Investigating the Dust Flux in the Meteoric Smoke Sampler (MESS) Instrument for Sampling Dust in the Mesosphere

We report and discuss the design of a rocket instrument to collect mesospheric dust particles that are composed of ice and include smaller refractory meteoric smoke particles (MSP). We expect that the ice components melt and that MSP are collected. The instrument consists of a collection device with an opening and closure mechanism and an attached conic funnel. Attaching the funnel increases the sampling area in comparison to the collection area which is kept small since this determines the size of the closure device which is a critical component to be designed for sea recovery. The instrument will collect primary 5 particles that directly hit the collection area and secondary particles that form from mesospheric dust hitting the funnel. We simulate the entry and impact of dust onto the detector considering their trajectories in the airflow and the fragmentation at the funnel. We estimate the collection efficiency of the instrument and the impact energy of particles at the collecting area. The design considered has a sampling area of 5 cm diameter and a collection area of 1.8 cm diameter. To estimate the expected amount of collected dust we assume collection during rocket flight through a 0.5 to 4 km dust layer with dust number densities 10 and dust sizes at 85 km as derived from lidar observations (Kiliani et al. 2015). Assuming the collected particles contain 3% volume fraction of MSP, we find that the instrument would collect of the order of 10 to 10 amu of refractory MSP particles. The estimate basis on the assumption that the ice components are melting and the flow conditions in the instruments are for typical atmospheric pressures at 85 km. Copyright statement. TEXT 15 1 https://doi.org/10.5194/amt-2020-278 Preprint. Discussion started: 28 July 2020 c © Author(s) 2020. CC BY 4.0 License.

The paper investigates the movement of particles in a conceptual instrument for collecting dust particles during a sounding rocket flight and later sea retrieval. This is accomplished by combining several models and simulations. For a future deployment, it is important to understand how dust moves from the atmosphere into the instrument and onto the collecting surface.
General comments: At the present stage of development, the investigation is clearly aiming to find the boundaries of the design, as no closing mechanism or collecting surface is defined. It would be nice to elaborate this more clearly, e.g. which collection principles exists and C1 what are their requirements. What would be other requirements or degrees freedom, e.g. from the rocket and environment?
Further I would recommend focusing less on the work that has clearly been done, but more on the meaning of it. For example in Fig. 2, the shock sure looks nice but what is it that you want to clarify to the reader, especially as the Mach number is not further discussed? Or the different trajectories in Fig. 5 & 6, what should the reader see in those figures? Even if crude, it would be more helpful to give e.g. the collection efficiency, in a table or figure, to see more clearly which altitude and velocity is preferred.
Why are only 80 and 85 km simulated? The reader is forced through half the paper before knowing in the results section that PMSE are limited to those altitudes. In Rapp and Lübken (2004) the altitude range is given with 80 to 90 km for PMSE with a clear peak at 85 km, while NLCs (large ice particles) peaking between 80 and 85 km. Thus particle size is a function of altitude, with the heaviest being lowest. This was not considered in the present paper and it feels like 90 km is missing in the simulations, especially if one could assume different particle sizes at different altitudes, which would also lead to different ratios of primary and secondary particles.
In 4.2.1 it is stated that primary particles (not colliding with funnel) are simulated, but in Figures 5 & 6 plenty of particles hit the funnel? It is further not clear, from the figures if they reach the surface or if they just move out of the plane? As the pressure regime is within the Knudsen flow (if i am not mistaken), particle trajectories should have more of a statistical outcome? 8 or 9 primary particle trajectories could be not enough?
In the results section a lot of work seems to be swept away by assuming a collection efficiency of unity and calculate the total amount of particles when flying a known collection area through a layer of an assumed density and then vary layer thickness and collection efficiency without taking into account the simulation results or other constraints.

C2
Usually the assumption of an angle of attack = 0 • is always wrong, as most rockets do not have attitude control. Maybe this could be more reasoned as insignificant for typical angles of attack in the given scenarios. A slower rocket at higher altitude as proposed might show significantly higher angles of attack.
If best results are obtained for lower pressures, could there be a more optimised shape of the funnel? Line 232 formatting of citation Line 238 why not make it larger for even more particles? why is it a reasonable funnel size or aspect ratio (diameter / funnel length)? Why not sample as much as possible? e.g. 80 to 90 km Line 243 the energies of the particles increase with the square of velocity, why is the number density the dominating factor and why does this not just increase or decrease a probability, e.g. the collection efficiency via number of air molecule collisions?