Smoothed Particle Hydrodynamics in cosmology: a comparative study of implementations
Bibliographic record
Abstract
We analyse the performance of 12 different implementations of Smoothed Particle Hydrodynamics (SPH) using seven tests designed to isolate key hydrodynamic elements of cosmological simulations which are known to cause the SPH algorithm problems. In order, we consider a shock tube, spherical adiabatic collapse, cooling flow model, drag, a cosmological simulation, rotating cloud-collapse and angular momentum transport. In the implementations special attention is given to the way in which force symmetry is enforced in the equations of motion. We study in detail how the hydrodynamics are affected by different implementations of the artificial viscosity including those with a shear-correction modification. We present an improved first-order smoothing-length update algorithm that is designed to remove instabilities that are present in simple forward prediction algorithms. Gravity is calculated using the adaptive particle–particle, particle–mesh algorithm. For all tests we find that the artificial viscosity is the single most important factor distinguishing the results from the various implementations. The shock tube and adiabatic collapse problems show that the artificial viscosity used in the hydra code prior to version 4.0 performs relatively poorly for simulations involving strong shocks when compared to a more standard artificial viscosity. The shear-correction term is shown to reduce the shock-capturing ability of the algorithm and to lead to a spurious increase in angular momentum in the rotating cloud-collapse problem. For the disc stability test, the shear-corrected and previous hydra artificial viscosities are shown to reduce outward angular momentum transport. The cosmological simulations produce comparatively similar results, with the fraction of gas in the hot and cold phases varying by less than 10 per cent amongst the versions. Similarly, the drag test shows little systematic variation amongst versions. The cooling flow tests show that implementations using the force symmetrization of Thomas & Couchman are more prone to accelerate the overcooling instability of SPH, although the problem is generic to SPH. The second most important factor in code performance is the way force symmetry is achieved in the equation of motion. Most results favour a kernel symmetrization approach. The exact method by which SPH pressure forces are included in the equation of motion appears to have comparatively little effect on the results. Combining the equation of motion presented by Thomas & Couchman with a modification of the Monaghan & Gingold artificial viscosity leads to an SPH scheme that is both fast and reliable.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.017 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".