University of Central Florida Exolith Lab

Source: [2]. CC BY 4.0.

The Exolith Lab in Orlando, Florida is a regolith research and space hardware testing facility managed by the Florida Space Institute at the University of Central Florida, operating within the Center for Lunar and Asteroid Surface Science, a NASA SSERVI node [1]. Its output is material rather than test time: it develops, produces and characterizes lunar, Martian and asteroid regolith simulants, publishing the recipes, production methods and characterization data as open standards so that other groups can reproduce the material [3]. Standard simulants are produced in bulk and are readily available for purchase and shipment [4]. Testing hardware against a manufactured lunar soil analog is not new; Boeing ran the Apollo-era lunar roving vehicle wheels through a simulant decades before any lab published a reproducible recipe for one [11]. What distinguishes Exolith from that earlier, single-purpose material is that the recipe, production method and characterization data are published as an open, purchasable standard rather than consumed inside one program.

Laboratory data

Parameter	Value
Operator	University of Central Florida, Florida Space Institute, within CLASS, a NASA SSERVI node [1]
Location	Orlando, Florida, United States
Commissioned	Not published. MGS-1 production capacity under construction 2019
Type	Simulant production plant and characterization laboratory, not a test chamber [1][3]
Floor area	Not published
Capabilities	Production, characterization, shear box
Simulant or terrain	LHS-1, LMS-1, LHS-1D, LMS-1D, MGS-1, MGS-1C, MGS-1S, JEZ-1, CI-1 [1]
Instrumentation	Fused bead XRF, XRD with corundum standard, CILAS 1190 particle size analyzer [2]
Ground truth	Composition referenced to Apollo samples and to Curiosity XRD of Rocknest [3]
Fidelity limits	Terrestrial feedstock: Mg and Al enriched, Fe poor; no trace elements [2]. See below
Access	Standard simulants offered in bulk for purchase and shipment; retail listing published [4]
Cited by	regolith-simulants

Capabilities

Production line

Parameter	Value
Working volume	Batch, not chamber-limited. High-volume production equipment, standard simulants offered in bulk [4]
Test article limits	MGS-1 capacity tens to hundreds of kilograms at publication [3]
Vacuum	Not applicable. Ambient pressure
Temperature	Process only: 1000 C graphite fusion; 900 C loss on ignition; 140 C bake [2][5]
Illumination	Not applicable
Simulant or terrain	Anorthosite, glass-rich basalt, pyroxenite, olivine, ilmenite feedstocks
Slope	Not applicable
Gravity offload	Not applicable
Instrumentation	Jaw crusher, hammer mill, ball mills, ASTM sieve shakers, cement mixers

Simulant design is stated as a constrained maximization of fidelity against cost, safety and material availability, with mineralogy and particle geometry as the tuned parameters [2]. Mineralogy of the target body is estimated first from returned samples and remote sensing; terrestrial rocks and minerals with well characterized sources are then found that either arrive pre-processed to a grain size range or are crushed in house. Feedstocks are bought in bulk both to avoid downtime and to raise product consistency, and where possible the lab uses materials already established in other simulants: Merriam crater basalt, as used in JSC-1, and Greenspar anorthosite.

Size reduction is percussive. A hammer mill and a jaw crusher are used specifically to produce irregular, jagged grains resembling space-weathered regolith, and ASTM standard sieves with sieve shakers size sort the output. When maximum particle sizes below about 50 micrometers are wanted, ball mills reduce the material further; the dust grades LHS-1D and LMS-1D are made by ball milling the parent LHS-1 and LMS-1, and constituents are then combined to within less than 0.5 wt% difference between production runs and mixed in cement mixers [2].

The product description for LHS-1 states the design intent in the same terms: the simulant is not made from a single terrestrial lithology but combines mineral and rock fragments as polymineralic grains in proportions referenced to returned Apollo regolith [4]. Free consulting on simulant selection, dust mitigation and other ISRU topics is offered alongside the material itself [1].

The Martian line uses a different route to grain structure. Rather than mixing dry powders, plagioclase, pyroxene, olivine, basaltic glass, magnetite and hematite are combined with water and sodium metasilicate pentahydrate in a 100:20:2 ratio by weight, fused into polymineralic cobbles, and mechanically ground to a power law size distribution, so that the product is made of polymineralic grains as eroded basalt is [3]. Most MGS-1 prototypes were ground to below 6.3 mm.

Characterization

Parameter	Value
Working volume	Sample scale
Instrumentation	Thermo ARL Perform’X XRF; PIXcel3D XRD; CILAS 1190 particle size analyzer
Simulant or terrain	Every constituent and every mixed product

Bulk elemental composition is measured by fused bead XRF at the Hamilton Analytical Laboratory, Hamilton College, on a Thermo ARL Perform’X spectrometer at 45 kV and 45 mA, using a low dilution graphite fusion at 1000 C with a 2:1 flux to powder ratio, doubly fused, surface finished to 15 micrometers and cleaned in ethanol, with loss on ignition measured by 15 to 17 h at 900 C, and mineral phases measured by XRD with a PIXcel3D detector scanning 2 theta from 5 to 90 degrees at 0.3 degrees per second over five iterative rotations, samples ball milled and spiked with 10 to 15 percent corundum as an internal standard for amorphous content; particle size is measured on a CILAS 1190 volumetric analyzer over 0.04 to 2500 micrometers, liquid dispersion in deionized water with 640 and 830 nm lasers, three independent samples per simulant with 2 sigma confidence intervals [2]. An earlier lab briefing had already reported modal mineralogy and XRF oxide chemistry for LHS-1 and LMS-1 by the same methods [7].

Measured D50 is 59.79 micrometers for LHS-1, 72.27 for LMS-1, 9.74 for LHS-1D, 5.92 for LMS-1D, 49.30 for MGS-1, 15.50 for MGS-1C, 63.13 for MGS-1S and 46.92 for JEZ-1 [2]. LHS-1 is 74.4 wt% anorthosite, 24.7 wt% glass-rich basalt, 0.4 wt% ilmenite, 0.3 wt% pyroxenite and 0.2 wt% olivine; LMS-1 is 32.8 wt% pyroxenite, 32.0 wt% glass-rich basalt, 19.8 wt% anorthosite, 11.1 wt% olivine and 4.3 wt% ilmenite [2].

Direct shear box

Parameter	Value
Working volume	Shearing surface 10.17 x 10.17 cm, depth 6 cm
Test article limits	Normal stress range of interest 0.097 to 0.68 kPa
Temperature	Samples baked 24 h at 140 C and transferred hot where adsorbed water is the variable
Instrumentation	Actuonix L16-R servo under Arduino control; HP-500 force gauge, 0.01 N

Source: [5].

A custom polycarbonate direct shear box built at UCF, used rather than commercial geotechnical equipment because the normal stress range of interest is not reachable with commercial apparatus. The lower half of the box rides a rail driven by the linear servo against the force gauge, normal stress is applied by calibrated aluminum sheets, and density is set by sprinkling to an uncompressed state and then vibrating at 20 to 40 Hz in pulses [5]. Measured internal friction angle varies non-monotonically with density: MGS-1C peaks near 29 degrees at 1.14 g/cm3 and falls to about 25 degrees at 1.35 g/cm3, while MGS-1S rises from about 34 degrees at 1.31 g/cm3 to near 53 degrees at 1.83 g/cm3.

Product line and pricing

The standard line is LHS-1 lunar highlands, LMS-1 lunar mare, LHS-1D and LMS-1D dust grades below 25 micrometers, LHS-1-25A with 25 wt% sintered anorthosite-iron agglutinates, MGS-1 Mars global basaltic, MGS-1C clay-enhanced, MGS-1S sulfate-enhanced, JEZ-1 Jezero crater and CI-1 carbonaceous asteroid, with Phobos and CR simulants made on request [1]. The Phobos line, PCA-1 for the captured-asteroid origin hypothesis and PGI-1 for the giant-impact hypothesis, was developed and characterized separately, each mixed to a different candidate parent composition rather than to one reference body [8]. LHS-1 lists at 45 USD per kilogram in the 1 kg retail size [4]; an earlier 2020 snapshot priced LHS-1 and LMS-1 at 25 to 30 USD per kilogram and the sub-25-micrometer dust grades at 55 to 60 USD per kilogram, with bulk quantities to tens of tons available [7]. Free consulting on simulants, dust mitigation and related in-situ resource utilization topics is offered. No bulk price list is published.

Instrumentation

Source: [2]. CC BY 4.0.

Measured particle size distributions for the lunar line give D50 of 59.79 micrometers for LHS-1 and 72.27 micrometers for LMS-1, falling to 9.74 and 5.92 micrometers for the LHS-1D and LMS-1D dust grades, with distribution spans of 3.26, 3.83, 1.97 and 2.45 and uniformity coefficients of 10.20, 16.38, 6.45 and 7.39 [2]. The Martian line gives D50 of 49.30 micrometers for MGS-1, 15.50 for MGS-1C, 63.13 for MGS-1S and 46.92 for JEZ-1, with MGS-1C and MGS-1S multi-modal because the added weathered and altered phases occupy a distinct 2 to 11 micrometer fraction while the non-crystalline silicates sit at 10 to 1000 micrometers.

Geotechnical characterization of the products has been done on a custom polycarbonate direct shear box built at UCF rather than on commercial geotechnical equipment, because the normal stress range of interest, 0.097 to 0.68 kPa, was not reachable with commercial apparatus [5]. The box has a 10.17 cm by 10.17 cm shearing surface and a 6 cm depth, the lower half riding a rail driven by an Actuonix L16-R linear servo under Arduino control against an HP-500 force gauge of 0.01 N stated resolution, with normal stress applied by calibrated aluminum sheets and density set by sprinkling to an uncompressed state and then vibrating at 20 to 40 Hz in pulses [5].

What it does not reproduce

The material is made of terrestrial rock, and the lab’s own characterization paper is explicit that the exact composition of planetary regolith is impossible to recreate perfectly from it, because terrestrial feedstocks are generally enriched in Mg and Al and carry less Fe than the regolith being simulated; the differences have to be quantified and accounted for by the user rather than designed away [2]. The lab’s own public accounting of what is and is not simulated draws the same line: major mineralogy, bulk major element chemistry with the stated Mg, Na and K excess and Fe and Ca deficiency, particle size distribution and total water release are targeted properties, while particle shape, trace elements and isotopes, and reactivity toward an atmosphere the material has already equilibrated with are not attempted [6].

Trace elements are the clearest case. The trace element profile of a simulant is inherently that of a terrestrial material, since the simulant is made solely from materials found on Earth and terrestrial rock-forming processes differ completely from those on the target body, so the lab states that a simulant should never be correlated to actual planetary regolith in trace element analyses [2].

Mineral substitution is the second. For the asteroid simulants the Fe-rich cronstedtite and tochilinite of the target material are not commercially available, so Mg-rich, non-asbestiform serpentines are used instead alongside iron powder for the Fe budget. The resulting Fe depletion and Mg enrichment is described by the lab as not ideal but the best constrained approximation available while keeping the product safe, available and affordable [2]. A Figure of Merit analysis of the asteroid line against the Orgueil CI meteorite quantifies what that substitution costs: the mineralogical component of fidelity is the one most degraded by the swap, so a test sensitive to mineralogy rather than to grain size or cobble strength recovers less of its value from the simulant than a test that is not [9].

Three further limits recur in the technical literature.

Feedstock drift is acknowledged but not quantified. Material source location, supplier and equipment all change over time, and such changes alter chemical composition, processing requirements that affect particle shape and size distribution, and grain density. The lab states that quantitative analysis of variation in simulant mineralogy and particle size distribution through time is not performed, and that the material and source used in a given simulant can only be recovered by contacting the lab with the purchase date. The switch to Merriam crater basalt was itself a recent change at the time the 2023 paper was written [2].

Measurement of the delivered product has its own limits. XRD patterns are best obtained from material of uniform particle size, and Exolith simulants span at least four orders of magnitude in particle size, which lowers peak intensity and can make phase identification difficult even with proper sample preparation. The asteroid cobbles are bound with a water-soluble agent and are relatively weak, so published asteroid particle size distributions describe only the loose powder forms [2].

The Martian standard is bounded by what is known about Mars. Geotechnical properties of actual Martian regolith are poorly constrained relative to returned lunar regolith, particle shape was not controlled in the MGS-1 prototypes, and the Martian regolith particle size distribution is itself poorly known because of camera resolution limits, with estimates of the clay-sized fraction ranging from about 1 vol% below 4 micrometers at the Phoenix site to 15 to 25 wt% of typical soils [3]. Crystalline nitrate and perchlorate salts were included in some early prototypes made for agricultural studies but are excluded from the root MGS-1 standard.

Adsorbed terrestrial water changes the mechanics. Direct shear tests on LHS-1D found that removing roughly 0.4 wt% of adsorbed atmospheric water, by baking for 24 hours at 140 C and transferring the sample to the shear box hot, changed cohesion non-monotonically with density and normal stress: at midrange densities that 0.4 wt% of water raised cohesion by about 15 percent, while above about 0.4 kPa normal stress the intrinsic shear strength dominated the water effect [5]. The material is synthesized and stored in the terrestrial atmosphere, so that adsorbed water is present unless a user drives it off deliberately.

A geotechnical properties database drawing on decades of mission reports and simulant papers, including Exolith’s own, finds bulk density, friction angle and cohesion values from this lab sitting inside the wider spread already measured across other lunar simulants and returned regolith, rather than marking out a distinct range of its own [10].

Campaigns run there

MGS-1 standard definition, 2019. The Rocknest windblown soil at Gale crater was taken as the reference and its crystalline phases, constrained by Curiosity XRD, were reproduced mineral by mineral, with the amorphous component fitted by a least squares combination of basaltic glass, ferric sulfate, ferrihydrite and iron carbonate against the estimated composition of the Rocknest amorphous material [3]. The paper explicitly rejects the practice of matching bulk chemistry with arbitrary compounds, and recommends against MMS-1 and MMS-2 on the grounds of missing documentation and discrepancies against the original Mojave Mars Simulant [3].

Constituent and product characterization, 2020 to 2023. The full standard line was characterized by fused bead XRF, XRD with corundum internal standard and CILAS 1190 particle size analysis, and published with the production methods as an open standard [2].

Cohesion and shear strength survey, 2024. Direct shear measurements to ASTM D3080 were run on the Exolith lunar and Martian lines [5]. Cohesion rises exponentially with bulk density, and the exponential growth constant rises linearly with the full width at half maximum of the particle size distribution. Martian simulant cohesion varied between 0.181 and 0.349 kPa over the achievable density range of about 1.40 to 1.60 g/cm3 for some products [5]. Internal friction angle did not vary monotonically with density: MGS-1C peaks near 29 degrees at 1.14 g/cm3 and falls to about 25 degrees at 1.35 g/cm3, while MGS-1S rises from about 34 degrees at 1.31 g/cm3 to near 53 degrees at 1.83 g/cm3. The authors record that all of it applies to terrestrial atmosphere and gravity at normal stresses below 0.68 kPa, and that repeating the measurements in a vacuum chamber was already under way [5].

References