Teorı́as de gravedad modificadas

Una clasificación sistemática

Yuri Bonder

Instituto de Ciencias Nucleares

Universidad Nacional Autónoma de México

Seminario hı́brido
Instituto Avanzado de Cosmologı́a

30 de abril de 2024
Motivation

• GR is our current theory of gravity.

• Gravity is the result of spacetime curvature (geometry!).
• It works really well...
Motivation

• GR is our current theory of gravity.

• Gravity is the result of spacetime curvature (geometry!).
• It works really well...

until it does not!

• Dark matter and dark energy.
• Singularities.
• Incompatible with quantum matter.
Proposal

• A new theory could solve some of these issues.

• Expect small deviations from a geometrical
description.
• Construct new theories within the geometrical
paradigm.
• Usually, only the GR action is modified [e.g.,
f (R)].
• Alternative theories can be put forward more
systematically.
Principles

• Use GR’s principles:

1. Spacetime is a 4-dim. pseudo-Riemannian differential manifold.
2. Gravity is completely described by a metric tensor.
3. Minimum coupling.
4. Lorentz invariance and (active) diffeomorphisms.
5. The theory is local.
6. The EOM are, at most, 2nd-order differential equations.
Principles

• Use GR’s principles:

1. Spacetime is a 4-dim. pseudo-Riemannian differential manifold.
2. Gravity is completely described by a metric tensor.
3. Minimum coupling.
4. Lorentz invariance and (active) diffeomorphisms.
5. The theory is local.
6. The EOM are, at most, 2nd-order differential equations.
• These principles lead to GR with a cosmological constant
(Lovelock, 1971).
• Another set of principles for the matter sector.
When is ∇a T ab = 0?

• To define Tab , we need

1
S[g , {ψ}] = SG [g ] + SM [g , {ψ}].
2κ
• We are implicitly using hypothesis 1, 2, 3, and 5.
• Diff. invariance (hypothesis 4) ⇒
Z Z
δSM δSM
0 = δdiff SM [g , {ψ}] = d 4x δdiff g ab + d 4x δdiff ψ.
δg ab δψ
When is ∇a T ab = 0?

• To define Tab , we need

1
S[g , {ψ}] = SG [g ] + SM [g , {ψ}].
2κ
• We are implicitly using hypothesis 1, 2, 3, and 5.
• Diff. invariance (hypothesis 4) ⇒
0
Z Z
4 δSM δSM
7
0 = δdiff SM [g , {ψ}] = d x δdiff g ab + 4
d x  δdiff ψ.
δg ab | {z } δψ
| {z } −2∇(a ξb)
√
−g
− 2 Tab

• This is valid on shell (i.e., δSM /δψ = 0 assuming hyp. 3).

When is ∇a T ab = 0?

• Integrating by parts (neglecting boundary terms):

√
Z
0 = − d 4 x −g ξ b ∇a Tab

• This is valid ∀ ξ a ⇒ ∇a Tab = 0.

• We can still relax hypothesis 6 and have ∇a T ab = 0.
• In particular, in f (R) gravity ∇a T ab = 0.
• We could also change the dimension of spacetime.
• Sometimes ∇a T ab = 0 is considered unremovable.
• GR:
D Eno local gravitational energy.
• Ĥ jumps during measurements [measuring apparatus cannot
compensate ∀ cases (Maudlin+Okon+Sudarsky 2019)].
Principle negation

• GR’s principles:
1. Spacetime is a 4-dim. pseudo-Riemannian differential manifold.
2. Gravity is completely described by a metric tensor.
3. Minimum coupling.
4. Lorentz invariance and (active) diffeomorphisms.
5. The theory is local.
6. The EOM are, at most, 2nd-order differential equations.
• Negation of the first principle leads to nongeometric (e.g.,
discrete theories) and theories in higher dimensions.
Principle negation

• GR’s principles:
1. Spacetime is a 4-dim. pseudo-Riemannian differential manifold.
2. Gravity is completely described by a metric tensor.
3. Minimum coupling.
4. Lorentz invariance and (active) diffeomorphisms.
5. The theory is local.
6. The EOM are, at most, 2nd-order differential equations.
• Negation of the second principle leads to additional
gravitational DOFs (e.g., scalar, nonmetric connections,...).
• Torsion: Einstein–Cartan theory.
• In principle, can be tested with polarized neutrons
(Arderucio+Bonder 2024).
• Currently: studying semiclassically and proposing
configurations for quantum interference.
Principle negation

• GR’s principles:
1. Spacetime is a 4-dim. pseudo-Riemannian differential manifold.
2. Gravity is completely described by a metric tensor.
3. Minimum coupling.
4. Lorentz invariance and (active) diffeomorphisms.
5. The theory is local.
6. The EOM are, at most, 2nd-order differential equations.
• Theories with nonminimal couplings include the well-known
Brans–Dicke theory (1961).
Principle negation

• GR’s principles:
1. Spacetime is a 4-dim. pseudo-Riemannian differential manifold.
2. Gravity is completely described by a metric tensor.
3. Minimum coupling.
4. Lorentz invariance and (active) diffeomorphisms.
5. The theory is local.
6. The EOM are, at most, 2nd-order differential equations.
• Requires nondynamical fields (Corral+Bonder 2019)
⇐⇒ principle of general covariance (Wald’s terminology).
• Nondynamical fields ↔ more fundamental DOFs.
Unimodular gravity
Basic aspects

• Action:
1 √ √
Z
d 4x


S= −g R + λ −g − F + LM (g , {ψ})
2κ
where λ is Lagrange mult., F nondynamical scalar density.
√
• EOM: {δLM /δψ} = 0, −g = F ,

1 TL 1 TL
Gab − λgab = κTab ⇐⇒ Rab ≡ Rab − Rgab = κTab .
2 4
• Semiclassically: VEVs of fields do not gravitate.
• If ∇a T ab = 0, ∃ integration constant Λ ⇒ possible solution to
the cosmological constant problem.
• Despite having different constraints w.r.t. GR, well posed
initial value formulation (Herrera+Bonder in press).
Unimodular gravity
Modified conservation law

• F reduces diffeo inv. to volume preserving diffeos: δSG ∝

 √ :0 √ √ 
 0
Z
d 4 x 2 −g ξ b a 
Gab + −g λ∇a ξ a + δλ−g
:

∇  −F ,

which requires ∇a ξ a = 0 ⇒ ξ a = ϵabcd ∇b αcd .

|{z}
arbitrary
• Invariance of SM under volume preserving diffeos:
√
Z
δSM = d 4 x −g αde ϵbcde ∇c ∇a Tab ⇒ ∇[b ∇a Tc]a = 0.

• Includes ∇a Tab = 0.
• Ja ≡ ∇b Tab (1-form) ⇒ dJ = 0 ⇒ Ja = −∇a Φ
(Poincaré Lemma, trivial top.).
• Proposal for Φ that produces additional contributions to Λ of
right size and sign (Josset+Perez+Sudarsky 2017).
Unimodular gravity
Other results

• Test Φ with test part./light (Bonder+Herrera+Rubiol 2023).

• Using Papapetrou’s method (1951):

u b ∇b u a = J b (δba + u a ub ) ≡ J⊥
a
.
TL = 0 + spherical symm. ⇒
• Gen. Birkhoff theorem: Rab

dr 2 2M
ds 2 = −fdt 2 + + r 2 dΩ2 , f =1− + Λr 2 .
f r
• Static!
• Assuming Φ = Φ(r ), reduce the problem to 1 DOF:

f l2
 
1 1
E − Φ(r ) = r˙2 + − κ ⇐⇒ E = r˙2 + Veff (r ; l)
2 2 r2 2
• E = integration constant, not conserved quantity.
Empirical limits

c2 1 M ℓ2 ℓ2 M 1
E− = r˙2 + Veff (r ; ℓ), Veff = − + 2 − 2 3 + 2 Φ.
2 2 r 2r c r c

t t t
• Consider Φ = br1 + br 22 + br 33 to keep the form of Veff .
• Orbits = bounded trajectories: Veff (rmin ) ≤ E − c 2 /2 < 0,
other restrictions for M, ℓ, and bit .

Veff
r1 r2
c2
E− 2 r
Empirical limits
• Deflexion angle in terms of semi-major axis (a) and
eccentricity (ϵ) (c = 1):
b2t 6πb3t
 
6πM
δϕ ≈ 1 − − .
a(1 − ϵ2 ) 3M 2 Ma2 (1 − ϵ2 )2
| {z }
43′′ (Mercury)

• PPN bounds from MESSENGER mission (NASA, 2011)

 
PPN ′′ 2(γ − 1) − (β − 1)
δϕ = 43 1 + .
3

where γ − 1 = (2.1 ± 2.3) × 10−5 (light bending)

β − 1 = (−4.1 ± 7.8) × 10−5 (Mercury)
b2t b3t
⇒ −6.8 < (3.3 × 104 ) 2
+ 1.1 3
< 1.4.
M⊙ M⊙
• Similar bounds on null curves.
Lorentz and diffeomorphisms noninvariance

Examples:
• Scalar/tensors nonminimally coupled to Riemann.
• I worked with the “t term”: t abcd Cabcd .
• For many years, thought to be unphysical. No reason found
(Bonder 2015).
• Effects/bounds in inflation (Bonder+Leon 2017) and in static
+ spherical-symmetry (Bonder+Peterson 2020).
• Indication that it is incompatible with any asymptotically flat
solution (Bonder+Peterson 2021).
• A nondynamical tensor coupled to a quantum scalar field
leads to unitarily inequivalent inertial vacua
(Arderucio+Bonder+Juárez 2023).
• It does not excite an Unruh-De Witt detector.
Principle negation
• GR’s principles:
1. Spacetime is a 4-dim. pseudo-Riemannian differential manifold.
2. Gravity is completely described by a metric tensor.
3. Minimum coupling.
4. Lorentz invariance and (active) diffeomorphisms.
5. The theory is local.
6. The EOM are, at most, 2nd-order differential equations.
• Tensor fields depend on one spacetime point: local objects.
• Trivial generalization: bitensors.
• We study a bitensorial connection (Bonder+Morales 2022).
• Hints of nonlocality:
• Bell’s theo. (1964) + experiments (e.g., Aspect et al. 1981).
• Expectation value of energy-momentum operator.
Principle negation

• GR’s principles:
1. Spacetime is a 4-dim. pseudo-Riemannian differential manifold.
2. Gravity is completely described by a metric tensor.
3. Minimum coupling.
4. Lorentz invariance and (active) diffeomorphisms.
5. The theory is local.
6. The EOM are, at most, 2nd-order differential equations.
• Leads to f (R) theories and other generalizations.
Conclusions

• Relaxing GR’s principles leads to interesting gravity theories.

• Generically: energy-momentum nonconservation.
• Proposal: negate one principle at a time.
• Sometimes interesting to negate more than one
(e.g., Bonder+Corral 2018).

• Understand the role of each principle within GR.

• Motivates new empirical and mathematical tests.
• Potential to uncover features of a more fund. theo. of gravity.