Gurney equations

The Gurney equations are a set of mathematical formulas used in explosives engineering to relate how fast an explosive will accelerate a surrounding layer of metal or other material when the explosive detonates. This determines how fast fragments are released by military explosives, how quickly shaped charge explosives accelerate their liners inwards, and in other calculations such as explosive welding where explosives force two metal sheets together and bond them.[1]

The equations were first developed in the 1940s by Ronald Gurney[2] and have been expanded on and added to significantly since that time.

Underlying physics

When an explosive surrounded by a metallic or other solid shell detonates, the outer shell is accelerated both by the initial detonation shockwave and by the expansion of the detonation gas products contained by the outer shell. Gurney modeled how energy was distributed between the metal shell and the detonation gases, and developed formulas that accurately describe the acceleration results.

Gurney made a simplifying assumption that there would be a linear velocity gradient in the explosive detonation product gases. This has worked well for most configurations, but see the section Anomalous predictions below.

Definitions and units

The Gurney equations use and connect the following quantities:

C - The mass of the explosive charge
M - The mass of the accelerated shell or sheet of material (usually metal). The shell or sheet is often referred to as the flyer, or flyer plate.
V or Vm - Velocity of accelerated flyer after explosive detonation.
N - The mass of a tamper shell or sheet on the other side of the explosive charge, if present.
\sqrt{2E} - The Gurney Constant for a given explosive. This is expressed in units of velocity (millimeters per microsecond, for example) and compares the relative flyer velocity produced by different explosives materials.

For implosion systems, with a hollow explosive charge accelerating an inner mass towards their center, the calculations additionally take into account:

Ro - Outside radius of the explosive charge.
Ri - Inside radius of the explosive charge.

Values of \sqrt{2E} and detonation velocity for various explosives

As a simple approximate equation, the physical value of \sqrt{2E} is usually very close to 1/3 of the detonation velocity of the explosive material for standard explosives.[1] For a typical set of military explosives, the value of \frac{D}{\sqrt{2E}} ranges from between 2.79 and 3.15.

Gurney velocity \sqrt{2E} for some common explosives[1]
Density Detonation Velocity \sqrt{2E}
Explosive \frac{g}{cm^3} \frac{mm}{\mu s} \frac{mm}{\mu s}
Composition B 1.72 7.92 2.70
Composition C-3 1.60 7.63 2.68
Cyclotol 75/25 1.754 8.25 2.79
HMX 1.835 8.83 2.80
LX-14 1.89 9.11 2.97
Octol 75/25 1.81 8.48 2.80
PBX 9404 1.84 8.80 2.90
PBX 9502 1.885 7.67 2.377
PETN 1.76 8.26 2.93
RDX 1.77 8.70 2.83
Tetryl 1.62 7.57 2.50
TNT 1.63 6.86 2.44
Tritonal 1.72 6.70 2.32

Note that \frac{mm}{\mu s} is dimensionally equal to kilometers per second, a more familiar unit for many applications.

Fragmenting versus nonfragmenting outer shells

The Gurney equations give a result which assumes that the flyer plate remains intact throughout the acceleration process. For some configurations, this is true; explosion welding, for example, uses thin sheets of explosives to evenly accelerate flat plates of metal and collide them, and the plates remain solid throughout. However, for many configurations where materials are being accelerated outwards the expanding shell fractures due to stretching as it expands. When it fractures, it will usually break into many small fragments due to the combined effects of ongoing expansion of the shell and stress relief waves moving into the material from fracture points.[1]

For brittle metal shells, the fragment velocities are typically about 80% of the value predicted by the Gurney formulas.

Effective charge volume for small diameter charges

Effective charge mass for thin charges - a 60° cone

The basic Gurney equations for flat sheets assume that the sheet of material is large diameter.

Small explosive charges, where the explosives diameter is not significantly larger than its thickness, have reduced effectiveness as gas and energy are lost to the sides.[1]

This loss is empirically modeled as reducing the effective explosive charge mass C to an effective value Ceff which is the volume of explosives contained within a 60° cone with its base on the explosives/flyer boundary.

Putting a cylindrical tamper around the explosive charge reduces that side loss effectively, as analyzed by Benham.


Anomalous predictions

In 1996, Hirsch described a performance region, for relatively small ratios of \frac{M}{C} in which the Gurney equations misrepresent the actual physical behavior. [3]

The range of values for which the basic Gurney equations generated anomalous values is described by (for flat asymmetrical and open-faced sandwich configurations):

\frac{M}{C} \left [ \left ( 4 \frac {N}{C} \right ) + 1 \right ] < \frac{1}{2}

For an open-faced sandwich configuration (see below), this corresponds to values of \frac{M}{C} of 0.5 or less. For a sandwich with tamper mass equal to explosive charge mass ( \frac{N}{C} \ge 1.0 ) a flyer plate mass of 0.1 or less of the charge mass will be anomalous.

This error is due to the configuration exceeding one of the underlying simplifying assumptions used in the Gurney equations, that there is a linear velocity gradient in the explosive product gases. For values of \frac{M}{C} outside the anomalous region, this is a good assumption. Hirsch demonstrated that as the total energy partition between the flyer plate and gases exceeds unity, the assumption breaks down, and the Gurney equations become less accurate as a result.

Complicating factors in the anomalous region include detailed gas behavior of the explosive products, including the reaction products' heat capacity ratio, γ.

Modern explosives engineering utilizes computational analysis methods which avoid this problem.

The equations

Cylindrical charge equation

Cylindrical charge of mass C and flyer shell of mass M

For the simplest case, a long hollow cylinder of metal is filled completely with explosives. The cylinder's walls are accelerated outwards as described by:[1]

\frac{V}{\sqrt{2E}} = \left(\frac{M}{C}+\frac{1}{2}\right)^{-1/2}

This configuration is a first-order approximation for most military explosive devices, including artillery shells, bombs, and most missile warheads. These use mostly cylindrical explosive charges.


Spherical charge equation

Center-initiated spherical charge - spherical explosive charge of mass C and spherical flyer shell of mass M

A spherical charge, initiated at its center, will accelerate a surrounding flyer shell as described by:[1]

\frac{V}{\sqrt{2E}} = \left(\frac{M}{C}+\frac{3}{5}\right)^{-1/2}

This model approximates the behavior of military Grenades, and some Cluster bomb submunitions.


Symmetrical sandwich equation

Symmetrical sandwich - flat explosives layer of mass C and two flyer plates of mass M each

A flat layer of explosive with two identical heavy flat flyer plates on each side will accelerate the plates as described by:[1]

\frac{V}{\sqrt{2E}} = \left(2\frac{M}{C}+\frac{1}{3}\right)^{-1/2}

Symmetrical sandwiches are used in some Reactive armor applications, on heavily armored vehicles such as main battle tanks. The inward-firing flyer will impact the vehicle main armor, causing damage if the armor is not thick enough, so these can only be used on heavier armored vehicles. Lighter vehicles use open-face sandwich reactive armor (see below). However, the dual moving plate method of operation of a symmetrical sandwich offers the best armor protection.


Asymmetrical sandwich equation

Asymmetrical sandwich - flat explosives layer of mass C, flyer plates of different masses M and N

A flat layer of explosive with two different mass flat flyer plates will accelerate the plates as described by:[1][4][5]

Let: A = \frac{1 + 2 \frac{M}{C}}{1 + 2 \frac{N}{C}}

\frac{V_M}{\sqrt{2E}} = \left(\frac{1+A^{3}}{3(1+A)}+A^{2}\frac{N}{C}+\frac{M}{C}\right)^{-1/2}


Infinitely tamped sandwich equation

Infinitely tamped sandwich - flat explosives layer of mass C, flyer plate of mass M, and infinitely heavy backing tamper

When a flat layer of explosive is placed on a practically infinitely thick supporting surface, and topped with a flyer plate of material, the flyer plate will be accelerated as described by:[1]

\frac{V_M}{\sqrt{2E}} = \left(\frac{M}{C}+\frac{1}{3}\right)^{-1/2}


Open-faced sandwich equation

Open-faced sandwich (no tamping) - flat explosives layer of mass C and single flyer plate of mass M

A single flat sheet of explosives with a flyer plate on one side, known as an "Open-faced sandwich", is described by:[1]

Since:

N = 0

Then:

A = 1 + 2\left(\frac{M}{C}\right)

Which gives:

\frac{V}{\sqrt{2E}} = \left [ \frac{ 1 + \left ( 1 + 2 \frac{M}{C} \right )^{3}}{6 \left ( 1 + \frac{M}{C} \right ) } + \frac{M}{C} \right ] ^{-1/2}

Open-faced sandwich configurations are used in Explosion welding and some other metalforming operations.

It is also a configuration commonly used in reactive armour on lightly armored vehicles, with the open face down towards the vehicle's main armor plate. This minimizes the reactive armor units damage to the vehicle structure during firing.


Imploding cylinder equation

Uniformly initiated cylindrical charge imploding an inner mass - cylinder shell explosive charge of mass C, outer tamper layer of mass N, and inner imploding cylindrical flyer shell of mass M, with inner explosive charge radius Ri and outer charge radius of Ro

A hollow cylinder of explosives, initiated evenly around its surface, with an outer tamper and inner hollow shell which is then accelerated inwards ("imploded") rather than outwards is described by the following equations.[6]

Unlike other forms of the Gurney equation, implosion forms (cylindrical and spherical) must take into account the shape of the control volume of the detonating shell of explosives and the distribution of momentum and energy within the detonation product gases. For cylindrical implosions, the geometry involved is simplified to include the inner and outer radii of the explosive charge, Ri and Ro.

\beta = \frac{R_o}{R_i}

a = 1

A = \frac{V_o}{V_i} = \frac {\left ( \frac{M}{C} + a \left ( \frac {M}{C} \right ) \left ( \beta - 1 \right ) + \frac { \beta + 2 } { 3 \left ( \beta + 1 \right ) } \right ) } {\left ( \frac{N}{C} + \frac{2 \beta + 1}{3 \left ( \beta + 1 \right )} \right ) }

\frac{V_m}{ \sqrt{2E} } = \left [ A \left \{ \frac { \left ( \frac {M}{C} + \frac { \beta + 3 } {6 \left ( \beta + 1 \right ) } \right ) } { A } + A \left ( \frac{N}{C} + \frac{3 \beta + 1}{6 \left ( \beta + 1 \right ) } \right ) - 1/3 \right \} \right ] ^{-1/2}

While the imploding cylinder equations are fundamentally similar to the general equation for asymmetrical sandwiches, the geometry involved (volume and area within the explosive's hollow shell, and expanding shell of detonation product gases pushing inwards and out) is more complicated, as the equations demonstrate.

The constant a was experimentally and analytically determined to be 1.0.


Imploding spherical equation

Uniformly initiated spherical charge imploding an inner mass - spherical shell explosive charge of mass C, outer tamper layer of mass N, and inner imploding spherical flyer shell of mass M

A special case is a hollow sphere of explosives, initiated evenly around its surface, with an outer tamper and inner hollow shell which is then accelerated inwards ("imploded") rather than outwards, is described by:[6]

\beta = \frac{R_o}{R_i}

a = 1

A = \frac{V_o}{V_i} = \frac {\left [ \frac{M}{C} + \left ( a \frac{M}{C} \right ) \left ( \beta^2 - 1 \right ) + \frac { \beta^2 + 2 \beta + 3 } {4 \left ( \beta^2 + \beta + 1 \right ) } \right ] }{ \left ( \frac{N}{C} + \frac {3 \beta^2 + 2 \beta + 1 } { 4 \left ( \beta^2 + \beta + 1 \right ) } \right )}

\frac{V_m}{ \sqrt{2E} } = \left [ A \left \{ \frac { \left ( \frac {M}{C} + \frac { \beta^2 + 3 \beta + 6} {10 \left ( \beta^2 + \beta + 1 \right ) } \right ) } { A } + A \left ( \frac{N}{C} + \frac{6 \beta^2 + 3 \beta + 1}{10 \left ( \beta^2 + \beta + 1 \right ) } \right ) - \frac { 3 \beta^2 + 4 \beta + 3} {10 \left ( \beta^2 + \beta + 1 \right ) } \right \} \right ] ^{-1/2}

The spherical Gurney equation has applications in early nuclear weapons design.


Applications of the Gurney equations

[7]

See also

References

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 Cooper, Paul W. (1996). "Acceleration, Formation, and Flight of Fragments". Explosives Engineering. Wiley-VCH. pp. 385–394. ISBN 0-471-18636-8.
  2. ↑ Gurney, R. W. (1943). "The Initial Velocities of Fragments from Bombs, Shells, and Grenades, BRL-405". Ballistic Research Laboratory, Aberdeen, Maryland.
  3. ↑ Hirsch, E. (1995). "On the Inconsistency of the Asymmetric-Sandwich Gurney Formula When Used to Model Thin Plate Propulsion". Propellants, Explosives, Pyrotechnics 20 (4): 178–181. doi:10.1002/prep.19950200404.
  4. ↑ Jones, G. E.; Kennedy, J. E.; Bertholf, L. D. (1980). "Ballistics calculations of R. W. Gurney". Am. J. Physics 48 (4): 264–269. doi:10.1119/1.12135.
  5. ↑ Kennedy, J. E. (March 1979). Explosive Output for Driving Metal. Behavior and Utilization of Explosives Symposium (12th). ASME/UNM.
  6. ↑ 6.0 6.1 Hirsch, E. (1986). "Simplified and Extended Gurney Formulas for Imploding Cylinders and Spheres". Propellants, Explosives, Pyrotechnics 11 (1): 6–9. doi:10.1002/prep.19860110103.
  7. ↑ Nuclear weapon