Gravitational Tension Screening

• Effective gravitational tension screening is a family of mechanisms that dynamically suppress, renormalize, or saturate gravitational responses via scalar adjustments, coupling variations, or geometric constructions.
• It includes local scalar-tensor strategies such as chameleon, symmetron, and kinetic screening that modify effective Newton constants to preserve GR in dense or oscillatory regimes.
• The concept extends to cosmological and gravitational-wave contexts where screening affects lensing, clustering, and waveform propagation to reconcile observations with modified gravity theories.

Effective gravitational tension screening denotes a family of mechanisms in which the gravitational response associated with matter, scalar hair, vacuum energy, bulk curvature, or wave propagation is dynamically suppressed, renormalized, or saturated. Across the literature surveyed here, the exact phrase is generally not the standard label; instead, the field uses more specific terms such as screening mechanism, kinetic screening or $k$-mouflage, weakening gravity, over-screening, screening bulk curvature, critical coupling surfaces, and, in one recent geometric construction, gravitational tension screening (Brax et al., 2022, Muñoz, 2024, Agarwal et al., 2011, Estrada, 1 Jun 2026). What unifies these uses is not a single formalism, but a common objective: preserving observationally acceptable gravity in one regime while allowing modified behavior in another.

1. Conceptual scope and terminology

In the narrow scalar-tensor sense, screening usually means suppression of a scalar fifth force near matter sources. The suppression can occur because the scalar becomes heavy, because its coupling to matter becomes small, or because nonlinear kinetic terms increase the effective normalization of scalar fluctuations and thereby reduce the sourced response (Brax et al., 2022). In this usage, the “screened” quantity is the scalar contribution to local gravity.

A broader usage concerns the effective gravitational coupling itself. In Brans-Dicke-like theories with nonminimal coupling $F(\varphi)R$, the screened quantity need not be only a fifth force; screening can also induce an environment-dependent Planck mass or Newton constant, with $F(\varphi)=M^2/A^2(\chi)$ and $G_{\rm eff}(\varphi)\sim 1/F(\varphi)$ (Muñoz, 2024). This shifts the focus from force suppression to an environment-dependent renormalization of gravitational strength.

A third usage is phenomenological rather than mechanistic. Large-scale analyses of $f\sigma_8$, $E_G$, strong lensing, and related probes sometimes describe departures from the GR/$\Lambda$CDM expectation as an effective weakening or screening of gravity, even when no specific local screening mechanism is established (Skara et al., 2019, Lian et al., 2022). Here the screened quantity is an inferred coupling such as $G_{\rm eff}$, $\mu$, $\Sigma$, or $F(\varphi)R$0.

Finally, some works use genuinely geometric language. In cascading gravity, scalar boundary terms can screen the effect of large brane tension on bulk curvature (Agarwal et al., 2011). In $F(\varphi)R$1 gravity, hypersurfaces where $F(\varphi)R$2 are interpreted as gravitational screening surfaces between attractive and repulsive phases (Teruel, 10 Jun 2026). In black-bounce regularization, a vacuum tidal “gravitational tension” is explicitly screened by a Schwinger-like saturation factor (Estrada, 1 Jun 2026). These constructions are the closest to a literal interpretation of “gravitational tension screening.”

2. Local scalar-tensor mechanisms and effective coupling renormalization

From the effective-field-theory viewpoint, screening can be organized by how the scalar modifies the local effective Newton coupling. In the weak-field limit, the review literature writes

$F(\varphi)R$3

or, with a finite scalar mass,

$F(\varphi)R$4

so suppression can arise through large $F(\varphi)R$5, small $F(\varphi)R$6, or large $F(\varphi)R$7 (Brax et al., 2022). Chameleon, symmetron, Damour-Polyakov, K-mouflage, and Vainshtein screening fit naturally into this classification.

The local theory can also support more than one screening regime. “Double screening” studies a shift-symmetric scalar EFT with both a cubic galileon operator and a $F(\varphi)R$8-type operator, giving a Vainshtein radius

$F(\varphi)R$9

and a kinetic-screening radius

$F(\varphi)=M^2/A^2(\chi)$0

Because these radii scale differently with source mass, the dominant screening mechanism depends on the source, and sufficiently massive objects can be kinetically screened on larger radii and Vainshtein screened further inside (Gratia et al., 2016).

Kinetic screening is developed explicitly for stars in $F(\varphi)=M^2/A^2(\chi)$1-essence. In the model

$F(\varphi)=M^2/A^2(\chi)$2

the paper focusing on $F(\varphi)=M^2/A^2(\chi)$3 identifies the screening radius

$F(\varphi)=M^2/A^2(\chi)$4

inside which nonlinear derivative terms suppress the scalar gradient and hence the fifth force (Haar et al., 2020). The local recovery of GR is therefore not achieved by removing the scalar globally, but by forcing the sourced solution into a nonlinear kinetic branch.

A different perspective is provided by the particle-level analysis of Brans-Dicke screening. There the Einstein- and Jordan-frame descriptions are treated as equivalent descriptions of an environment-dependent $F(\varphi)=M^2/A^2(\chi)$5 or $F(\varphi)=M^2/A^2(\chi)$6, with

$F(\varphi)=M^2/A^2(\chi)$7

The paper argues that only sectors whose masses arise from explicit scale-symmetry breaking contribute to screening, while sectors with masses generated by dynamical scale breaking can evade the coupling. It further introduces over-screening, defined as the regime in which the $F(\varphi)=M^2/A^2(\chi)$8-dependent mass rescaling becomes more significant than the screened fifth force itself (Muñoz, 2024). This makes screening a statement about effective gravitational response, not merely about scalar exchange.

The explicitly broken symmetron, or asymmetron, extends this logic to domain-dependent gravity. With

$F(\varphi)=M^2/A^2(\chi)$9

the effective potential becomes

$G_{\rm eff}(\varphi)\sim 1/F(\varphi)$0

and GR is restored for

$G_{\rm eff}(\varphi)\sim 1/F(\varphi)$1

At low density, however, the vacua are no longer degenerate, so domain walls can separate regions with different effective gravitational couplings (Perivolaropoulos et al., 2022). This shifts screening from a purely local property of dense environments to a domain-structured modification of low-density gravity.

Post-GW170817 DHOST models illustrate a more constrained version of the same theme. In the subclass satisfying $G_{\rm eff}(\varphi)\sim 1/F(\varphi)$2 and avoiding graviton decay, GR can be recovered outside matter only if the Lagrangian functions satisfy a tuning condition, while inside matter one still finds a different effective Newton constant and nonzero slip, with

$G_{\rm eff}(\varphi)\sim 1/F(\varphi)$3

Screening in this setting does not erase modified gravity inside matter; it leaves a renormalized gravitational response and a residual splitting of the metric potentials (Hirano et al., 2019).

3. Compact stars, nonlinear oscillations, and collapse

Static compact objects provide one of the clearest demonstrations that screening can be locally effective and yet dynamically limited. In $G_{\rm eff}(\varphi)\sim 1/F(\varphi)$4-mouflage $G_{\rm eff}(\varphi)\sim 1/F(\varphi)$5-essence, weak-field stellar solutions show the expected screened branches, but the naive interior scaling $G_{\rm eff}(\varphi)\sim 1/F(\varphi)$6 is singular at the center. Regularity instead requires

$G_{\rm eff}(\varphi)\sim 1/F(\varphi)$7

so the physical screened solution matches an FJBD-like core near the center and only then transitions into the screened regime (Haar et al., 2020). The result is not screening all the way to $G_{\rm eff}(\varphi)\sim 1/F(\varphi)$8, but a regularized screened profile over most of the star and surrounding screened zone.

The same work shows that fully relativistic stars, including neutron stars, remain screened in the nonlinear Einstein-scalar-matter system. The scalar profile transitions near the center, surface, and screening radius, and metric deviations from GR become tiny in the screened region. In that sense, the effective scalar contribution to local gravitational binding is strongly reduced even for $G_{\rm eff}(\varphi)\sim 1/F(\varphi)$9 matter coupling (Haar et al., 2020).

Nonlinear time evolutions sharpen this picture. When exact static screened stars are used as initial data, the system remains static only if high-resolution shock-capturing methods are used; standard finite-difference methods fail, likely because of strong micro-shocks in the scalar sector (Haar et al., 2020). Perturbations that do not increase compactness enough to cause collapse lead to bounded oscillations. For $f\sigma_8$0, the stars are effectively unscreened and evolve like FJBD solutions, while for screened configurations with $f\sigma_8$1, substantial perturbations remain stable provided collapse is not triggered (Haar et al., 2020).

The oscillation problem was revisited in a later nonlinear study of stellar pulsations and collapse. That work confirms that screening remains efficient during nonlinear stellar oscillations and that monopole scalar radiation is suppressed to undetectable levels for cosmologically relevant $f\sigma_8$2-essence scales (Bezares et al., 2021). The scalar perturbation propagates on the effective metric

$f\sigma_8$3

and in the screened oscillatory regime the field stays on the nonlinear branch that suppresses its coupling to the stellar motion (Bezares et al., 2021). This suggests that effective gravitational screening can survive genuinely nonlinear, though non-collapsing, dynamics.

Collapse changes the conclusion. In the earlier stellar-dynamics study, gravitational collapse causes the scalar characteristic speeds

$f\sigma_8$4

to diverge as

$f\sigma_8$5

before apparent black-hole or sound horizons form (Haar et al., 2020). The later study regularizes the evolution with a “fixing equation” inspired by dissipative hydrodynamics and evolves collapsing neutron stars past this divergence (Bezares et al., 2021). Its physical conclusion is that screening is less efficient in collapse than in static or oscillating stars, because the star must shed all scalar hair before forming a black hole (Bezares et al., 2021). Static success therefore does not guarantee dynamical predictivity.

4. Cosmological and extragalactic manifestations

On large scales, screening language often becomes phenomenological. The $f\sigma_8$6 statistic is designed to compare lensing and growth while reducing dependence on galaxy bias and $f\sigma_8$7. A galaxy-lensing measurement using CFHTLenS and BOSS CMASS found

$f\sigma_8$8

in agreement with the GR prediction

$f\sigma_8$9

corresponding to a $E_G$0 measurement (Alam et al., 2016). In this regime, the data do not show evidence for unscreened deviations from GR, and the result is interpreted as a broad consistency test of the relation between lensing and clustering rather than as a direct probe of nonlinear screening (Alam et al., 2016).

A later compilation of 66 $E_G$1 measurements and 16 $E_G$2 measurements instead found a marked preference for effective weakening of gravity. In the phenomenological parameterization

$E_G$3

the reported tensions with Planck/$E_G$4CDM were $E_G$5 from $E_G$6 alone, $E_G$7 from $E_G$8 alone, and $E_G$9 from the combined fit, with best fits favoring $\Lambda$0 and especially $\Lambda$1 at low to intermediate redshift (Skara et al., 2019). The same paper repeatedly emphasizes that overlapping datasets and simplified covariance treatment likely overestimate the formal significance, so the result is evidence for phenomenological gravitational weakening rather than a settled detection of a screening mechanism (Skara et al., 2019).

Galaxy-scale strong lensing provides another phenomenological implementation of screening. A step-function model assumes that GR is restored inside a screening radius $\Lambda$2, while outside $\Lambda$3 lensing responds to a modified slip parameter $\Lambda$4. Using 99 current strong lenses, the inferred $\Lambda$5 constraints were strongly model-dependent, and the authors stated that there is no noticeable evidence indicating some specific cutoff scales on kpc–Mpc scales (Lian et al., 2022). For 5000 simulated LSST lenses, however, the same methodology forecasts roughly $\Lambda$6 precision on $\Lambda$7 and sensitivity up to $\Lambda$8 kpc (Lian et al., 2022). In this usage, screening denotes suppression of extra gravitational degrees of freedom within a galactic radius rather than a fully specified microphysical mechanism.

A separate cosmological use of screening language appears in Energy-Momentum Log Gravity. There the bare cosmological constant is not removed from the action, but the modified matter redshifting causes the reconstructed effective dark energy density to cross zero and become negative at high redshift, screening $\Lambda$9 in the effective background dynamics. The constrained parameter is

$G_{\rm eff}$0

and the effective-dark-energy crossing occurs at

$G_{\rm eff}$1

in the observational fit (Akarsu et al., 2019). The paper is explicit that this screening is driven mainly by anomalous matter dilution, not by direct cancellation of the bare $G_{\rm eff}$2 term (Akarsu et al., 2019).

5. Propagation-sector screening and screened observables

Screening can also act on gravitational-wave observables. In scalar-tensor theories with luminal tensor speed and a time-varying effective Planck mass $G_{\rm eff}$3, unscreened propagation gives

$G_{\rm eff}$4

However, if both source and observer lie in locally screened regions with $G_{\rm eff}$5, then the standard-siren modification cancels and one recovers

$G_{\rm eff}$6

The wave can still traverse unscreened regions, but the observable amplitude correction depends only on the endpoint ratio $G_{\rm eff}$7 (Dalang et al., 2019). Screening in this setting does not remove modified propagation everywhere; it screens the integrated observable effect.

An analogous idea appears in massive-graviton phenomenology. If screened regions around the host galaxy and the Milky Way suppress the non-GR propagation sector, the graviton-mass correction accumulates only over an effective unscreened distance $G_{\rm eff}$8. The resulting waveform phase shift keeps its usual frequency dependence but with a coefficient proportional to $G_{\rm eff}$9. In the model-independent forecasts, future observations could constrain screening radii of $\mu$0–$\mu$1 Mpc and graviton masses of $\mu$2–$\mu$3 eV (Perkins et al., 2018). In specific dRGT and bigravity examples, the induced screening radii are typically too small for large effects with current detector sensitivities (Perkins et al., 2018).

Screening can also affect emission rather than propagation. In a chameleon-like “hairy wave” model of black-hole binaries embedded in a halo, the effective coupling is taken to switch to

$\mu$4

in the thin-shell interval $\mu$5 (Honardoost et al., 2019). In that shell, the modified binding energy and quadrupole power lead to a late-inspiral frequency shift; when one companion lies between $\mu$6 and $\mu$7, the gravitational-wave frequency is reduced relative to GR (Honardoost et al., 2019). This is again a screened observable, but now the screening is tied to environmental structure in the near zone rather than to endpoint effects.

6. Literal tension screening, critical surfaces, and recurring limitations

The most literal use of tension screening in the surveyed literature appears in higher-dimensional and geometric constructions. In the effective 5D cascading-gravity proxy, a brane with tension $\mu$8 can remain flat because the brane-bending scalar contributes to the junction condition through

$\mu$9

so the $\Sigma$0 term screens the extrinsic-curvature response to large tension (Agarwal et al., 2011). The bulk Jordan-frame geometry can remain exactly Minkowski even for arbitrarily large positive tension at the background level. The obstruction is perturbative stability: positive-tension solutions are generically ghostly, while ghost-free solutions exist only for sufficiently small negative tension (Agarwal et al., 2011). This is a genuine tension-screening mechanism, but not a viable one for the physically most interesting sign of the tension.

In $\Sigma$1 gravity, the field equations

$\Sigma$2

admit critical hypersurfaces $\Sigma$3 where the matter source term is screened away (Teruel, 10 Jun 2026). The apparent singularity of the rewritten non-conservation law is shown to be an artifact; the fundamental equations remain regular at $\Sigma$4. On a regular critical surface, matter must satisfy

$\Sigma$5

equivalently $\Sigma$6, so the normal projection of stress-energy vanishes there (Teruel, 10 Jun 2026). The paper interprets these surfaces as boundaries between attractive ($\Sigma$7) and repulsive ($\Sigma$8) gravitational phases. This is the closest surveyed example to screening of the gravitational action of stress-energy itself.

A recent black-bounce construction introduces an explicit gravitational tension screening inspired by Schwinger-like saturation. The unscreened vacuum tidal tension is identified as

$\Sigma$9

and replaced by the screened value

$F(\varphi)R$00

with $F(\varphi)R$01 (Estrada, 1 Jun 2026). This $F(\varphi)R$02 then enters the scale function

$F(\varphi)R$03

producing regular spherical, planar, and hyperbolic black-bounce geometries without prescribing an ad hoc core (Estrada, 1 Jun 2026). The bounce location is dynamical, and hyperbolic and planar regular black holes may satisfy the standard energy conditions near the bounce, whereas wormholes violate them near the throat (Estrada, 1 Jun 2026). Here “gravitational tension screening” is not metaphorical: it is the stated mechanism by which curvature growth is saturated.

Taken together, these works show that effective gravitational tension screening has no single canonical meaning. It can denote suppression of a scalar fifth force, renormalization of an effective Newton coupling, erasure of a gravitational-wave propagation signal, screening of bulk curvature sourced by brane tension, vanishing of the matter-curvature coupling on critical hypersurfaces, or saturation of a geometric vacuum tension (Brax et al., 2022, Dalang et al., 2019, Agarwal et al., 2011, Estrada, 1 Jun 2026). A consistent pattern nevertheless emerges. Screening can succeed statically yet fail dynamically, as in collapsing screened stars (Haar et al., 2020). Observational evidence for effective weakening can be suggestive yet covariance-limited and model-dependent (Skara et al., 2019). Geometric tension-screening mechanisms can regularize backgrounds while remaining vulnerable to ghosts or other consistency conditions (Agarwal et al., 2011). The subject is therefore best understood not as a single theory, but as a set of strategies for reducing the effective gravitational action of selected sectors, with the central questions being which quantity is screened, in which regime, and at what cost in stability, UV completion, or observational consistency.