Theoretical Research · Stellar Fabric
What states of matter and fields could support a traversable wormhole—and remain consistent when the system is disturbed or loaded?
THE RESEARCH QUESTION
A traversable wormhole is a proposed connection between regions of spacetime through which matter or signals could pass. For a candidate to become physically meaningful, its supporting matter, boundaries and response to disturbances have to agree with the same laws.
Stellar Fabric studies the connection between a proposed spacetime geometry, the source that would support it, and the cost of keeping that source under control. The work combines analytical derivations with explicitly bounded numerical models. The results below follow that chain from source matching and constrained motion to state preparation and quantum stress.
In the Einstein-gravity throat framework studied here, the source must meet a negative null-stress requirement near the throat. Quantum theory also constrains negative energy. These are established starting points, not results invented by this project. [1] [2]
RESEARCH STATUS · OCTOBER 15
The work has produced conditional mathematical results, numerical cross-checks and restrictions on candidate mechanisms. A stable physical wormhole and a realizable supporting material have not been established.
Continue with the October 16 update: thermal states and the source–geometry test. The ten results below retain their original assumptions and date.
TWO PARTS OF THE PROBLEM
FOCUS I
Geometry, source and motion
These calculations test geometry, supporting stresses, allowable motion and loading. A favorable result in one part of a system can coexist with a failure elsewhere. Those failures help narrow the next model.
FOCUS II
State preparation and response
Here we follow the source more closely: how a state is prepared, how the control fields react, what happens at the boundary and which stresses the geometry requires. These calculations use one specified interaction.
The ten entries below show selected derivations and their implications. The calculations use different stated approximations; their partial successes do not combine into a self-consistent loaded spacetime.
TEN RESULTS · EQUATIONS & CHECKS
The first five results examine matching, energy and permitted motion. The next five follow one specified interacting seed. Each result shows the starting theory, what we calculated, how we checked it and what remains unresolved.
Notation is local to each result. The interacting-source calculations use normalized model units (c=ℏ=1); residuals describe the numerical tests, not certified error bounds. Exact finite-model identities and numerical evidence are labelled separately. Long equations scroll horizontally on small screens.
Conditional analytical obstruction
For the fixed spherical shape b(r)=r₀²/r, imposing a traceless Einstein source determines the lapse. The resulting family cannot stay positive and approach a nonzero finite value at both ends.
The input is the standard Einstein tensor and the zero-trace condition. Scalar-flat wormhole geometries are established in Dadhich and collaborators; their wider family is not excluded by this fixed-shape calculation. [7]
We solved the lapse equation for this chosen shape and checked the proposed source against the complete tensor, rather than matching only its negative radial component.
l is signed proper radial distance; r₀ is the throat radius; N is the positive lapse in gₜₜ=−N²; A and B are integration constants. Setting B=0 makes both end limits zero.
| Check | Result |
|---|---|
| Independent ODE integration | Runge–Kutta integration agrees with the closed lapse at maximum error 8.57×10⁻¹². |
| Two-end test | For B≠0 the limits have opposite signs. For B=0 neither limit is the required nonzero constant. |
Scope: smooth, static spherical geometry, the specified shape, Einstein gravity and the stated two-ended lapse requirement. Changing the shape, source trace or asymptotic conditions is a different problem.
Tensor matching and lapse, checked by the closed-form solution and numerical integration.
What remains unresolvedA different compatible source and geometry are still needed. This is not a no-go theorem for all wormholes.
Analytical identity in a restricted field class
In the selected shift-symmetric multi-field class, the homogeneous field velocities carry conserved currents. Those currents force one common direction to be blind to every time-dependent stiffness harmonic.
Autonomous dynamical stabilization and its stored-energy cost have prior literature, including Allahverdyan and Khalafyan. That mechanism motivates the control question; it does not provide our field-class identity. [9]
We isolated a conserved-current restriction that a leading fast-averaged stabilizer must respect, including mixed three-field examples.
V is the constant homogeneous field-velocity vector; J is its conserved shift-current vector; G(t)=−Paa at zero spatial derivatives is the spatial-gradient Hessian of the Lorentz-scalar P(XAB,uᵢ); Ḡ is its time average. K is the constant positive temporal Hessian. The temporary aA are spatial derivatives, unrelated to the radius a in the interacting-source model.
| Check | Result |
|---|---|
| Mixed-field identity audit | Twelve mixed three-field fixtures: max |GV−J|=4.57×10⁻⁸; a separate stress-identity residual is 5.71×10⁻¹¹. These are numerical fixture residuals, not rigorous error bounds. |
| Leading averaged direction | Each drive harmonic annihilates V. After canonical normalization the corresponding direction is K¹ᐟ²V; the leading fast-drive stiffening vanishes there. |
Scope: local shift-symmetric P(XAB,uᵢ), the specified homogeneous solution, a closed linear block, constant positive K and leading fast averaging. Mixing with additional modes, variable K, other couplings or higher-order corrections require separate analysis.
Conserved-current blind direction in the specified field class.
What remains unresolvedThe identity does not rule out every controller. It rules out treating this proposed drive as independent support in its blind direction.
Superseded candidate · corrected boundary calculation
An early thin-shell calculation admitted a candidate near γ=600. Recomputing radial and angular conditions with one consistent two-sided moving-junction geometry removed that window at the selected point.
The starting method is the established junction-condition treatment of dynamic thin shells, including conservation and momentum flux, as in Lobo and Crawford. [10]
We corrected the geometry used in our candidate boundary law and compared angular positivity with the selected positive radial acceleration response on the same configuration.
γ is the coefficient of the postulated surface operator −γ(div u)²/2, in the declared model units; it is not a Lorentz factor. ξ=l/r₀=0.215 identifies the inherited proper-coordinate anchor, with reference r₀=32. R=r/r₀, v=dR/dξ and f₊ is the assigned exterior metric coefficient. I is the pressure response to shell acceleration.
| Check | Result |
|---|---|
| Angular condition | The corrected angular sector requires γ>379.558972. |
| Radial condition | The selected positive acceleration response I>0 requires γ<356.339152. The interval is empty; γ=600 fails this response condition. |
This rejects the earlier tested window in the inherited static bulk branch, with local surface matter and a hypothetical foliation operator. The angular bound comes from a nonconstant phase mode on a fixed round surface. No coupled angular metric spectrum, physical cutoff or microscopic realization is supplied.
Corrected two-sided moving-junction calculation. The earlier result is retained as a correction.
What remains unresolvedA revised boundary interaction would have to pass the corrected conditions and have a realizable source law.
Restricted-sector analytical & numerical result
A positive fixed-boundary calculation can miss a permitted collective motion. In the selected local planar layer, solving the constraints and releasing the common proper-length mode exposes a negative reduced kinetic coefficient.
The lapse–shift decomposition and elimination of nondynamical constraints are standard methods. Gleyzes and collaborators provide related ADM/EFT methodology; their cosmological result is not the result claimed here. [11]
We kept the region’s proper-length response in the same constrained model and checked its short-patch coefficient against the numerical reduction.
The radial metric scale is B=exp(b), the region’s length is ℓ=∫B dx, and the normalized length coordinate is qℓ=√(2L)b. Its reduced kinetic term is Jℓ q̇ℓ²/2. L is the half-width. βℓ is the positive higher-derivative coefficient, unrelated to the coupling β in the interacting-source model. Jraw is the unconstrained kinetic integral; C is the shift quadratic form and D its coupling to length velocity (these are not the covariance variables of that interacting-source model).
| Check | Result |
|---|---|
| Background constraint control | On |x|≤0.08, the local background has maximum constraint residual below 2.57×10⁻⁹. |
| Reduced length sector | On [−0.03,0.03], Jℓ=−2.16791331 in the declared normalization. The analytical short-patch limit gives the same negative sign. |
The model is local and planar, with anchored induced lapse/transverse scale and zero endpoint shift, while proper length is free. The inherited higher-derivative term is −βℓ Aφ¹² Ktrace²/(2N⁶), where Aφ is the selected field amplitude. The finite projection and constitutive choice are restricted. This does not classify the full continuum or prove a global gravitational ghost theorem. The length is a geometric variable; no physical hardware support is identified.
Reacting local layer and proper-length calculation.
What remains unresolvedA favorable fixed-support result must be replaced by a compatible model of the allowed joint motion.
Exact field identity & independent finite checks
For the specified positive normalization Z=Zref exp(βη), changing a constant Z scales both signal inertia and stiffness. Canonical normalization cancels this common factor. Gradients and time dependence contribute through the effective operator.
Field redefinition, integration by parts and boundary variation are established methods. Hollands and Wald supply conservation context; Ishibashi and Wald explain why the operator’s boundary law must be specified. [4] [3]
We derived the effective operator for the chosen η–χ–φ action, kept its boundary reaction, and tested the force identities in an independent spectral representation.
φ is the signal with mass parameter Mφ, χ and η are classical mean fields in this approximation, and Zref>0 is the reference normalization. h and β are couplings; U is the canonical quadratic effective mass term. □ and ∇ use the assigned spacetime metric.
| Check | Result |
|---|---|
| Exact identity audit | The polynomial checks of the canonical transform and variational force identities have zero residual coefficients. |
| Boundary force balance · N=32 fixture | Interior stationary η force −2.00585245×10⁻⁶ is cancelled by endpoint reaction +2.00585245×10⁻⁶; net uniform-shift force is 1.43×10⁻¹⁸. |
This restricts the constant-background normalization mechanism of the selected action. It does not exclude effects of varying η, a different coupling or boundary work. A full interacting renormalized source and physical boundary apparatus remain separate requirements.
Canonical normalization, exact identities and an independent spectral cross-check.
What remains unresolvedAn independent physical advantage from the proposed normalization has not been established.
Established boundary formalism · project consistency result
On the fixed Euclidean product H²×S², counting only propagating vector bulk modes misses a harmonic boundary contribution. Including the full edge tower makes our local and spectral heat-kernel routes agree.
Vassilevich gives the standard local heat coefficients. Banerjee, Gupta and Sen, and Larsen and Lisbao establish the vector boundary-mode and product-space accounting used here. [12] [6] [13]
We applied that formalism to the selected mixed field seed, with its unequal product radii, and reconstructed the same ultraviolet coefficient by a second spectral route.
K is the local heat-kernel density, s is proper time, a is the H² radius, and R is the S² radius. The convention is Keff(s;x,x)∼(4πs)⁻²[a₀+a₂s+a₄s²+…]. Both routes use the same retained benchmark parameters and operator/ghost normalization. a₂ and a₄ name local short-time coefficients, not the radius a.
| Check | Result |
|---|---|
| Independent a₄ routes · selected benchmark | Local: 1.080769540509. Spectral cubic extrapolation: 1.080769541522. Difference ≈1.01×10⁻⁹ in model units; this is not a certified error interval. |
| Domain and tail controls | Four uncoupled radius-pair fixtures agree to ≤4.45×10⁻¹⁶; independent quadrature/tail controls differ relatively by ≤2.30×10⁻¹⁶ at the tested proper times. |
This is fixed-product Gaussian ultraviolet consistency. It is not the full finite absolute stress tensor. Numerical extrapolation uncertainty, physical wall/exterior, charge ensemble, zero-mode measure and finite effective-action coefficients are not resolved by this match.
Product heat kernel and independent vector-edge check.
What remains unresolvedThe boundary sector is necessary accounting, not an identified supporting source.
Finite Hartree evolution · independent representations
Starting from a finite self-consistent equilibrium, a χ-field pulse prepares signal correlations through a trilinear interaction. A separate run begins with a finite initial energy reserve and evolves without continued forcing. Both include the mean fields’ own contribution. The negative changes below are relative to the reference state; the absolute bare null observable stays positive.
Trilinear pair preparation and a stabilizing quartic potential have direct antecedents in Dufaux and collaborators. The standard mechanism is not a project invention. [5]
We combined reciprocal mean-field response, work accounting, total-null histories, independent mode-function controls and parameter ablations in the assigned curved radial domain.
These equations describe the separate closed-reserve run: M₀ is its positive mass matrix, F its normalized kick profile, A the initial amplitude and Jχ the external force. The driven comparison below uses a different force amplitude f₀. Numerical quantities are dimensionless model units, not joules or laboratory times.
| Check | Result |
|---|---|
| Weak external pulse · N=24, duration 12a | f₀=0.002: work W=9.38816×10⁻⁵; post-drive minimum ΔQ=−2.15527×10⁻⁴. Energy/work residual ≤2.12×10⁻¹¹; independent null-history difference ≤9.38×10⁻¹³. |
| Closed reserve · N=24, duration 48a | A=0.002 gives ΔE(0)=2×10⁻⁶ and minimum ΔQ=−3.40634×10⁻⁵. Energy drift ≤1.46×10⁻¹¹; independent total-null-history difference ≤1.51×10⁻¹². |
| Ablations and stronger drive | Removing h and λ disconnects the signal. Removing β leaves the weak-reserve minimum almost unchanged. At external f₀=0.05, post-drive minimum ΔQ is positive: +0.015094. |
Hartree approximation: Gaussian signal, classical χ and η, assigned geometry and boundary law, finite grid and finite test durations. ΔQ=Qbare−Qequilibrium is a reference-state change. The absolute bare Q remains positive in both runs. A microscopic force carrier and preparation of the initial kick are unresolved.
Driven preparation, closed-reserve evolution and independent mode-function checks.
What remains unresolvedThe run demonstrates finite-model preparation, not a physical negative source or autonomous wormhole maintenance.
Exact finite-model theorem · no continuum claim
With positive masses and the specified coupling relation, the Hartree potential can be written as a positive square plus a positive remainder. Conserved absolute energy then bounds the finite variables and the selected bare null observable.
Square completion, symplectic covariance transport and finite-dimensional ODE continuation are standard mathematics. The underlying trilinear/quartic potential has established preheating antecedents. [5]
We derived the Hartree remainder, explicit sitewise energy bounds and the positive-window null budget for this same closed finite action.
C=⟨φ²⟩ is the local signal variance in the zero-mean Gaussian sector; λ=h²/(2mχ²). Eabsolute includes the finite vacuum contribution. Qnull uses finite bare stress with window weights between zero and one, including both endpoints of that range.
| Check | Result |
|---|---|
| Exact algebra | Six sparse-polynomial identities have zero residuals, including square completion, energy cancellation, the symplectic generator and the null budget. |
| What establishes the theorem | Energy inequalities and finite-dimensional continuation give the exact result. No new trajectory or numerical evaluation of the bound was used as its proof. |
Fixed finite grid, time-independent boundaries, physical initial covariance, positive coefficients and exact autonomous equations. This proves no finite-time blowup, not closeness to equilibrium. The bound depends on cutoff and absolute finite energy; it is not a uniform continuum, full gravitational-stability or renormalized-sign theorem.
Exact coercivity and the raw-null energy budget.
What remains unresolvedThe sign of a covariantly renormalized stress requires physical subtraction and matched finite terms. It is not determined by the positive bare bound.
Exact first-order identity · unresolved quantum response
Expanding around the constant-radius product makes the source requirement precise. A small localized flare asks for a signed first-order null stress, while the declared classical positive-kinetic sector has no linear null contribution there.
The relation between flaring geometry and null stress is established by Hochberg and Visser. Our calculation specializes that requirement to the selected product background and its allowed source response. [1] [4]
We derived the normalized first-order equation, checked its contact and counterterm signs, and evaluated the spatial sign pattern of the chosen localized test direction.
ds²=−exp(2Φ)dt²+dx²+r²dΩ², r=R(1+εu), Φ₀=log cosh(x/a), and k=(exp(−Φ),1,0,0). Primes denote x derivatives. Gkk and Tkk are contractions with this metric-normalized radial null vector; ε is the small flare amplitude.
| Check | Result |
|---|---|
| Exact first-order audit | Eight exact rational-polynomial checks have zero residuals, including Gkk, the zero classical linear term and varied-null-vector contact cancellation. |
| Localized test profile | u=−sech²(x/4a) gives (δGkk/ε) at x=0 equal to −1/(4a²); its sign changes at |x|/a≈1.049914. |
| Independent integral checksum | A 4096-panel Simpson check differs from the closed-form weighted integral by 2.43×10⁻¹⁶. This checks an identity, not physical source admission. |
This is a necessary radial condition, not the complete tensor equations. The full renormalized response Πeff was not computed, no operator norm was established, and the semiclassical background was not closed. The absence of a localized solution to Lu=0 does not prove a uniform gap on the infinite line.
First-order product-branch requirement and localized sign check.
What remains unresolvedA physically specified source, boundary/exterior and finite effective-action coefficients must satisfy this equation together with the remaining field and constraint equations.
THE NEXT SCIENTIFIC REQUIREMENT
The next task is a consistently defined interacting source, its quantum state, boundary and exterior, with a physically specified prescription for finite quantum stresses. Geometry and loading then have to be tested with that same source.
Success requires compatible answers to all five questions below. A positive result for one sector cannot stand in for the whole system.
Do geometry, matter, boundaries and the exterior solve one compatible set of equations?
A complete physical throat solution remains open.
What happens when the source, support and geometry move together?
Selected sectors are tested; full joint stability remains open.
Does the candidate survive finite matter or energy passing through it?
Partial loading tests do not establish sustained passage.
What supplies control, reaction time, work and heat removal?
A physical actuator and complete operating budget remain open.
Can the required stress be produced, prepared and measured?
A realizable supporting source has not been established.
METHODS & SOURCES
We tested how proposed sources match the geometry, examined their allowed motion and followed how a state could be prepared. Analytical derivations and numerical cross-checks show where the chosen models work and where they fail. They do not yet establish a physically realizable supporting material or a stable wormhole. Scientific priority has not been established; the references below identify the established theory behind the calculations.
David Hochberg & Matt Visser · 1998
Established geometric context for the null-energy requirement at a throat in the stated general-relativistic framework.
L. H. Ford & Thomas A. Roman · 1996
Restrictions on the magnitude and duration of negative energy under the paper's field and sampling assumptions.
Akihiro Ishibashi & Robert M. Wald · 2004
Why dynamics on anti-de Sitter backgrounds requires a specified boundary law.
Stefan Hollands & Robert M. Wald · 2005
Established conditions for consistent stress-energy accounting in interacting quantum field theory.
J. F. Dufaux, G. N. Felder, L. Kofman, M. Peloso & D. Podolsky · 2006
Known antecedents for the bounded trilinear potential and transfer into pair excitations used in the preparation comparison.
Shamik Banerjee, Rajesh K. Gupta & Ashoke Sen · 2010 preprint / 2011 journal
Established vector boundary-mode and zero-mode accounting on the product background; these methods do not establish the Stellar Fabric source.
Naresh Dadhich, Sayan Kar, Sailajananda Mukherji & Matt Visser · 2002
Established scalar-flat geometries. Result 01 fixes a particular shape and stricter two-ended lapse requirement; it does not exclude the wider family.
Matt Visser, Sayan Kar & Naresh Dadhich · 2003
Established integrated energy-condition measures. Result 02 uses a negative-density proper-volume functional and a narrower profile class, not the same measure.
A. E. Allahverdyan & E. Khalafyan · 2022
Antecedent for autonomous stabilization and resource accounting; the project’s conserved-current restriction is a separate field-class calculation.
Francisco S. N. Lobo & Paulo Crawford · 2005
Junction conditions, surface conservation and momentum flux. Result 04 corrects our own proposed boundary-response calculation.
Jerome Gleyzes, David Langlois, Federico Piazza & Filippo Vernizzi · 2013
Related ADM and effective-action constraint methodology. Its cosmological conclusions are not imported into the local length-mode result.
D. V. Vassilevich · 2003
Established Laplace-type heat coefficients used for the local route in result 07.
Finn Larsen & Pedro Lisbao · 2014 preprint / 2015 journal
Established boundary-state and harmonic one-form accounting. The boundary tower is prior literature, not a Stellar Fabric discovery.
The question continues.
The constraints become more precise.