer who wants both protection and passage. The mechanism, the plasma supply that feeds it and every parameter of the skin below are hypotheses of this paper. They are not findings about the Gungan generators, and the records bound them only loosely.
2. System Description
Gungan generators were carried by large beasts, and the field they raised was dome-shaped and extended over the host, where it stopped the tank and artillery fire directed at it. It admitted battle droids on foot, and the droids crossed it without, so far as any record says, being slowed or stopped. The Gungans are reported to use the plasma that lies close to the planet's core, but how that plasma was fed to the generators is not recorded, and we treat the supply as a hypothesis (Section 5).
Our principal source is a set of copies of the occupation command's engagement log and fire-control returns, released to the Republic by the Naboo authorities from the captured series (Trade Federation Naboo Occupation Command, 32 BBY). The log records formation movements by time between grid markers set at known spacing, and the fire-control returns list each round fired with its nominal speed at range and a disposition code. The order of battle is taken from Bastable (30 BBY). The speeds of walking droids are further constrained by gait data for line-infantry droids (Aldrenne, 30 BBY).
The shield passed more than marching columns. The log notes, without times or speeds, that heavy droids carrying personal emitters rolled into the shielded area, and that the host's catapults and thrown energy charges struck the attackers from inside the dome while it stood. The markers were set for walking columns, so the rolling droids have no usable speed, and we leave them out of the 12 crossing entries below and out of every bound. Their absence from the tally does not remove the tension they create: a rolling droid moves faster than a marching one, and if the skin alone governed its passage the lower bound would rise. We offer as a hypothesis, and as no more than that, that the emitter's own deflector field interacts with the skin, so that such a droid is not a clean probe of it (Section 5). The log records the effects of the outbound ordnance on the attackers' side and no outbound speeds.
In the copies there are 12 formation crossing entries, 11 at marching pace and one at double time, covering 816 droids in all (Table 1). There are 41 fire-control entries with a speed and the disposition "terminated at field", and 9 further returns reporting that blaster fire was turned, without speeds. Speeds in the log come from dead reckoning between markers, and we assign them a 15% error throughout. Whether the field was stationary in the ground frame is not stated, and we assume it was.
3. Analysis / Model
Suppose, as a first model, that the dome is a thin sheet of plasma, held by the generator, of thickness δ and with a single relaxation time τ. An intruder crossing the sheet at speed v spends a time δ/v inside it. If that time is long against τ, the sheet has time to flow around the intruder and resists only weakly, with a stress that grows in proportion to v. If the time is short against τ, the sheet has no time to flow and responds as an elastic solid, with a stiffness G. The ratio vτ/δ, which we call the relaxation number, separates the two regimes, and the transition speed is v<sub>c</sub> = δ/τ.
The idealisation is convenient because it needs no sensor, no logic and no choice by the generator, since selection falls out of the physics of a fluid with memory, and bench work on thin field sheets has already reported stress that depends on deformation rate (Brandreth, 31 BBY). We do not claim that the Gungan skin is a fluid with a single relaxation time in series with an elastic element. We claim only that it is the simplest barrier that could produce what the record shows.
To confront the model with a record that has no intermediate cases, we describe the chance that a body of speed v is admitted as 1/(1 + (v/v<sub>c</sub>)n). The exponent n measures how sharply the skin changes character. A large n is a switch, and a small one is a slow fade.
Stopping a round of areal mass μ at speed v<sub>max</sub> within the skin requires that the elastic stress acting over the thickness absorb ½μv<sub>max</sub>², so Gδ is fixed by the worst round the shield is meant to stop. That product does not depend on v<sub>c</sub>. What v<sub>c</sub> fixes is how the product is split. At a given τ, the thickness is δ = v<sub>c</sub>τ, so raising the threshold by a factor k thickens the skin by k and lets the stiffness fall to 1/k of its value. A designer who wants a shield that lets faster traffic through must therefore carry more plasma per unit area, and may use a softer and perhaps more tolerant medium. This is the central cost in the model, and it holds only if τ is a property of the medium and not of the generator.
A body crossing slowly drags against the viscous stress, which is small at a few metres per standard second, and whatever energy the passage dissipates is heated into the skin and is not available to stop the next round.
4. Validation Against Field Data
The fastest speed in any admitted record is 4.6 m/s, for the formation that advanced at double time, and the slowest in any refused record is 38 m/s, for one of four low lobs at the end of the range. The marching columns lie well below both. No admitted speed exceeds any refused speed, and because the groups are completely separated, a fitted gate is not unique, and the likelihood is flat as n grows. The records fix an interval for v<sub>c</sub> and no more.
Between 4.6 m/s and 38 m/s lies the transition speed, an interval spanning a factor of 8.3 whose geometric midpoint is 13.2 m/s. Allowing the 15% dead-reckoning error, the lower edge becomes 4.0 to 5.3 m/s and the upper 33 to 44 m/s, which widens the span to a factor of 10.9. We do not call the midpoint an estimate, only a reference value for the cost table.
We searched over v<sub>c</sub> within the interval for the smallest n at which the likelihood of the 53 recorded outcomes stays within 1.92 log units of its maximum, the usual approximate 95% criterion for one parameter, which is only approximate here because the maximum lies at the boundary. Each formation entry is treated as one outcome, because the log records crossing by formation and not by droid. The 11 marching entries run at 1.6, 1.8, 1.9, 2.0, 2.0, 2.1, 2.2, 2.3, 2.4, 2.6 and 2.9 m/s, with a standard deviation of 0.4 m/s. The smallest exponent is 1.6, at v<sub>c</sub> near 9.7 m/s. Weighting by droid instead, so that each of the 816 admitted bodies counts singly, tightens the minimum exponent to about 2.8, at v<sub>c</sub> near 23.5 m/s.
The records are thus consistent with a very sharp switch and with a gate as gentle as the one at 1.6. At 9.7 m/s and an exponent of 1.6 that gate admits a marching droid with probability 0.87 to 0.95, a double-time droid at 4.6 m/s with 0.76, and a round at 38 m/s with 0.11. It passes the marching columns as units and, counted by droid, would have turned back about one in ten.
Since the skin thickness is not recorded, we cannot compute τ. If δ were 0.5 m, then τ = δ/v<sub>c</sub> would lie between 13 and 109 ms (thousandths of a standard second), and if δ were 1.0 m, between 26 and 217 ms. The 0.5 and 1.0 m values are illustrations and are not inferred from anything in the record.
The 9 returns reporting bolts turned have no speeds, so they cannot bound v<sub>c</sub>. In the model, a bolt carries so little mass and moves so fast that its relaxation number is very large and the skin should respond elastically, which is at least not contradicted by their being turned aside. A personal emitter on a heavy droid is documented to reject bolts (Ashkelan, 30 BBY), but whether it also admits slow bodies, or how its field behaves against another, has not been tested, and we do not extend our model to it beyond the hypothesis of Section 2.
Table 2 prices alternative thresholds in the model against the records, taking the midpoint of 13.2 m/s as reference and assuming a sharp gate (large n). Under the gentlest gate the data allow, the columns blur. A light vehicle at 20 m/s would then be admitted with probability 0.24 at 9.7 m/s and not refused outright, and rounds near each threshold would be admitted in part.
Thicker skins admit more traffic and more ordnance, and thinner ones block the marching columns that the record shows passing.
5. Failure Modes
A selective barrier admits whatever is slow, whether it is a droid or a shell. Four of the 41 recorded rounds, those at 38 to 55 m/s, are lobbed rounds at the end of their range. They were stopped, which is why we place the threshold below 38 m/s. At a threshold of 40 m/s one of them would have been admitted, and at 60 m/s all four. A designer who moves the threshold upward to admit light vehicles at 20 m/s therefore buys, in the model, a leak for the slowest ordnance of an attacker.
In a skin that yields to slow intrusion, a body that stops inside the sheet is held by nothing. A body crossing the barrier slowly, pausing and crossing again meets the same permissive response each time, and the model offers no way to keep it out. An admitted body that stops is a threat inside the dome: the log has the generator brought down by fire from inside it after the crossings. Nothing in the log records a droid halting at the field, so the case is untested. The occupation command's log cannot test it, because it records only the attackers' side of the surface. Section 6 names records that could.
A droid that carries its own deflector field meets the skin with a field of its own, and the model has nothing to say about two fields in contact. The rolling droids in the log crossed, so the contact did not bar them, but without their speeds we cannot say whether the skin admitted them as slow bodies or the emitter field altered the skin where it passed. The record leaves this open.
Our model supposes a uniform sheet, but where a dome meets the ground, or where a gap opens between two generator fields, the skin thins, and the threshold δ/τ falls. The records are silent on the margin.
Plasma for the skin comes from a supply we have not seen. If the plasma that forms it is bled continuously from a reservoir or a tap on a source in the planet's core, and if the carrying beasts moved the generators, the thickness and so the threshold may have drifted over the engagement. The log records the generator brought down by fire from inside the dome after the crossings and gives no thickness history before that, so we cannot say whether the threshold drifted.
Section 3's cost argument assumes that τ belongs to the medium. If the generator could tune τ, for example by changing the coupling to the plasma, the threshold could be moved without changing the thickness and the trade would shift. The records cannot distinguish a fixed skin from a tuned one.
6. Conclusion
The records bound the transition speed of the Gungan shield between 4.6 and 38 m/s, and they say little about its sharpness. A viscoelastic skin is one barrier that would produce the pattern, and in it the cost of passage appears as thickness. We have not shown that the Gungan generators work this way.
Five tests would narrow the interval and could overturn the skin. A captured generator, if one survives, should be run against probes launched at several speeds between 5 and 40 m/s, which would give the gate's shape directly. The thickness of the sheet could be measured by the time a fast probe takes to cross it. Gungan-side accounts of the engagement, if any exist, may record bodies halting at the dome, which would test the stalled-body case, and a captured generator would test it too. The same accounts could give the range and elevation of the catapults, from which the speed of the outbound lobs follows, and those speeds are a fifth constraint, since a symmetric barrier predicts that every outbound body slower than v<sub>c</sub> passes. The plasma supply is the largest unknown, and a survey by the Naboo plasma geologists of the sites where the generators stood would show whether a source was tapped.
The column that marched through the dome on the grass plains would, on this account, have met a skin it was too slow to harden. The same dome would have been a wall to every round the host's attackers fired at it. Its sharpness is another matter. What we cannot yet say is how wide the margin was between the droids and the threshold, and whether the Gungans knew it.
References
- Trade Federation Naboo Occupation Command (32 BBY). Ground engagement log and fire-control returns, grass plains, Naboo. Trade Federation Droid Operations Records (captured), Series N, engagement log 3.
- Aldrenne, T. (30 BBY). Walking speed and gait of line-infantry droids on soft ground. Coruscant Journal of Droid Engineering, 12(2), 77–101.
- Brandreth, V. (31 BBY). Rate dependence of stress in thin repulsor-field sheets, a bench study. Journal of Repulsorlift Engineering, 9(1), 18–44.
- Barrowdene, L. (31 BBY). Energy absorption per unit area in layered deflector barriers, bench results. Journal of Repulsorlift Engineering, 9(3), 150–171.
- Ashkelan, M. (30 BBY). Personal deflector emitters on heavy droids, coverage and bolt rejection. Journal of Repulsorlift Engineering, 9(4), 250–268.
- Bastable, H. (30 BBY). The Naboo ground engagement of 32 BBY, an order-of-battle reconstruction. Journal of Republic Military History, 6(1), 3–29.
Open in Uncited Press →