Structural Constraints on Entropy-Gradient Gravity from Cluster Merger Geometry
The Bullet Cluster has become a stress test for any gravity theory that tries to work without particle dark matter. Its importance is not just that “mass is missing,” but that the geometry of mass and gas is split apart. The hot X-ray gas — most of the normal matter — lags behind after the collision, while the gravitational lensing signal peaks near the galaxies. The separation is on the order of 150 kiloparsecs. Any viable alternative to dark matter must reproduce that spatial structure, not merely match total mass.
This new constraint study asks a narrow but decisive question: Can entropy-gradient gravity (within the Energy-Flow Cosmology framework) place lensing mass where the galaxies are, rather than where the gas is? The goal is not to validate the theory, but to eliminate versions that fail on geometry alone.
Starting from first principles
Two physical facts drive the test:
- Gravitational lensing traces the effective gravitational field, not just visible matter.
- Gas and galaxies behave differently in collisions
- Gas is collisional, shock-heated, and mixes.
- Galaxies are mostly collisionless and pass through.
If gravity responds only to local baryonic density gradients, then it will naturally highlight sharp gas features — shock fronts and edges. But observations show the opposite: the strongest lensing sits with the galaxies, not the shocked gas cloud. That tension becomes the core diagnostic.
The first model: Local gradient coupling (falsified)
The simplest entropy-gradient idea assumes gravity strengthens wherever the baryonic density changes rapidly — mathematically, where the gradient of density is large. This sounds reasonable: structure forms where gradients exist.
But gradients are strongest at edges, not at centers. In a cluster, that means the model enhances gravity in rings or shells around gas cores, not at galaxy concentrations.
When tested across 42 parameter combinations, this local formulation failed every time. It consistently placed lensing peaks hundreds of kiloparsecs away from the observed galaxy positions. The mechanism itself guarantees the wrong geometry.
Conclusion: A purely local entropy-gradient coupling cannot explain cluster mergers. This class of models is ruled out at this scale.
This is not a tweakable mismatch — it is a structural failure. The math of the operator forces the wrong outcome.
The second model: Non-local, component-sensitive response
The next step relaxes one assumption: gravity may respond non-locally and may weight collisionless matter more strongly than shocked gas.
Physically, this reflects entropy flow. Shocked gas loses fine structure through mixing, while stellar distributions preserve it. If gravity amplifies preserved structure more than erased structure, the galaxies can dominate the effective mass map.
This version introduces:
- A non-local smoothing scale (L_0), on the order of a cluster core (~200 kpc)
- A component weighting (w), giving galaxies stronger influence than gas
Under these conditions, some parameter sets do reproduce the observed separation. But there is a critical caveat: the model preserves a geometry already present in the stellar distribution — it does not yet predict that geometry from first principles.
So the result is conditional:
Local models are eliminated. Non-local, component-aware models survive — but only as constrained possibilities.
From fitting to prediction: the w(M, t) model
To move beyond ad-hoc tuning, the study introduces a minimal physical model for the weighting factor (w). It links the strength of the collisionless component to:
- The Mach number of the merger shock (compression physics)
- The mixing time of the gas
- A single global parameter (η) that controls entropy mixing efficiency
The key design choice is a saturating time dependence, not an exponential one. That makes predictions less sensitive to uncertain merger ages — a practical requirement if the theory is to be testable.
Using the Bullet Cluster only to fix this global parameter, the model then generates a priori predictions for two other mergers: MACS J0025 and Abell 520. No per-cluster retuning is allowed. These predictions can be tested directly against real lensing reconstructions.
What this study does and does not claim
What it establishes
- A clean falsification of local entropy-gradient gravity at cluster scales
- The necessity of non-local response in any viable entropy-based gravity model
- A physically motivated, minimally parameterized predictive framework
What it does not establish
- That entropy-gradient gravity is correct
- That the stellar–gas weighting has a fully independent derivation
- That the predictions match real shear-catalog lensing data (synthetic reconstructions were used here)
In other words, this is a theory-space constraint, not an observational confirmation.
Why this matters
The Bullet Cluster is often presented as a knockout blow to modified gravity. This work reframes it more precisely: it is a geometric filter. It doesn’t say alternative gravity is impossible — it says most naive versions are.
That is real scientific progress. Instead of arguing in slogans (“Bullet Cluster proves dark matter”), we now have quantitative structural requirements. Any entropy-gradient gravity theory that survives must be:
- Non-local on cluster scales
- Sensitive to the dynamical state of matter
- Predictive across multiple mergers with locked parameters
The next step is decisive: apply these locked predictions to real lensing datasets. Either the geometry matches — or the surviving version fails too.
That’s how a theory grows up: not by fitting everything, but by surviving the places it should break.
DOI: https://doi.org/10.6084/m9.figshare.31173850
About the Author
