SPACE: Suborbital Particle Aggregation Experiment

• SPACE is a microgravity laboratory campaign that examines submillimeter, porous dust aggregate collisions using controlled agitation and high-speed imaging.
• It employs precise velocity control (0.1–5 cm s⁻¹) via a DC motor system to simulate early planetesimal formation conditions.
• Empirical data on collision regimes and energy thresholds now refine dust aggregation models, improving predictions of mm-scale aggregate growth.

The Suborbital Particle Aggregation and Collision Experiment (SPACE) is a microgravity laboratory campaign designed to characterize the collision properties of submillimeter-sized, highly porous dust aggregates under conditions analogous to the early stages of planet formation. Utilizing the REXUS 12 suborbital rocket, SPACE enabled the controlled investigation of dust–dust interactions at ultra-low relative velocities (≲1 cm s⁻¹), directly addressing critical gaps in the empirical data underpinning planetesimal growth models. Through a combination of carefully engineered vacuum-encapsulated agitation, high-speed imaging, and automated data acquisition, SPACE provided the first statistically robust measurements of collisional regimes—sticking, bouncing, and fragmentation—in the size and velocity domains most relevant for the transition from micrometer-scale grains to macroscopic planetary precursors (Brisset et al., 2013).

1. Experimental Infrastructure and Flight Environment

SPACE was flown aboard the REXUS 12 rocket, a 5.6-m-long, 0.356-m-diameter vehicle employing spin stabilization (20 g peak), subsequently derotated to a residual 11°/s via yo-yo despin. The experiment exploited the microgravity window between 70 s and 220 s post-launch, corresponding to atmospheric drag <10⁻⁴ g and a sustained centrifugal acceleration of ~1.45×10⁻³ g. Approximately 150 s of reduced-gravity enabled hundreds of dust–dust collisions at precisely calibrated, gentle velocities.

The hardware consisted of three Borofloat 33 glass containers: two cells of 11×10×15 mm³, one of 24×10×15 mm³; certain internal walls were aluminum, the others coated anti-adhesively to minimize dust wall-stick. Dust samples were retained by narrow slits (allowing gas escape during evacuation) and a turbopump umbilical achieved pre-launch vacuum levels of p < 10⁻⁴ mbar.

2. Dust Samples and Agitation Protocol

Three types of SiO₂ aggregates were investigated:

• Monodisperse spheres (mean diameter 180 μm) in the large cell.
• Monodisperse spheres (370 μm) and polydisperse, irregular grains (mean 370 μm) in the small cells. Bulk porosity was φ≈0.35 with constituent monomer size ≈1 μm.

A DC motor and cogwheel system generated two-dimensional circular motion (amplitude A ≈ a few millimeters, offset ≈40 mm from rocket axis). The shaker was modulated through three voltage phases, controlling the frequency f and thus the induced aggregate velocities according to:

$v_{\rm rel} \approx 2 \pi f A$

Motion phases included a “shake-up” (full voltage: 5–10 s), reduced-voltage intervals (20–50%, lasting 10–25 s), and rapid ramps to higher voltages (up to 125%) to explore aggregation versus fragmentation thresholds.

3. Collision Detection and Analysis Methods

High-speed imaging was performed with a Prosilica GE680 camera at 170 frames per second (640×480 resolution, 8-bit grayscale), with back-lighting from 86 diffused blue LEDs and mirrored optics to compact the imaging path. Frames were streamed in Jumbo Ethernet packets to an onboard embedded computer, storing data on a 16 GB CompactFlash.

Raw images underwent correction for background non-uniformity and cell rotation. Aggregates were autonomously tracked across frames, permitting velocity extraction before and after collision. The primary observables included:

• Coefficient of restitution:

$e = \frac{v_{\rm after}}{v_{\rm before}}$

• Collisional outcome classification (sticking, bouncing, fragmentation).
• Cluster formation tracked via boxcar averaging over 200 frames to reveal mass accretion rates.

4. Empirical Findings: Collision Regimes, Thresholds, and Key Parameters

The relative velocities sampled the 0.1–5 cm s⁻¹ range, peaking near 1–2 cm s⁻¹. Quantitative regime boundaries and statistical outcomes were mapped as follows:

• Sticking dominated for $v_{\rm rel} < 1.0$ cm s⁻¹, with sticking probability $P_{\rm stick}(v) \approx 1$ for $v_{\rm rel} < 0.5$ cm s⁻¹ and dropping to zero at 2–3 cm s⁻¹.
• Between 1–3 cm s⁻¹, bouncing was the predominant outcome. The coefficient of restitution $e$ decreased from ≈0.7 at $v_{\rm rel} ≈ 1$ cm s⁻¹ to ≈0.4 at 3 cm s⁻¹.
• Fragmentation became measurable above ≈3 cm s⁻¹, with the fragmentation threshold set by the specific energy

$Q^* \approx \frac{1}{2}\,v_{\rm frag}^2$

For $v_{\rm frag} ≈ 3$ cm s⁻¹, $Q^* ≈ 4.5 \times 10^{-5}$ J kg⁻¹.

Collisional cross section for two equal-radius $e = \frac{v_{\rm after}}{v_{\rm before}}$0 aggregates was given by

$e = \frac{v_{\rm after}}{v_{\rm before}}$1

These data establish, for the first time, the full continuum from sticking through bouncing to fragmentation in a monodisperse and polydisperse aggregate ensemble under true microgravity.

5. Integration with Dust Growth and Collision Modeling

SPACE results directly address empirical uncertainties in contemporary dust aggregation models. In particular:

• Previously extrapolated sticking thresholds are now replaced with empirical values: the onset of bouncing occurs at $e = \frac{v_{\rm after}}{v_{\rm before}}$2 cm s⁻¹.
• The experimentally determined restitution-velocity relationship $e = \frac{v_{\rm after}}{v_{\rm before}}$3 informs the modeling of bouncing-induced compaction, which strongly influences aggregate porosity evolution.
• The measured fragmentation energy $e = \frac{v_{\rm after}}{v_{\rm before}}$4 sets the kinetic energy boundary for net mass loss during high-velocity encounters.

Compared to the G{\"u}ttler et al. (2010) collision-outcome map, SPACE populates the “sticking–bouncing transition” regime with direct measurements ($e = \frac{v_{\rm after}}{v_{\rm before}}$5, $e = \frac{v_{\rm after}}{v_{\rm before}}$6, $e = \frac{v_{\rm after}}{v_{\rm before}}$7, $e = \frac{v_{\rm after}}{v_{\rm before}}$8). Updates to Monte Carlo growth simulations (e.g., Zsom et al. 2010) are warranted by these findings; the maximal aggregate size before reaching the bouncing barrier can now be more accurately constrained. A plausible implication is a somewhat larger sticking window than previously assumed, potentially enabling more efficient mm-scale aggregate formation and improving the fidelity of planetesimal growth predictions.

6. Broader Significance and Future Directions

SPACE provides the first statistically robust microgravity dataset on submillimeter, highly porous dust aggregate collisions in the 0.1–5 cm s⁻¹ domain (Brisset et al., 2013). The comprehensive hardware design, calibrated agitation, and high-speed, high-fidelity imaging stand as a reference for subsequent experiments in protoplanetary dust physics. The refined collisional kernels contributed by SPACE serve to bridge the knowledge gap between micrometer-scale grain coagulation and the formation pathways of macroscopic planetesimals.

Continued development of microgravity platforms and even more precise control of collision parameters is expected to further constrain the physical mechanisms underlying early planetary growth, as well as elucidate the critical roles of aggregate composition, porosity, and morphology across the full spectrum of pre-planetesimal evolution.