Internal Ballistics: Muzzle Velocity Prediction Models

The Le Duc model, developed by the French military engineer Felix Le Duc and published in 1873, relates muzzle velocity (v) to the peak chamber pressure (Pm) and other ballistic parameters through an algebraic expression of the form:

v = a * Pm / (1 + b * Pm)

where the constants a and b are empirically determined from test-fire data for a particular combination of propellant, bullet weight, and barrel. The model predicts that muzzle velocity increases with peak pressure but does so with diminishing returns as pressure rises: velocity is a hyperbolic function of pressure, consistent with the physical observation that doubling charge weight does not double muzzle velocity. This saturation behaviour arises because at higher pressures the bullet has already accelerated to high velocity early in its bore travel, so the later portion of the combustion event contributes proportionally less to exit velocity.

In forensic reconstruction, the Le Duc model is encountered in its practical form: an examiner knows the calibre, can estimate the charge weight from recovered unburned powder or from standard loading data (such as published in Lyman Reloading Handbook or Sierra Bullets Reloading Manual in the US, or Norma Reloading Manual in Europe), and has either a measured or a SAAMI/CIP-published MAP for the cartridge. The model then allows a velocity estimate, which is compared against chronograph data or against the physical evidence (wound depth, ricochet angle, impact deformation on a known material). The accuracy of Le Duc estimates for modern propellants is typically within 3-5% of measured velocity when calibrated against a few test-fire data points. Without calibration, the estimate may drift to 8-12% depending on how well the empirical constants apply to the specific propellant-bullet combination.

Indian Ordnance specifications and procurement documents reference the Le Duc and similar classical models for initial projectile design calculations. The UK's Defence Science and Technology Laboratory (DSTL) at Porton Down and the Armament Research and Development Establishment (ARDE, now part of DRDO in India) have historically used Le Duc as a first-pass check before committing to more expensive computational codes. In US casework, the ATF Firearms Technology Branch uses similar classical approximations in rapid-response forensic reconstruction before detailed computational modelling is warranted.

The Heydenreich model, developed in the German ordnance tradition and formalised during the early 20th century, extends the Le Duc approach by explicitly incorporating the specific volume of propellant gas (a thermochemical property of the propellant), the chamber volume, and the propellant burn rate as distinct parameters. It represents the pressure-time relationship as a function of propellant mass fraction burned versus time, providing a more physically interpretable model than Le Duc's purely empirical formulation.

The practical descendants of Heydenreich's approach are the ballistic coefficient systems used by propellant manufacturers to characterise their products. Systems such as the British CE (Coefficients of Energy) system, the German Ballistische Leistungsverzeichnis, and the SAAMI-affiliated Powley computer (and its digital successors) provide tabulated burn-rate coefficients that allow ballisticians to predict the performance of a given propellant in a given cartridge without case-by-case calibration. The Powley system, developed by Homer Powley in the 1960s, became a widely used computational aid for the US reloading community and remains referenced in US forensic reconstruction practice.

In the context of forensic reconstruction, the Heydenreich-class models matter in two scenarios. First, when the propellant type can be inferred (from GSR chemistry, from headstamp data, or from unburned powder recovered from the scene), the burn-rate coefficient for that propellant family allows a velocity estimate without a test-fire. Second, when a test-fire is conducted and the measured velocity differs from the model prediction, the discrepancy can be used to narrow down what propellant class was present: a measured velocity significantly above the Le Duc or Heydenreich prediction for a standard military load suggests a higher-energy propellant (possibly double-base or a surplus military overcharge) was present.

Modern internal-ballistic prediction relies on lumped-parameter codes, which solve the differential equations of propellant combustion numerically, treating the propellant charge and gas as a system of well-mixed, spatially-uniform (lumped) quantities. This simplification is computationally tractable and sufficiently accurate for most small-arms and artillery applications, while avoiding the full three-dimensional fluid dynamics of the actual combustion event.

PRODAS (Projectile / Rocket Design and Analysis System) is a suite of computational tools developed by Arrow Tech Associates and used extensively by the US Army Research Laboratory, the UK DSTL, and NATO-nation ordnance programs. PRODAS includes an internal-ballistic module that solves the lumped-parameter equations for the propellant combustion event, outputs the pressure-time curve and muzzle conditions (velocity, spin rate at muzzle exit), and interfaces with the external-ballistic and terminal-ballistic modules. The internal-ballistic input parameters include: propellant burning rate law coefficients, specific force (force constant) of the propellant, chamber volume, bullet engraving force, barrel length, bore area, twist rate, and bullet mass. PRODAS is subject to US ITAR (International Traffic in Arms Regulations) export controls; access for forensic reconstruction purposes in non-US jurisdiction laboratories requires export licence approval.

IBHVG2 (Interior Ballistics of High Velocity Guns, version 2) is a US Army Ballistic Research Laboratory code developed in the 1980s, primarily for artillery and large-calibre systems but applicable to small-arms calibres. It implements the Lagrange model of internal ballistics, which accounts for the spatial pressure distribution along the bore (the pressure at the base of the bullet is not identical to the chamber pressure during rapid combustion) alongside the lumped-parameter combustion model. IBHVG2 is available in declassified form from the US Army Research Laboratory technical report library and has been used in academic forensic ballistics research. The European STANAG-4367 working group has developed equivalent computational standards for 120mm tank gun charges.

For forensic purposes, PRODAS or IBHVG2 outputs are typically used in high-stakes reconstruction cases: weapon-qualification disputes in military procurement litigation, failure-mode investigation for catastrophic barrel failures, or sniper-reconstruction cases requiring sub-percent accuracy in velocity estimation. The code requires a validated propellant data set, which is not always available for unknown or improvised ammunition; in those cases, the examiner falls back to Le Duc or Heydenreich estimates with explicitly stated uncertainty bounds.

The relationship between barrel length and muzzle velocity is a direct consequence of the pressure-distance curve. In the early part of bullet travel (the first few inches from the chamber), pressure is at or near peak and falling rapidly; the bullet gains most of its velocity in this region. In the later part of barrel travel, pressure has fallen to much lower values and contributes relatively little additional velocity per inch of barrel. The relationship is therefore a diminishing-returns curve, typically well approximated by:

v(L) = v_ref * (L / L_ref)^n

where n is an empirical exponent between 0.15 and 0.35 depending on the propellant and cartridge, and L_ref is a reference barrel length at which v_ref was measured. For 5.56x45mm M855, the velocity loss from cutting from a 20-inch service rifle barrel to a 14.5-inch M4 carbine barrel is approximately 120-140 m/s (roughly 10-12%). The velocity loss from cutting further to a 10.5-inch short-barrelled configuration is a further 80-100 m/s.

This relationship is critical to forensic reconstruction in cases where the recovered weapon has a non-standard barrel (cut down from a longer configuration, which is the common scenario in illicit weapon modification). An examiner who test-fires a recovered weapon and measures its muzzle velocity, then consults SAAMI or military velocity tables for the nominal barrel length of that weapon pattern, will find a discrepancy if the barrel has been cut. The magnitude of that discrepancy, compared against the barrel-length formula, gives an estimate of how much barrel was removed, corroborating the examiner's direct barrel-length measurement.

In India, the National Ballistics Centre operated by the Bureau of Police Research and Development, and the CFSL laboratories, use barrel-length-velocity reference tables derived from IOF ammunition test data and from published SAAMI/CIP specifications for imported calibres. In the UK, NABIS maintains a reference database of weapon-specific muzzle velocities tested under controlled conditions. The FBI Laboratory in the US maintains analogous reference tables for common service calibres.

For a projectile to fly stably, it must be gyroscopically stabilised: the bullet must spin fast enough that gyroscopic resistance prevents aerodynamic forces from tipping it end-over-end. The spin is imparted by the rifling in the barrel, which is characterised by its twist rate (the rate at which the rifling grooves complete one full revolution, typically expressed in inches per turn, such as 1 turn in 12 inches = 1:12).

The Greenhill formula (published by Sir Alfred George Greenhill in 1879) gives the minimum twist rate (T, in calibres per turn) required to stabilise a bullet of given length (L) and calibre (d):

T = C * sqrt(SG * d / L)

where C is a constant (Greenhill used C = 150 for lead-core bullets and C = 180 for mild-steel bullets), SG is the specific gravity of the bullet material (approximately 10.5 for gilding-metal jacketed lead core), d is the diameter, and L is the length. For a 147-grain full-metal-jacket 9x19mm bullet (approximately 29mm long, 9mm diameter), Greenhill predicts a required twist of about 1:10 inches, consistent with the 1:10 standard twist used in most 9mm service barrels. For the longer 77-grain open-tip 5.56x45mm bullet, Greenhill predicts a requirement for a faster 1:8 twist, which is why the US military adopted 1:7 twist barrels with the M16A2 in the 1980s to stabilise the then-new SS109 / M855 62-grain projectile; the original 1:14 twist of the earliest M16 (later revised to 1:12 for the M16A1) was adequate only for the shorter 55-grain M193 projectile.

The Miller stability formula, developed by Donald Miller and published in 2005, refines Greenhill for modern spitzer (pointed) bullets whose centre of pressure behaviour differs from the blunt ogive bullets for which Greenhill was calibrated. Miller expresses a dimensionless stability factor Sg:

Sg = (30 * m) / (p^2 * d^3 * L * (1 + L^2))

where m is bullet mass in grains, p is twist rate in calibres per turn, d is calibre in inches, and L is bullet length in calibres. A bullet is considered adequately stabilised when Sg is approximately 1.5 or greater. Values below 1.0 indicate the bullet will be marginally or unstable; values above 3.0 indicate over-stabilisation, which reduces the bullet's ability to compensate for yaw at close range but generally does not cause stability problems in normal use.

The forensic relevance of stability formulas is most acute in two scenarios. First, when a bullet is recovered from a wound or intermediate target in an unstable or yawed orientation (causing a keyhole-shaped rather than round impact hole, or an asymmetric wound channel), and the examiner needs to determine whether this instability arose from the weapon-ammunition combination or from impact destabilisation of an initially-stable projectile. Second, when an improvised weapon with an unknown or anomalous twist rate is recovered, and the examiner needs to determine whether it was capable of producing the wounds observed.

Chronograph instruments measure the time a bullet takes to pass between two sensors a fixed distance apart, converting the transit time to a velocity in metres per second or feet per second. Three instrument types are encountered in forensic ballistic reconstruction work.

The Oehler Model 35P is a two-screen optical chronograph widely used in US laboratory and field testing. It uses two optical skyscreens separated by a precision-measured base distance (typically 2 feet), recording the shadow of the bullet on each screen as it passes overhead. The Oehler 35P includes a proof channel (a third screen) that flags inconsistent readings from defective screens or bullet tumble. It is accurate to within 0.2% of reading under controlled conditions and is the standard for SAAMI test-fire data collection in many US manufacturer and ATF labs. Published velocity data for US-manufactured ammunition, including the velocity specifications in SAAMI Z299.4-2015, are frequently produced on Oehler-type equipment.

The MagnetoSpeed V3 uses a muzzle-attached bayonet sensor that detects the magnetic perturbation of the passing bullet. It requires no alignment with a distant screen, making it practical for field reconstruction work. Its accuracy is typically within 0.5% and it is unaffected by lighting conditions that can compromise optical chronographs. The muzzle-attachment design introduces a minor pressure perturbation (the bayonet slightly obstructs muzzle gas flow) that can shift velocity readings by 1-3 fps for very short barrels; this is negligible for most forensic purposes but should be noted in the method section of a reconstruction report.

The LabRadar is a Doppler radar chronograph that tracks the bullet's velocity continuously from muzzle to several hundred metres downrange, producing a velocity-versus-distance profile rather than a single muzzle-velocity reading. It requires no physical attachment to the weapon and does not affect muzzle gas flow. The LabRadar is increasingly used in forensic reconstruction when the examiner needs to know not just muzzle velocity but velocity at an estimated crime-scene distance: for example, estimating whether a bullet recovered from a victim at 150 metres was supersonic or subsonic at the point of impact, which matters for wound-pattern interpretation.

The forensic reconstruction protocol for a velocity dispute typically proceeds as follows: (1) recover or seize the suspect weapon; (2) recover or acquire matched ammunition, or the closest available substitute ammunition identified from headstamp analysis; (3) test-fire under controlled conditions, measuring muzzle velocity with a calibrated chronograph instrument, recording barrel temperature, ambient temperature, and humidity; (4) compare measured velocity against SAAMI/CIP/military specification tables, noting any barrel-length or condition correction required; (5) compare against the Le Duc or Heydenreich prediction for that charge weight and pressure; (6) express the result as a best estimate with stated uncertainty, referencing the comparison standards. In UK Crown Court proceedings, this reconstruction chain must be disclosed in the expert's report under the Criminal Procedure Rules Part 19 (expert evidence) and the FSR Codes of Practice. In US Federal Court proceedings, the methodology must survive Daubert scrutiny. In Indian courts under the Bharatiya Sakshya Adhiniyam 2023 Section 39 (expert evidence), a similarly documented technical basis for the opinion is expected.

1. Step 1: Weapon and barrel condition
   Measure barrel length from chamber face to muzzle. Inspect bore for erosion, obstruction, or non-standard rifling. Photograph and document twist rate by rod-and-rod method or borescope.
2. Step 2: Ammunition identification
   Read headstamp. Identify primer type (Berdan vs Boxer, single flash hole vs offset). If unfired rounds are available, weigh the charge after bullet pull. Cross-reference against SAAMI, CIP, or NATO STANAG published specifications for the identified cartridge.
3. Step 3: Model prediction
   Apply the Le Duc or Heydenreich equation with the identified propellant burn-rate coefficients, charge weight, bullet mass, and peak chamber pressure from published MAP. State the uncertainty range, typically plus or minus 5-8% without test-fire calibration.
4. Step 4: Test-fire measurement
   Fire a minimum of 5 rounds over a calibrated chronograph (Oehler 35P, MagnetoSpeed V3, or LabRadar). Record ambient conditions. Calculate mean velocity and standard deviation. Compare against model prediction and published specifications.
5. Step 5: Stability assessment
   Apply Greenhill or Miller formula for the bullet length and calibre. Confirm the tested barrel twist rate is adequate for the projectile. If tumble or keyholing was observed, document whether the stability factor Sg is below 1.5 for the weapon-ammunition combination.
6. Step 6: Report and court disclosure
   State the measured muzzle velocity, the comparison standard (SAAMI / CIP / NATO), the prediction model used, instrument model and calibration status, and the stated uncertainty. Disclose the full chain under the applicable expert-evidence rules (Criminal Procedure Rules Part 19 UK; Federal Rules of Evidence 702 US; BSA 2023 s. 39 India).

Frequently asked questions