The extra-large light-gas gun of the Fraunhofer EMI: Applications for impact cratering research


* Corresponding author E-mail:


Abstract– The extra-large light-gas gun (XLLGG) at the Fraunhofer Ernst-Mach-Institut (EMI, Efringen-Kirchen, Germany) is a two-stage light-gas gun that can accelerate projectile masses of up to 100 g up to velocities of 6 km s−1. The accelerator’s set-up allows various combinations of pump and launch tubes for applications in different fields of hypervelocity impact research. In the framework of the MEMIN (Multidisciplinary Experimental and Modeling Impact Research Network) program, the XLLGG is used for mesoscale cratering experiments with projectiles made of steel and of iron meteorites, and targets consisting of sandstone and other rocks. The craters produced with this equipment reach a diameter of up to 40 cm, a size unique in laboratory cratering research. With the implementation of neural networks, the acceleration process is being optimized, currently yielding peak velocities of 7.8 km s−1 for a 100 g projectile. Here, we summarize technical aspects of the XLLGG.

Introduction: Two-Stage Light-Gas Guns

Experimental impact research is a very important tool for understanding cratering and ejection mechanics, and the influence of target and projectile properties on the impact process. However, significant results can only be obtained by (1) systematic variations of experimental parameters (e.g., size, velocity, and density of the projectile) and the target (e.g., porosity, water-saturation, composition), (2) realistic dimensions (e.g., grain size of the target in relation to the shock plateau), and (3) a complete documentation in real time, combined with a thorough postmortem analysis of the crater, the subcrater basement, the ejecta, and the fate of the projectile. In this context, access to hypervelocity acceleration facilities is a basic requirement for successful research. The research unit MEMIN, which is focused on impact cratering into solid geological materials (e.g., Kenkmann et al. 2011; Poelchau et al. 2013), is using the extra-large light-gas gun (XLLGG) at the EMI.

Two-stage light-gas gun accelerators are most suited for accelerating projectiles with masses ranging from several hundred micrograms up to several kilograms to impact velocities of several km s−1. These types of gun accelerators have been operated since the end of the 1940s (Crozier and Hume 1957). Overviews of working principles and specific set-ups are given, for example, by Asay and Shahinpoor (1993), Schneider and Schäfer (2001), or Chhabildas et al. (2005).

During the operation of a two-stage light-gas gun, gas pressures in excess of 1 GPa (10,000 bar) are generated temporarily, exerting enormous loads onto the gun components. Safe control of such tremendous pressures is very demanding, with regard to both engineering design as well as operational procedures. At increasing projectile kinetic energy, the most heavily loaded parts of the light-gas guns experience considerable wear and erosion (e.g., Bogdanoff 1998; Cornelisen and Watts 1998), causing experiments at high kinetic energies to be very expensive or even impossible. Thus, the main factor limiting the performance of such guns is of technological nature and related to handling the required high pressures. For this reason, the highest muzzle velocity ever achieved by a two-stage light-gas gun, which amounted to 11.3 km s−1 obtained at NASA Ames facility (Seigel 1965), was never attained again. Further factors that limit the muzzle velocity are real gas behavior of the highly compressed gas, and frictional forces between the piston or sabot and the inner surfaces of the pump tube or the gun barrel, respectively. Three-stage light-gas guns provide an interesting alternative to two-stage gun technology. Using a three-stage set-up, Piekutowski and Poormon (2006) were able to accelerate masses of 0.154 g to velocities of 8.65 km s−1 with only limited damage to the launcher components. While still under development, the advancement of this technology could lead to reproducible experiments in excess of 9 km s−1.

Table 1 summarizes the performance of the largest and fastest two-stage light-gas gun accelerators worldwide by listing the facility parameters and selected reported experimental results. The list includes only accelerators that do not change the physical state of the projectile during acceleration. The geometries of all facilities vary considerably: 1.8 to 30.5 m pump tube length, 40 to 355.6 mm pump tube diameter, 1.5 to 58.5 m launch tube length, 5.6 to 203.2 mm launch tube diameter. Only two facilities (EMI SLGG and NASA Ames) exist worldwide that have reported test results above 9 km s−1 within the last 20 yr, and only two institutions are able to accelerate projectile masses of 100 g or above to velocities in excess of 5 km s−1: these are the Arnold Engineering Design Center (AEDC) and EMI XLLGG.

Table 1. Examples of acceleration performance of the world’s most powerful two-stage light-gas guns.
  1. CAPS Kent: Centre for Astrophysics and Planetary Science, Kent, UK; Burchell et al. (1999).

  2. TiTech: Tokyo Institute of Technology, Tokyo, Japan; Moritoh et al. (2001).

  3. EMI SLGG: EMI’s Space Light-Gas Gun, Freiburg, Germany; Schneider et al. (1995).

  4. NASA AMES: Ames Research Center, California, USA; Berggren and Reynolds (1970); Glenn (1987).

  5. LLNL: Lawrence Livermore National Laboratories, California, USA; Glenn (1987).

  6. XLLGG: EMI’s Very Large Light-Gas Gun, Efringen-Kirchen, Germany; (this study).

  7. AEDC: Arnold Engineering Development Center, Tennessee, USA; Carver et al. (2008).

  8. *Projectile mass excluding sabot.

  9. –: no information available.

Pump tube length (m)
Pump tube diameter (mm)12.7504045.064.490150355.6355.6
Launch tube length (m)0.73.581.53.99.01230.558.5
Launch tube diameter (mm)4.311.88.55.612.7285083.8203.2
Hydrogen fill pressure (MPa)40.60.2760.0691.01.5
Diaphragm burst press. (MPa)34137.982.725.0
Piston mass (g)12290.4128100888454013000
Max. piston velocity (m s−1)830833
Projectile mass (g)780.60.0048*0.0453*0.9415.910050012000
Projectile velocity (km s−1)

Hypervelocity impact experiments for basic planetological research have mainly been performed using smaller to medium caliber light-gas guns, for example, at NASA Ames Vertical Gun Range (e.g., Smrekar et al. 1986; Schultz et al. 2007; Hermalyn and Schultz 2011), the Open University (McDonnell 2006), and the University of Kent (Burchell et al. 1999; Burchell and Whitehorn 2003). In these facilities, mostly small projectiles with diameters up to a few millimeters were fired onto consolidated or unconsolidated geomaterials. A major shortcoming is that millimeter-sized projectiles are in the order of the grain size of relevant targets in earth and planetary sciences (e.g., sandstone). The XLLGG, however, allows hypervelocity experiments to be performed with larger projectiles (and, hence, higher kinetic energies).


The XLLGG accelerator (Fig. 1) has a maximum length of 52 m and can be set up as a two-stage light-gas gun or single-stage launcher. As a single-stage launcher, only a small powder chamber and launch tube with a maximum diameter of 50 mm is used. In the two-stage set-up, configurations based on two different pump tubes with either 100 or 154 mm caliber and lengths between 7 and 22 m are employed. Launch tube calibers vary between 25 and 70 mm and between 6 and 12 m length.

Figure 1.

 Components of the impact facility of the two-stage light-gas gun XLLGG.

In the blast tank, the projectile velocity is first measured by a laser light barrier consisting of two laser sheets as the projectile and sabot exit the muzzle. Due to aerodynamic effects of the blast, the muzzle velocity may not represent the final velocity. Therefore, a series of three X-ray tubes that are triggered by the laser light barrier expose the position of the projectile on X-ray film further downrange in the blast tank. Evaluation of the X-ray film yields velocities with a precision of ±1 m s−1. The projectile travels several meters until impact in the target chamber and is decelerated through aerodynamic drag of the chamber’s atmosphere by between 5 and 20 m s−1, depending on the set-up.

The target chamber is a 5.5 × 3.5 × 3.5 m bunker that is accessed from the back through an air-tight, double-walled door (Fig. 1). The target chamber can accommodate meter-sized target rocks for cratering experiments. For a better handling of the MEMIN experiments, a jib crane with a load capacity of up to 5 tons was fitted into the target chamber. The crane has armor plating and withstands vacuum pressures down to 20 mbar.

The target chamber walls have eight portholes available for different types of diagnostics. For impact cratering experiments, optical portholes allow documentation of the cratering process with various high-speed cameras, including Photron or Shimadzu video cameras capable of recording at up to 10fps (Kenkmann et al. 2011; Hoerth et al. 2013). The target itself can also be directly equipped with a variety of sensors. Sensor signals are typically transmitted out of the chamber through a flange with BNC connectors to a transient recorder. As an example, target rocks can have ultrasound sensors or calibrated pressure gauges fastened to the rock surface or embedded in the rock to measure the pressure pulse of the impact or postshock reverberations (e.g., Schäfer et al. 2006). Light sensors can be applied to the target surface for a detailed spectral analysis of the initial impact phase or a simple time-resolved registration of the impact flash. The mechanical recovery of ejected material is also possible. For the MEMIN experiments, a plywood and Plexiglas container was placed on the target surface and equipped with Vaseline and phenolic foam plates for a localized retrieval of ejected rock and projectile particles (Sommer et al. 2013).

Performance Optimization of the XLLGG

The projectile velocity of a light-gas gunshot depends on many parameters (see Table 1) and cannot be predicted easily. Therefore, neural networks are being used to optimize the performance of the XLLGG at EMI. These neural networks are able to “learn” coherencies while being trained with data sets from experiments. Our network was trained with data from over 170 shots, and is used to carry out parameter studies to optimize parameter sets for further tests with the XLLGG. The Stuttgart Neural Network Simulator (SNNS) package was initially used and refined for data analysis. Results from the analysis indicate that, in brief, heavy pistons, low light-gas prepressures, short pump tubes, and long launch tubes will result in the highest gun performance.

Previously, both the theoretical and measured maximum velocity for a 100 g projectile was at about 6.3 km s−1 for the XLLGG. Using an experimental set-up optimized by neural networks, the highest measured velocity is now 7.766 km s−1 for a plastic cylinder with a mass of 100.2 g (Fig. 2). This increased performance was realized by reducing the pump tube length from 22 to 14 m, increasing the hydrogen prepressure from 10 to 15 bar and changing the powder mass, so that the maximum pressure in the high-pressure section was less than 7000 bar. This also reduces the probability of damaging the high-pressure section and the membrane section while firing the gun.

Figure 2.

 Performance of the XLLGG using different launch tube calibers. The red circles show test shots with a 50 mm caliber launch tube after using neural networks (NN) for parameter optimization.

Conclusion: The XLLGG in Planetary Science

The two-stage light-gas gun XLLGG at the EMI is an ideal hypervelocity launcher for planetary applications, including the study of impact cratering processes. Due to the size of the XLLGG, meter-sized solid-rock targets can be impacted with real-time monitoring. The size of the experimental craters (D < 40 cm), produced using 12 mm projectiles accelerated to 4.5 km s−1 helps to close the gap between smaller scale laboratory cratering experiments, small high-explosive cratering experiments, and planetary-scale impacts in the strength-dominated regime. An additional benefit of these “mesoscale” MEMIN experiments is the large amount of material (ejecta, crater floor, subcrater basement) available for postimpact analysis. However most importantly, the critical issue of a small ratio of projectile to grain size of the target material is reduced. Potentially, higher impact velocities can be implemented at the XLLGG in the near future for different projectile types, due to the use of neural networks for performance optimization. These improvements can further assist the MEMIN project and planetary scientists in understanding cratering processes at high velocities and energies.

Acknowledgments— Research for the MEMIN project is funded by the German Research Foundation (DFG), grant TH805/4-1 in the framework of the DFG research unit FOR-887 “Experimental Impact Cratering—The MEMIN Program.” We also thank the technicians at EMI for their indispensable work and support. Comments and reviews by M. Price, M. Cintala, A. Deutsch, and N. Artemieva greatly helped to improve this manuscript and are gratefully acknowledged.

Editorial Handling— Dr. Natalia Artemieva