NASA LaRC Landing and Impact Research Facility

NASA Bob Nye. Public domain (NASA / US government work).

A 240 ft steel A-frame gantry at NASA LaRC, built in 1965 as the Lunar Landing Research Facility to fly Apollo crews through the last 150 ft of a lunar descent under a suspension carrying five sixths of the vehicle weight, converted at the end of Apollo into a full-scale aircraft crash test facility, and now the Landing and Impact Research Facility used for spacecraft landing attenuation, airframe crashworthiness and water impact [1][3].

Laboratory data

Parameter	Value
Operator	NASA LaRC, Research Directorate [1][6]
Location	Hampton, Virginia, United States [1]
Commissioned	Gantry completed 1965; hydro impact basin added 2011 [1][6]
Type	Outdoor A-frame gantry, water impact basin and four drop towers
Floor area	Impact strip 121 x 9.0 m; basin 35 x 27 m [2][6]
Capabilities	Gantry, Hydro Impact Basin, drop towers [6]
Simulant or terrain	No simulant. Reinforced concrete strip and an adjacent soft soil area [1][2]
Instrumentation	Photogrammetry: Phantom 9 cameras, 1000 fps at 1632 x 1200 [3][6]
Ground truth	Painted 1.0 m grid on the strip; 30 x 12 m gridded backboard on rails [2]
Fidelity limits	Pitch angle accuracy 4.25 degrees; uncompensated umbilical moment
Access	Named point of contact; no fee schedule or lead time published [6]
Cited by	Apollo Lunar Module descent training [1], Orion landing attenuation [3][5]

Capabilities

Gantry

NASA Mark Knopp. Public domain (NASA / US government work).

Parameter	Value
Working volume	240 ft (73 m) high, 400 ft (122 m) long, 265 ft (81 m) wide at the base [1][6]
Test article limits	Maximum bridge crane load 65,000 lb (29,500 kg) [6]
Vacuum	Not applicable. Outdoor structure at ambient pressure
Temperature	Not applicable. No thermal control; Virginia outdoor conditions
Illumination	Not published. Historic LLRF tests were run at night
Simulant or terrain	Reinforced concrete strip 121 x 9.0 m, 0.2 m thick, plus a soft soil area [1][2]
Slope	Not applicable. The impact surface is level
Gravity offload	Not offered. The historic 5/6 suspension is not a current capability [6]
Instrumentation	Phantom 9 cameras, 1000 fps at 1632 x 1200, on 24 mm lenses [3]

The gantry is a steel A-frame of truss elements with three sets of inclined legs giving vertical and lateral support and one set at the east end giving longitudinal support, with legs 81 m apart at the ground narrowing to 20 m apart at the 66 m level, where a movable bridge traverses the full length [1][2][6]. An enclosed elevator, a stairway and catwalks give access to the overhead work platforms [1][2][6]. Maximum drop height is approximately 200 ft (61 m).

Impact conditions are set independently of article mass: flight path angles to -60 degrees from swing cable length, suspension height to a 49 m maximum giving impact velocity along the flight path up to 27 m/s, and up to about 30 degrees in angle of attack and roll from suspension cable lengths, with only small yaw adjustment possible because of clearance between the pullback harness and the empennage [2]. Achieved accuracy is flight path within 8 percent, velocity within 6 percent and pitch angle to 4.25 degrees [2].

The operator’s current article figure is a maximum bridge crane load of 65,000 lb (29,500 kg), which supersedes the 30,000 lb (13,600 kg) article limit given in the 2013 history [6]. The original cable systems were operated with aircraft of 2721 kg nominal gross mass at a 2 g dynamic swing load; with block and tackle cable arrangements to reduce winch forces, the system could crash test aircraft to 14,000 kg gross mass [1][2][6].

The historic partial-gravity suspension is not part of the current capability. It supported five sixths of the total weight of the Lunar Excursion Module Simulator, leaving one sixth on the vehicle’s own thrusters, so that crews flew the last 150 ft of a lunar descent in a dynamically correct gravity field [1]. NASA now describes that arrangement only in the past tense, and it appears in neither the operator’s current Gantry Key Features nor its Other Facility Equipment list; whether the hardware was physically removed is not published [1][2][6].

Hydro Impact Basin

Sean Smith. Public domain (NASA / US government work).

Parameter	Value
Working volume	115 x 90 x 20 ft deep (35 x 27 x 6 m) [6]
Test article limits	Shares the gantry’s 65,000 lb maximum bridge crane load
Simulant or terrain	Water
Gravity offload	Not offered. Articles are released into free flight at 1 g
Instrumentation	Shared photogrammetry and data acquisition [3][6]

The basin was added in 2011 at the west end of the gantry and is where the water landing tests are run [6]. Orion crew vehicle water landings, and land impact testing of the capsule that will return Mars samples to Earth, have been run at the complex.

Drop towers

Parameter	Value
Working volume	72 ft outdoor; 50 ft three-rail outdoor; 30 ft four-rail and 15 ft two-rail indoor [6]
Instrumentation	Portable photogrammetric backboard, 16 coded targets in a 10 x 6 ft array [3]

Four towers cover the vertical-only drops that do not need the gantry’s pendulum swing: a 72 ft outdoor tower, a 50 ft three-rail outdoor tower, a 30 ft four-rail indoor tower and a 15 ft two-rail indoor tower [6]. The portable rolling calibration backboard was built for vertical drops under the 70 ft tower [3].

The operator’s current instrumentation list for the complex is anthropomorphic test devices, materials testing machines, high-speed cameras, ruggedized data acquisition systems, three-dimensional scanners, accelerometers, load cells and string potentiometers.

How a swing test is run

In its original configuration the facility was a partial-gravity flight trainer: the suspension supported five sixths of the total weight of the Lunar Excursion Module Simulator, so that the vehicle’s own thrusters worked against the remaining one sixth and the pilot flew the last 150 ft of a lunar descent in a dynamically correct gravity field [1]. The ground beneath the gantry was reshaped to resemble lunar terrain, and many tests were run at night to reproduce the lighting of the actual landing [1].

After Apollo the same structure became a crash test facility. Two pivot-point platforms at the west end carry 26.7 kN winches and sheave systems controlling swing cable length, with cable locks protecting the winches from overload and dynamic reaction; a pullback platform under the movable bridge carries a third 26.7 kN winch for the pullback cable, and the bridge is positioned so that the pullback and swing cables meet at the required 90 degrees [2].

The test sequence is three pyrotechnic events. The article, suspended on two swing cables and drawn back and up by the pullback cable, is released from the pullback harness and swings pendulum style toward the impact surface; the swing cables are pyrotechnically separated a short distance above the surface so that the article is completely unrestrained through the impact; and the umbilical cable is separated at a predetermined time after swing-cable separation, during the post-impact skid [1][2]. Flight path angle up to -60 degrees comes from swing cable length; velocity along the flight path, up to 27 m/s, comes from the 49 m maximum suspension height.

The suspension geometry is set so that the swing and pullback cable force vectors both pass through the article’s center of gravity, leaving it free to pitch about its own center of gravity, with two sets of pitch cables restraining the planned angle of attack during preparation and swing. A 0.76 cm steel support cable is sandwiched into the umbilical wiring to carry load and protect the data wires from unintended tensile loads; it passes through the top of the fuselage and attaches to the main spar near the center of gravity, while the data wires connect 1.1 m above the center of gravity [2].

The impact surface is the reinforced concrete strip, with a painted 1.0 m grid, backed by a 30 m by 12 m gridded photographic backboard on rails so it can be moved to whichever point along the strip a test uses [1][2]. A soft soil area exists alongside, and impact surface was a standard test variable in the general aviation crash program [1][2].

Instrumentation

Measurement at the facility is photogrammetric rather than contact-based wherever possible. The high-speed cameras are Phantom 9 units run at 1000 frames per second at the full 1632 x 1200 sensor [3]. Swing tests use 24 mm lenses with the cameras at least 50 ft from the test article and 80 to 90 ft from the calibration backboard, focused at infinity so that article and backboard are both sharp, which gives a calibrated volume of roughly 62 ft wide by 45 ft tall and more than 30 ft deep.

Calibration is the part that outdoor work at this scale changes. Handheld panels are useless over a 50 ft field, so the facility uses two custom objects instead: a large fixed backboard with 24 coded targets in a 50 x 36 ft array for gantry swing tests, and a portable backboard on rollers with 16 coded targets in a 10 x 6 ft array for vertical drops under the 70 ft tower [3]. The backboard is held still and the cameras are moved, 16 to 25 images each, from as much as 40 degrees off axis and 30 to 120 ft away. Sunlight is the dominant nuisance: sharp contrast and hard shadows defeat the software’s target recognition, so images are artificially flattened on bright days and camera settings held identical between calibration and test [3].

Some targets on the backboard are left inside the test field of view and reused for the coordinate reference frame. Because the backboard does not move, the residual displacement the software computes for its targets is the measurement noise: about plus or minus 0.05 in, a 0.10 in range across a 62 ft field, or 0.01 percent [3]. Vehicle pitch, yaw and roll are recovered as projection angles between lines drawn on the vehicle and lines on the backboard, and differentiated for angular rates at impact. Every camera and data system is tied to an IRIG-B time code generator so that otherwise incompatible systems share one time stamp [3].

The same photogrammetric approach was applied to crew exploration vehicle airbag landing attenuation systems, measuring the deforming airbag and the capsule together during impact [5].

What it does not reproduce

Pitch, exactly. The swing cable catenary introduces pitch error into the system, and the pitching velocity of the article at swing cable separation is a consequence of it [2]. Published accuracy is flight path within 8 percent, velocity within 6 percent, and pitch angle to 4.25 degrees [2]. The pitch figure is an absolute angle, not a percentage, and is the loosest of the three.

The umbilical. The mass of the umbilical acting 1.1 m above the article’s center of gravity creates a pitching moment that varies with drag on the umbilical, and it is not compensated for during the swing phase [2]. Analytical predictions of the swing do not consider aerodynamic forces after swing-cable separation nor catenary effects [2].

Yaw. Only small yaw adjustments are possible because of the clearance between the pullback harness and the article’s empennage. Additional yaw requires removing the stabilizers and simulating them with concentrated masses, which changes the structure being tested [2].

Velocity above the pendulum limit. The 27 m/s ceiling set by the 49 m suspension height was sufficient for the take-off and landing speeds of small general aviation aircraft, 81 to 88 fps, but not for the larger pressurized aircraft tested in the early 1980s, for which a Velocity Augmentation System using rockets had to be added [1].

Article mass. The operator’s current figure is a 65,000 lb maximum bridge crane load [6]; the 2013 history gives an article limit of 30,000 lb or less [1]. The original cable systems were rated for a 2721 kg article at a 2 g dynamic swing load, extended to 14,000 kg only with block and tackle arrangements to reduce winch forces [2].

Gravity, except in the original configuration. The five sixths weight relief that made the LLRF a lunar simulator applied to a tethered vehicle in powered descent, and it relieved five sixths of Earth weight rather than producing five sixths of lunar gravity [1]. It is not offered today: NASA writes about it only in the past tense and lists it among neither the gantry’s key features nor its other facility equipment [6]. A crash test article in free flight after swing-cable separation is at 1 g, and impact accelerations measured on the concrete or the soil are terrestrial.

Soil. The soft soil area is terrestrial soil, not a regolith simulant, and its preparation is not described in the facility literature [1][2]. Soft soil impacts were run specifically to study modifications that would let an aircraft skid along the surface rather than decelerate longitudinally [1], which is a crashworthiness question rather than a terramechanics one.

Campaigns run there

Apollo Lunar Module descent training, 1965 to end of Apollo. Crews flew a Lunar Excursion Module Simulator tethered to the gantry through the final 150 ft of descent, at five sixths weight relief, over a reshaped lunar-looking surface, frequently at night [1].

General aviation crash test program, 1974 to 1983. Thirty-three full-scale crash tests in ten years [1]. Test articles were mostly Piper Aztecs and Cherokees obtained for scrap aluminum value because they had been submerged in a flood at the Piper plant and could not be certified, retrofitted or sold; later tests used Cessna 172s and larger pressurized Piper Navajos [1]. Variables were impact velocity, airframe attitude at impact, and impact surface, hard or soft soil. The stated purpose was definitive data on structural response and on the loads transmitted to occupants, for correlation with analytical prediction and for evaluating crashworthy structure, seat and restraint concepts [1].

Rotorcraft crash testing. Helicopters are tested with lower forward and higher vertical velocity than fixed wing aircraft; the 1999 crash test of a Sikorsky prototype ran at 31.5 fps forward and 38 fps vertical, against 82 fps forward and 31 fps vertical for the 1994 Lear Fan 2100 test [1]. A series was conducted on both soft soil and concrete, and a further set on soft soil across a range of roll, pitch and yaw angles, on articles of nominal 9,000 lb gross weight including external fuel system, simulated fuel, swing fixture and instrumentation [1].

Orion and crew exploration vehicle landing, 2003 onward. The Constellation program ran five swing tests in the summer of 2009 on a 5146 lb, 8 ft diameter subscale Orion boilerplate into a 4 ft high by 20 ft wide by 78 ft long sand bed, varying vertical and horizontal velocity to assess abort land landing conditions; the sand bed obscured the bottom row of calibration targets and the cameras had to be tilted so the calibrated cube sat on the sand surface [3]. Water landing tests of the Orion crew vehicle, and land impact testing of the capsule that will return Mars samples to Earth, have since been run at the complex [6]. Photogrammetric measurement of airbag landing attenuation systems characterized airbag and capsule deformation through impact [5]. Numerical modeling of space capsule ground landings was correlated against the facility’s drop test data [4].

MD-500 helicopter crash test, 2009. A full-scale MD-500 mass simulator was swung into the impact surface and its conditions recovered photogrammetrically at 27.1 ft/s horizontal and 17.5 ft/s vertical with 1.8 degrees pitch, 6.7 degrees yaw and 3.4 degrees roll, against planned conditions of 28 ft/s horizontal, 18.2 ft/s vertical and zero attitude angles [3]. A later test in the series added a speckle pattern for full-field strain, which resolved a strain field on the tail skin but at an effective gauge length of about 6 in, too coarse to match a strain gauge rosette painted underneath it [3].

Current work. The complex is equipped for coupon through airframe testing with indoor and outdoor drop towers, anthropomorphic test devices, materials testing machines and three-dimensional scanners alongside the gantry [6].

Other ways landing and impact are tested

The Langley gantry reproduces impact at 1 g with a free-flying article [1][2]. Other facilities answer different parts of the landing question, and the differences set what each can and cannot say about a lander or rover.

Aircraft crash history. The facility’s crash programs from 1974 onward are surveyed separately from the history cited above, including helicopter qualification work, fuselage section drops, crew escape capsule drops and component work for the Mars Sample Return Earth Entry Vehicle [7].

Robot-held touchdown. DLR’s Landing and Mobility Test Facility in Bremen suspends the test object from a heavy-duty industrial robot on a linear rail through a weight offloading suspension, and used it for Philae landing gear touchdown testing [8]. The same facility dropped and dragged instrumented lander footpads onto prepared sand for a European lunar lander study [9], and has run drop tests of a preliminary model of the MMX rover [10]. The Philae work shows an offload rig contaminating the measurement it exists to make [8]. The robot approach controls release conditions more finely than the pendulum, at much smaller article mass.

Subscale dynamic similitude. JPL’s Mars Sample Return lander tests used a 3/8 scale article with length, acceleration and stress scaled by similitude, because a full-scale drop would need both a gravity offload and a very heavy article [16]. This trades fidelity of mass and velocity for feasibility, where Langley drops full-scale articles.

Airbag landing under Mars conditions. Mars Exploration Rover airbag loads were characterized in full-scale drops inside a vacuum chamber at Martian pressure, then turned into a qualification specification for a shaker rather than a centrifuge or pull tower [17][18]. Mars Pathfinder’s lander deployment actuators were qualified for the righting sequence after airbag landing [30]. Langley’s Orion airbag work measured the bag and capsule with photogrammetry in ambient air on soil [5].

Chinese landing test field. China reported hover and obstacle-avoidance tests of its first Mars lander at a Hebei test field with simulated Martian gravity [19], and a landing and take-off verification of the Lanyue crewed lander at the same field [20]. Both are press items; instrumentation and gravity method are not described in them.

Flight reconstruction of touchdown. On-orbit and surface data serve as the only full-fidelity check of ground landing tests. Philae’s four contacts with comet 67P were reconstructed from landing gear data, with a first-touchdown energy partly absorbed by the soil [21], and its rotational motion was fitted with a multibody model using solar panel power [22]. Huygens’ first seconds on Titan were reconstructed from accelerometry and photometry [23]. A rocket-borne flight test of terrain relative navigation with a Doppler lidar gave truth against DGPS for landing navigation rather than impact loads [29].

Small-body and low-gravity impact. Regolith response to slow contact changes with gravity level, as parabolic, shuttle and suborbital experiments showed [25]. Granular force laws fitted in laboratory gravity were extended numerically to a 1300 kg spacecraft pressing on asteroid regolith at about 10 cm/s [24], and a large crater from the Hayabusa2 impactor was reproduced only by simulation, since the regime cannot be reached on Earth [26]. A shear and compression cell built for drop tower falls measures regolith strength at small-body, lunar and Titan accelerations [27]. The Bremen drop tower supplies the free-fall time those experiments use [11]. The SLIM lunar lander’s results, with crushable porous metal shock absorbers, are reviewed by its project lead [28]. None of these reproduce an impact of a lander-class article on a hard surface at Earth gravity, which the gantry does.

Gravity offload for crew and rovers. The Langley suspension relieved five sixths of weight on a vehicle in powered descent [1]. Later offload systems target walking and manipulation instead. A pneumatic constant-force rig held reference lift to within 2.2 to 10 percent of command, degrading with the dynamic content of motion [12]. ARGOS offloads a suited subject with three active translational axes, and its fidelity depends on a gimbal pivot placed within a few centimeters of the combined center of gravity [13][14]. A student overhead trolley tracks a rover-mounted infrared beacon over a simulant sandbox [15].

Historical environment simulation. The Apollo-era vacuum chambers at Johnson Space Center, documented in a monograph, were built for thermal and vacuum work and not for landing dynamics [31].

What is not established

NASA Langley closed the Impact Dynamics Research Facility in September 2003 with demolition planned for 2007 [1]. The later Orion campaigns and the operator’s 2022 page show the gantry in use under the LandIR name [3][5][6]; the corpus holds no document describing the interval or what was refurbished, so whether the pre-2003 cable systems and instrumentation are the current ones is not established.

The gantry load figure differs by source: 64,000 lb (284.7 kN) in the 2009 Orion correlation paper [4] and 65,000 lb in the operator’s page [6]. Neither states the largest article that can be dropped under the swing arrangement, and the operator’s page gives a maximum drop height of about 200 ft without an achievable velocity, release-condition accuracy or calibration statement for its listed instruments [6].

The soil is characterized only as a sand and clay mixture standing in for a dry lakebed in the Orion airbag series [5]; its density, depth and reconditioning between drops are not given. The soil-vehicle friction coefficient of 0.45 is back-calculated by matching simulated to measured stopping distance, so it absorbs other modeling error [4]. The photogrammetric uncertainty (angle bias 0.06 degrees, 2-sigma precision 0.006 degrees) is derived from target pixel geometry and fit residuals rather than an independent reference [5]. No water basin wave, current or temperature capability is published [6].