Models of long-period variables of the globular cluster 47 Tuc

Basic Information

• Paper ID: 2511.09065
• Title: Models of long-period variables of the globular cluster 47 Tuc
• Author: Yu. A. Fadeyev (Institute of Astronomy, Russian Academy of Sciences)
• Classification: astro-ph.SR (Stellar and Solar Physics)
• Submission Date: Submitted to arXiv on November 12, 2025
• Paper Link: https://arxiv.org/abs/2511.09065

Abstract

This study presents evolutionary calculations for stars with zero-age main sequence mass $M_\mathrm{ZAMS}=0.86M_\odot$ and initial metallicities $Z=0.003$ and $Z=0.004$. Models selected from evolutionary sequences are used to calculate radial pulsations during the RGB (red giant branch), eAGB (early asymptotic giant branch), and TP-AGB (thermally pulsing asymptotic giant branch) phases. The research demonstrates that pulsating red giants in globular cluster 47 Tuc are not all Mira variables: the lower period limit for TP-AGB pulsations is approximately 70 days, while radial oscillation periods in the eAGB phase range from approximately 5 to 40 days. The periods and luminosities of hydrodynamic models for eAGB and TP-AGB pulsators distribute along a common period-luminosity relation. The small masses of Mira variables in 47 Tuc ($0.54M_\odot\le M\le 0.70M_\odot$) are the primary cause of irregular large-amplitude oscillations and outer-layer dynamical instability at periods $\Pi > 200$ days.

Research Background and Motivation

Research Questions

This study aims to determine the radial pulsation excitation conditions for red giants in globular cluster 47 Tuc at different evolutionary stages (RGB, eAGB, TP-AGB) and to clarify the evolutionary status of these long-period variables.

Importance of the Problem

1. Observational Peculiarity: 47 Tuc (NGC 104) is a globular cluster with anomalously high metallicity, nearly an order of magnitude higher than most globular clusters, resulting in a large population of pulsating red giants (14 out of 117 long-period variables discovered in 102 globular clusters by 2001 are in 47 Tuc)
2. Dust Condensation Phenomenon: Infrared observations reveal dust particle condensation in the outer atmospheres of 47 Tuc's long-period variables, closely related to periodic shock waves generated by large-amplitude stellar oscillations
3. Unclear Evolutionary Status: Existing research contains controversies regarding the evolutionary status of these pulsating red giants, particularly whether they all belong to the Mira class

Limitations of Existing Methods

1. Insufficient Linear Pulsation Analysis: Lebzelter and Wood (2005) used linear pulsation models for stars with $Z=0.004$ and $M=0.9M_\odot$, but failed to provide convincing evidence
2. Metallicity Estimation Bias: Earlier studies employed metallicities significantly higher than modern observational estimates
3. Lack of Consistent Calculations: Absence of consistent calculations combining stellar evolution and nonlinear pulsations

Research Motivation

Through consistent calculations of stellar evolution and nonlinear pulsations, using selected models from evolutionary sequences as initial conditions for hydrodynamic equations, systematically investigate pulsation characteristics of 47 Tuc red giants at different evolutionary stages.

Core Contributions

1. Established a complete evolution-pulsation calculation framework: Used MESA code (version r15140) for stellar evolution calculations, combined with radiative hydrodynamics and time-dependent convection equations for stellar pulsation calculations
2. Clarified pulsation characteristics at different evolutionary stages:
   • RGB phase: narrow or non-existent pulsation instability region ($Z=0.003$)
   • eAGB phase: period range 5-40 days with relatively regular amplitudes
   • TP-AGB phase: period lower limit approximately 70 days, irregular large-amplitude oscillations at long periods (>200 days)
3. Determined the primary evolutionary status of 47 Tuc long-period variables: Most pulsating red giants are in the eAGB phase rather than TP-AGB phase
4. Established period-luminosity relations: Period-luminosity data for eAGB and TP-AGB stars distribute along a common regression line
5. Explained the dynamical instability mechanism: Small mass ($0.54-0.70M_\odot$) is the primary cause of outer-layer instability at long periods

Detailed Methodology

Task Definition

Input: Stellar initial parameters (zero-age main sequence mass, metallicity, mass loss rate parameter)
Output: Pulsation characteristics at different evolutionary stages (period, amplitude, stability)
Constraints: Consistent with observational constraints for 47 Tuc (age 11.8-12.4 Gyr, metallicity $-0.78 \le [Fe/H] \le -0.66$)

Computational Architecture

1. Stellar Evolution Calculations

• Code: MESA version r15140
• Initial Parameters:
  • Zero-age main sequence mass: $M_\mathrm{ZAMS} = 0.86M_\odot$ (corresponding to AGB phase age ~11.9 Gyr)
  • Metallicity: $Z = 0.003$ (corresponding to $[Fe/H] = -0.67$) and $Z = 0.004$ (corresponding to $[Fe/H] = -0.54$)
  • Mass loss rate parameter (Reimers formula): $\eta_R = 0.3$ and $\eta_R = 0.5$
• Evolutionary Sequences: Total of 4 evolutionary sequences calculated (2 metallicities × 2 mass loss rate parameters)

2. Hydrodynamic Pulsation Calculations

• Equations: Radiative hydrodynamic equations + time-dependent convection equations
• Initial Conditions: Hydrostatic and thermal equilibrium models selected from evolutionary sequences
• Solution Method: Numerical solution of Cauchy problems
• Analysis Content:
  • Pulsation mode identification (fundamental/overtones)
  • Specific mechanical work analysis: $\oint PdV$ (P is total pressure, V is specific volume)
  • Amplitude evolution and stability determination

Technical Innovations

1. Consistent Calculation Method

Seamlessly connects stellar evolution calculations with nonlinear pulsation calculations, with evolutionary models directly serving as hydrodynamic initial conditions, avoiding parameter mismatch problems.

2. Multi-parameter Space Exploration

Systematically explores the metallicity and mass loss rate parameter space to assess the impact of uncertainties on results.

3. Physical Mechanism Analysis

Through the radial distribution of specific mechanical work $\oint PdV$, quantitatively analyzes the contributions of hydrogen and helium ionization regions to pulsation excitation/damping:

• Partially ionized hydrogen regions (adiabatic index $\Gamma_1 < 4/3$) are the primary driving regions
• Overtone node positions determine pulsation modes

4. Long-period Instability Mechanism

Reveals that small mass (leading to low-density outer convective zones and larger ionization regions) is the fundamental cause of dynamical instability at long periods.

Experimental Setup

Parameter Configuration

Stellar Evolution Parameters

• Zero-age main sequence mass: $0.86M_\odot$
• Metallicity: $Z = 0.003$ (solar value $Z_\odot = 0.014$) and $Z = 0.004$
• Mass loss rate: Reimers formula, $\eta_R = 0.3, 0.5$
• Evolutionary stages: RGB tip, eAGB, TP-AGB (4 thermal pulses)

Hydrodynamic Model

• Spatial discretization: Approximately 600-700 spherical shells
• Pulsation modes: Fundamental and first overtone
• Criteria:
  • Limit cycle oscillations: periodic oscillations with stable amplitude
  • Damped oscillations: amplitude gradually decreasing to zero
  • Irregular oscillations: $\Delta R/R \approx 1$

Evaluation Metrics

1. Pulsation period $\Pi$ (days)
2. Pulsation constant $Q = \Pi\sqrt{\bar{\rho}/\rho_\odot}$ (days, $\bar{\rho}$ is mean density)
3. Relative radial displacement amplitude $\Delta R/R$
4. Pulsation instability duration $\Delta t_\mathrm{ev}$ (million years)
5. Luminosity $\log(L/L_\odot)$

Comparison Benchmarks

• Linear pulsation results from Lebzelter and Wood (2005)
• Observed period lower limit for galactic Mira variables (Samus' et al. 2017)
• Period distribution of long-period variables observed in 47 Tuc

Experimental Results

RGB Phase Models

Main Findings (Table 1)

Key Conclusions:

1. The model with $Z=0.003, \eta_R=0.3$ does not pulsate (pulsation instability region does not exist)
2. Increased metallicity and mass loss rate cause evolutionary tracks to shift toward lower effective temperatures
3. Higher $Z$ and smaller mass make stars more unstable to radial oscillations (lower density outer convective zones, broader ionization regions)
4. Pulsation duration near the RGB tip is very short (<1 Myr)

Physical Mechanism (Figure 2)

• First overtone oscillations (overtone node $j \approx 318$, radius $r \approx 0.81\bar{R}$)
• Partially ionized hydrogen region ($\Gamma_1 < 4/3$) provides primary driving force
• For $Z=0.004, \eta_R=0.5$, hydrogen ionization region extends sufficiently deep to support fundamental mode oscillations

eAGB Phase Models

Evolutionary Characteristics (Figures 3-4)

1. Oscillation Appearance Sequence: First appears as first overtone, then transitions to fundamental mode as stellar radius and luminosity increase
2. Period Range: 5-40 days
3. Amplitude Characteristics: $\Delta R/R \lesssim 0.3$, oscillations relatively regular
4. Luminosity Upper Limit: Pulsations cease when $L \gtrsim 10^3 L_\odot$
5. Duration: Approximately $1.2 \times 10^7$ years (far longer than TP-AGB phase)

TP-AGB Phase Models

Thermal Pulse Characteristics (Figure 5)

• Mass Range: $0.59-0.71M_\odot$ at first thermal pulse, approximately $0.54M_\odot$ at post-AGB phase onset
• Thermal Pulse Interval: Average $1.6 \times 10^5$ years
• Number of Thermal Pulses: 4 (total TP-AGB phase duration $5-6 \times 10^5$ years)
• Pulsation Conditions: Appear only when $\log(L/L_\odot) \ge 3.4$, meaning substantial non-pulsating time within each pulse interval

Period and Stability

1. Period Range: 70-240 days
2. Irregular Oscillation Threshold: Transition to irregular large-amplitude oscillations ($\Delta R/R \approx 1$) at $\Pi \approx 170$ days
3. Period Lower Limit: Approximately 70 days (consistent with galactic Mira variables)

Period-Luminosity Relations (Figure 6)

For evolutionary sequences with $\eta_R = 0.3$, eAGB and TP-AGB models distribute along a common regression line:

$\log(L/L_\odot) = 0.743 \log \Pi + 1.79 \quad (Z = 0.003)$

$\log(L/L_\odot) = 0.724 \log \Pi + 1.78 \quad (Z = 0.004)$

• Coefficient standard errors < 3%
• Scatter increases at $\log \Pi \gtrsim 2$, attributed to evolutionary growth of carbon core mass and stellar luminosity

Key Experimental Findings

1. Evolutionary Status Determination: Most pulsating red giants in 47 Tuc are in the eAGB phase, not TP-AGB phase
2. Period Boundaries:
   • Period < 70 days → eAGB stars
   • Period ≥ 70 days → possibly TP-AGB stars
3. Mass Effect: Small mass ($0.54-0.70M_\odot$) is the primary cause of outer-layer dynamical instability at long periods (>200 days)
4. Dust Condensation Mechanism: Periodic shock waves from large-amplitude oscillations are the primary mechanism for dust particle condensation

Related Work

Observational Studies

1. Clement et al. (2001): Discovered 117 long-period variables in 102 globular clusters, with 14 in 47 Tuc
2. Lebzelter & Wood (2005), Lebzelter et al. (2005): Discovered additional 47 Tuc long-period variables, doubling the total
3. Percy & Gupta (2021): Continuous observations discovering new variables
4. Origlia et al. (2002), McDonald et al. (2011): Infrared observations revealing dust condensation evidence

Theoretical Studies

1. Lebzelter & Wood (2005): Linear pulsation analysis using $Z=0.004, M=0.9M_\odot$, but conclusions insufficient
2. Willson (2000): Proposed physical mechanism of periodic shock waves causing dust condensation
3. Fadeyev (2013, 2023, 2024): Author's previous work establishing radiative hydrodynamics and time-dependent convection equation framework

Advantages of This Work

1. More Accurate Parameters: Uses modern metallicity estimates ($Z=0.003$ corresponding to $[Fe/H]=-0.67$)
2. Nonlinear Calculations: Employs complete hydrodynamic equations rather than linear approximations
3. Consistent Framework: Seamless connection between evolution and pulsation calculations
4. Systematic Analysis: Covers three evolutionary stages (RGB, eAGB, TP-AGB)

Conclusions and Discussion

Main Conclusions

1. Evolutionary Status: Most pulsating red giants in 47 Tuc are in the eAGB phase, because:
   • eAGB duration ($1.2 \times 10^7$ years) is twice that of TP-AGB ($6 \times 10^5$ years)
   • Few thermal pulses in TP-AGB phase (≤4)
   • Only partial time within each pulse interval is in pulsation instability state
2. Period Distribution:
   • TP-AGB: period lower limit ~70 days (consistent with galactic Mira variables)
   • eAGB: 5-40 days
   • RGB tip: period ≥ 60-70 days, but instability region narrow or non-existent
3. Short-period Variables: 47 Tuc red variables with periods < 70 days should be classified as eAGB stars, not RGB or TP-AGB stars
4. Dynamical Instability: Outer-layer dynamical instability at periods > 200 days, with periodic shock waves as primary mechanism for dust condensation

Limitations

1. Mass Loss Uncertainty: Although two $\eta_R$ values considered, actual mass loss mechanisms may be more complex
2. Metallicity Range: $Z=0.004$ exceeds 47 Tuc observational upper limit, primarily for comparison with previous work
3. Convection Treatment: Time-dependent convection model has theoretical uncertainties
4. Number of Thermal Pulses: Only 4 thermal pulses simulated, may be insufficient for complete TP-AGB description

Future Directions

1. Extend to more thermal pulse cycles to study late TP-AGB evolution
2. Improve mass loss rate formulas considering pulsation-enhanced mass loss
3. Three-dimensional hydrodynamic simulations for more accurate convection and shock propagation description
4. Combine with observational data for statistical comparison to validate model predictions

In-Depth Evaluation

Strengths

1. Methodological Innovation

• Consistent Calculation Framework: First systematic combination of MESA evolution calculations with nonlinear hydrodynamic pulsation calculations, avoiding parameter mismatch
• Physical Mechanism Analysis: Quantitative analysis of driving/damping regions through $\oint PdV$ with clear physical picture

2. Experimental Sufficiency

• Parameter Space Coverage: 2 metallicities × 2 mass loss rates = 4 evolutionary sequences
• Complete Evolutionary Stages: Systematic study of RGB, eAGB, TP-AGB phases
• Detailed Data: Tables and figures comprehensively present computational results

3. Result Convincingness

• Resolves Observational Contradictions: Reasonably explains coexistence of different-period variables in 47 Tuc
• Quantitative Predictions: Provides clear period boundary (70 days) and period-luminosity relations
• Physical Self-consistency: Clear causal chain from small mass to low density to large ionization region to instability

4. Writing Clarity

• Rigorous structure with clear logic
• Large information content in figures with clear annotations
• Accessible physical explanations

Weaknesses

1. Methodological Limitations

• Single Mass: Only $M_\mathrm{ZAMS}=0.86M_\odot$ calculated, no exploration of mass effects
• Convection Model: Time-dependent convection theory has uncertainties, potentially affecting amplitude predictions
• Radiative Transfer Simplification: Radiative transfer in hydrodynamic calculations may involve simplifying assumptions

2. Experimental Setup

• Insufficient TP-AGB Coverage: Only 4 thermal pulses, may miss late evolution features
• Lack of Statistical Comparison: No quantitative statistical comparison with observed period distribution in 47 Tuc
• $Z=0.004$ Validity: This metallicity exceeds observational range, weakening applicability of some conclusions

3. Analysis Depth

• Period-Luminosity Scatter: While mentioning carbon core mass growth causes scatter, no quantitative analysis provided
• Irregular Oscillation Mechanism: Physical mechanism of irregularity at $\Pi > 200$ days not sufficiently detailed
• Dust Condensation: Mentions shock mechanism but doesn't calculate specific dust formation conditions

Impact

1. Contribution to Field

• Theoretical Breakthrough: Clarifies evolutionary status of 47 Tuc long-period variables, resolving long-standing controversy
• Method Demonstration: Evolution-pulsation consistent calculation framework generalizable to other star clusters and variable types
• Observational Guidance: Period boundary provides theoretical basis for future observational classification

2. Practical Value

• Variable Classification: Provides tools for variable classification in 47 Tuc and similar high-metallicity clusters
• Age Determination: Period-luminosity relations applicable to cluster distance and age estimates
• Dust Formation Research: Provides physical basis for understanding dust formation in globular clusters

3. Reproducibility

• Uses public MESA code (version clearly specified)
• Detailed parameter settings
• However, hydrodynamic code not public, limiting complete reproducibility

Applicable Scenarios

Directly Applicable

1. Long-period variable research in high-metallicity globular clusters (e.g., $\omega$ Cen)
2. Evolution and pulsation studies of low-mass (~$0.86M_\odot$) stars
3. Observational discrimination between eAGB and TP-AGB stars

Requires Adjustment

1. Different mass or metallicity stars: need recalculation of evolutionary sequences
2. Younger or older clusters: different zero-age main sequence mass and evolutionary timescales
3. Field stars (non-cluster): different mass loss rates

Not Applicable

1. Low-metallicity clusters ($[Fe/H] < -1.5$): different pulsation mechanisms
2. Massive stars ($M > 2M_\odot$): completely different evolutionary paths
3. Short-period variables (e.g., RR Lyrae): different pulsation instability strips

Key References

1. Paxton et al. (2019): MESA code, Astrophys. J. Suppl. Ser. 243, 10
2. Fadeyev (2013): Hydrodynamic equations, Astron. Lett. 39, 306
3. Lebzelter & Wood (2005): Linear pulsation analysis, Astron. Astrophys. 441, 1117
4. McDonald et al. (2011): Infrared observations and dust, Astrophys. J. Suppl. Ser. 193, 23
5. Willson (2000): Shock waves and dust condensation mechanism, Ann. Rev. Astron. Astrophys. 38, 573

Overall Assessment: This is a high-quality theoretical stellar physics research paper that resolves the long-standing controversy regarding the evolutionary status of 47 Tuc long-period variables through rigorous numerical calculations. The methodological innovation (consistent evolution-pulsation calculation), clear conclusions (mostly eAGB stars), and transparent physical picture (small mass causes instability) are notable strengths. Main limitations are single-mass calculations and lack of statistical comparison with observations. This work has significant theoretical value for globular cluster variable star research, with methods generalizable to other astrophysical problems.