Bi-Star Configurations in Binary Systems

• Bi-Star Configurations are binary star systems defined by geometrical arrangements, mass ratios, and additional components that determine their observable dynamics.
• Central configurations reduce the orbital evolution of binary systems to rigid rotation or homothetic scaling, aiding our understanding of binary-star planetary dynamics.
• Studies of circumbinary and misaligned circumstellar discs reveal that system architecture significantly influences planet formation and long-term stability.

Bi-Star Configurations denote binary-star system architectures in which the relative geometry of two stars, their mass ratio, and the presence of additional stellar, substellar, gaseous, or dusty components jointly determine the observable and dynamical state of the system. In current research, the term spans classical binary-star classes identified by astrometry, spectroscopy, and eclipses; coplanar and inclined circumstellar or circumbinary environments; hierarchical B-star multiples; and special solvable four-body central configurations adapted to binary-star planetary systems (Southworth, 2019, Veras, 2016, Kennedy et al., 2019).

1. Classification, geometry, and measurable structure

Binary stars are naturally classified along two axes. The first is the detection method: astrometric (visual), spectroscopic, and eclipsing. The second is the physical configuration: detached, semi-detached, and contact or overcontact, defined by how the stars fill their Roche lobes (Southworth, 2019).

Astrometric binaries provide the sky-projected orbit. With astrometry alone, one obtains orbital elements such as $P$, $e$, $\omega$, $\Omega$, and $i$, together with the angular semimajor axis. With a distance, the angular semimajor axis is converted into a true semimajor axis, and Kepler’s third law yields the mass sum. With radial velocities of both stars added, the individual masses, the mass ratio, and even the distance can be solved from the comparison between angular and linear orbit size (Southworth, 2019).

Spectroscopic binaries separate into SB1 and SB2 systems. SB1 observations directly yield $P$, $e$, $\omega$, $K_1$, and $\gamma$, together with the mass function and $e$0. SB2 systems add $e$1, so that the mass ratio $e$2, the relative $e$3, and the minimum masses $e$4 and $e$5 become available. Eclipsing binaries complement these data by supplying inclination and fractional radii from light-curve geometry. In SB2 eclipsing binaries, the combination of photometry, spectroscopy, and geometry yields true masses, radii, densities, luminosities, and distances. In good cases, masses and radii can be measured to better than $e$6, and in the best cases to approximately $e$7 precision (Southworth, 2019).

The underlying orbital relations are standard. Kepler’s third law connects semimajor axis and period to the mass sum, while the radial-velocity orbit satisfies

$e$8

For Roche geometry, the Eggleton approximation gives the lobe size of star 1 as

$e$9

with $\omega$0 (Southworth, 2019).

A common misconception is that binary configuration is exhausted by the mutual stellar orbit. The literature instead treats the stellar orbit as the backbone of a larger architecture that may include circumstellar discs, circumbinary discs, tertiary companions, or special few-body equilibria. This broader usage is essential in planetary, disc, and massive-star contexts (Kennedy et al., 2019, Frost et al., 5 May 2025).

2. Central configurations in binary-star planetary systems

A particularly restrictive meaning of Bi-Star Configurations appears in the study of four-body coplanar central configurations tailored to binary-star planetary systems. A central configuration is an arrangement of $\omega$1 point masses whose mutual gravitational accelerations are proportional to their displacements from the center of mass, so that the figure can rotate uniformly or evolve homothetically. The defining condition is

$\omega$2

with

$\omega$3

For coplanar relative equilibria in a rotating frame, $\omega$4 in appropriate units (Veras, 2016).

The classical $\omega$5 cases are Euler’s collinear configurations and Lagrange’s equilateral configurations. The Trojan architecture is the equilateral case. Its linear stability requires the planet-to-star mass ratio to be smaller than $\omega$6 (Veras, 2016).

The four-body systems studied for binary-star planetary applications consist of four planar bodies with one axis of symmetry and two equal masses; exactly two bodies are stars and two are substellar. The geometry is parameterized by two acute angles, $\omega$7 and $\omega$8, and a scale $\omega$9. For the convex cases,

$\Omega$0

whereas for the concave cases,

$\Omega$1

Here $\Omega$2 and $\Omega$3 are distances from one equal-mass object to star 1 and star 2, and $\Omega$4 is either the star-star or substellar-substellar separation, depending on the case (Veras, 2016).

Realistic planetary-system mass cuts drastically compress the allowed parameter space. The adopted constraints are that each star is at least ten times more massive than each substellar body, $\Omega$5, and that the stellar mass ratio lies within $\Omega$6. Typical stellar hosts span $\Omega$7--$\Omega$8 solar masses, while planets and substellar bodies are $\Omega$9--$i$0 solar masses. Under these restrictions, the allowed central-configuration phase space is reduced by over $i$1 (Veras, 2016).

Only four of the six planar symmetric cases survive the planetary mass cuts. Case #1 is the only admitted unequal-mass binary configuration. It places the stars on the symmetry axis and the two equal substellar bodies at the other corners of a near-rhombus, with $i$2. The geometry is a near-equilateral extension of the Sun-Jupiter-Trojan configuration, producing internal angles of about $i$3 at the stars and $i$4 at the substellar bodies. For $i$5, one has $i$6 and $i$7, so the side lengths are near-equilateral within approximately $i$8 (Veras, 2016).

Case #2 applies to equal-mass stars, with the symmetry axis connecting the substellar bodies. It is a near-rhombus rotated by $i$9 relative to Case #1, with $P$0. At $P$1, $P$2 and $P$3. Case #4 superposes a nearly collinear three-body central configuration with a near-equilateral one: $P$4 and $P$5, with the critical point at $P$6, $P$7. Case #6 corresponds to an equilateral triangle whose third vertex is doubled by the two substellar bodies coinciding at the critical point $P$8; under the planetary cuts, $P$9 may exceed $e$0 by up to approximately $e$1, while $e$2 may lie up to approximately $e$3 below $e$4 (Veras, 2016).

Cases #3 and #5 admit no solutions under the planetary mass cuts. More generally, departures from the ideal Trojan-derived geometry increase with substellar mass. Dust, pebbles, and asteroids force $e$5 and $e$6 to lie extremely close to their limiting values, whereas super-Jupiter masses allow deviations of a few degrees in Cases #1, #2, and #4, and up to approximately $e$7 in Case #6 (Veras, 2016).

These configurations are exceptional because the general orbital evolution of binary-star planetary systems is not integrable, whereas central configurations reduce the evolution to rigid rotation or homothetic scaling. The four-body planar symmetric systems have configuration constant

$e$8

so that $e$9 and $\omega$0 act as scale factors. A major limitation is that a full linear stability analysis for the admitted Cases #1, #2, #4, and #6 was not carried out; only existence regions were identified (Veras, 2016).

3. Circumbinary discs, polar equilibria, and disc-binary misalignment

A circumbinary disc orbits the binary center of mass rather than one stellar component. When such a disc is initially misaligned with the binary orbital plane, secular binary torques make its angular momentum vector precess. Dissipation drives the disc toward one of two equilibrium families: coplanar alignment with the binary plane, or a polar configuration in which the disc plane becomes nearly perpendicular to the binary orbital plane and its angular momentum aligns with the binary eccentricity vector, that is, the pericentre direction (Kennedy et al., 2019).

In the test-particle limit, the relevant secular invariant for a low-mass circumbinary ring around an eccentric binary is the Farago-Laskar constant

$\omega$1

where $\omega$2 is inclination relative to the binary plane, $\omega$3 is the longitude of ascending node measured from the binary pericentre direction, and $\omega$4 is the binary eccentricity. The separatrix between the coplanar and polar families is

$\omega$5

Rings with $\omega$6 precess about the binary angular-momentum vector, whereas rings with $\omega$7 precess about the pericentre direction. In a dissipative disc, trajectories do not cross this separatrix (Kennedy et al., 2019).

The prototypical observed case is the circumbinary disc around HD 98800 BaBb. HD 98800 is a hierarchical quadruple at $\omega$8 pc and age approximately $\omega$9 Myr, consisting of two inner binaries on roughly $K_1$0 au scales and an outer AB orbit of semimajor axis about $K_1$1 au. The inner binary BaBb has eccentricity $K_1$2, ascending node $K_1$3, inclination $K_1$4, and combined mass approximately $K_1$5 (Kennedy et al., 2019).

ALMA Band 6 continuum and CO $K_1$6--$K_1$7 observations resolve the disc. Its position angle is $K_1$8, and the inclination is either $K_1$9 or $\gamma$0 from the sky plane. The CO kinematics show the north side approaching, but do not uniquely fix the near side. The $\gamma$1 solution places the disc within approximately $\gamma$2 of the ideal polar state, with the disc angular momentum nearly aligned with the binary pericentre direction. The $\gamma$3 solution corresponds to a moderate misalignment of approximately $\gamma$4 and is dynamically short-lived in the simulations (Kennedy et al., 2019).

The dust and gas have different radial extents. The dust ring has inner and outer edges at $\gamma$5 au and $\gamma$6 au. The CO extends from $\gamma$7 au to $\gamma$8 au. The integrated $\gamma$9 mm dust flux is $e$00 mJy, which under the stated optically thin assumptions gives $e$01. The CO flux is approximately $e$02 Jy km s$e$03, implying $e$04 and $e$05 under the quoted assumptions; both masses remain uncertain because of optical-depth, temperature, and abundance effects (Kennedy et al., 2019).

Gas-free $e$06-body calculations show that test particles at $e$07--$e$08 au are generally ejected within less than $e$09 Myr for either orientation. In the polar case, survival islands appear near approximately $e$10 au and $e$11--$e$12 au, but these do not match the observed dust ring. This indicates that the solids are embedded in a more massive, stabilizing gas disc. Hydrodynamic simulations reinforce the interpretation: a moderately misaligned disc reorients into the polar configuration in several hundred years, whereas the polar state remains in the observed orientation for at least $e$13 BaBb orbits, or approximately $e$14 years. Since the alignment time is far shorter than the system age, the disc has likely been polar for most of its lifetime (Kennedy et al., 2019).

This system also constrains the incidence of polar configurations. Under assumptions of random initial disc orientations, test-particle dynamics, and a binary eccentricity distribution uniform from $e$15 to $e$16, the fraction of circumbinary discs that evolve to polar configurations is estimated as approximately $e$17; the main text quotes about $e$18. This suggests that polar circumbinary discs, and plausibly polar circumbinary planets, may be common around eccentric binaries, although transit surveys are strongly biased against detecting misaligned circumbinary planets (Kennedy et al., 2019).

4. Misaligned circumstellar discs and terrestrial planet formation

Binary configuration also governs planet formation in circumstellar, or S-type, discs. In the simulations of post-gas terrestrial planet formation, only S-type motion is modeled: the planetesimal and embryo disc orbits the primary, while the secondary acts as a secular perturber. Planar cases adopt $e$19, while misaligned cases use $e$20 relative to the initial disc plane. Two classes of inclined initial conditions are considered: PIC, an initially cold disc, and EIC, an initially excited disc (Zimmermann et al., 18 Dec 2025).

The simulations use the GPU-accelerated code GANBISS with a Bulirsch-Stoer integrator and adaptive tolerance $e$21. Each run contains two stars, $e$22 planetary embryos, and $e$23 planetesimals. The solid disc extends from $e$24 to $e$25 au, with total solid mass $e$26, divided roughly equally between embryos and planetesimals. The binaries are equal-mass, $e$27, with separation $e$28 au, eccentricity $e$29, and duration $e$30 Myr (Zimmermann et al., 18 Dec 2025).

Secular forcing is described in the paper through the Heppenheimer frequency

$e$31

with secular periods ranging from approximately $e$32--$e$33 yr across $e$34--$e$35 au for $e$36 au, and from approximately $e$37--$e$38 yr for $e$39 au. The simulations remain below the classical Kozai-Lidov threshold, so secular nodal and apsidal precession, rather than Kozai-Lidov cycles, controls the architecture (Zimmermann et al., 18 Dec 2025).

Two global results stand out. First, embryos migrate slightly inward in misaligned systems, especially during the first approximately $e$40--$e$41 Myr and most clearly in the outer disc where planetesimals survive longer. Second, the large initial oscillations in embryo inclinations and nodes about the secondary’s inclination and node damp over time. Damping is faster in the inner disc, where number densities, collision rates, and dynamical friction are higher (Zimmermann et al., 18 Dec 2025).

The collision statistics quantify the difference between planar and inclined binaries.

Collision class	Planar	Inclined, $e$42 EIC
Embryo-embryo	156 collisions; 100% accretive	173 collisions; 62.43% accretive, 0.58% destructive, 36.99% hit-and-run
Embryo-planetesimal	8,336 collisions; 99.18% accretive	6,090 collisions; 54.01% accretive, 0.67% destructive, 45.24% hit-and-run
Planetesimal-planetesimal	9,638 collisions; 35.98% accretive, 0.19% destructive, 63.83% hit-and-run	11,754 collisions; 7.99% accretive, 21.58% destructive, 70.43% hit-and-run

The physical interpretation is direct. Inclined discs have larger impact velocities and broader impact-angle distributions, which reduce gravitational focusing and shift collisions from accretive to hit-and-run or destructive regimes, especially for small bodies. In wide binaries, PIC discs remain cold longer, so embryo-planetesimal accretion is enhanced; in the examples quoted, wide PIC runs produce $e$43--$e$44 accretive embryo-planetesimal collisions per configuration, with fewer hit-and-run and almost no destructive outcomes. In tight binaries, PIC rapidly evolves toward EIC-like behavior because secular excitation acts quickly (Zimmermann et al., 18 Dec 2025).

The final planetary architectures remain viable in both aligned and misaligned binaries, but they differ systematically. Planar binaries typically produce $e$45--$e$46 terrestrial planets with masses $e$47--$e$48, semimajor axes approximately $e$49--$e$50 au, eccentricities $e$51, and inclinations $e$52. Inclined EIC binaries typically produce $e$53--$e$54 planets with masses $e$55--$e$56, semimajor axes approximately $e$57--$e$58 au, eccentricities up to about $e$59, and inclinations clustered around $e$60--$e$61 after damping. This suggests that bi-star misalignment does not prevent terrestrial planet formation, but it substantially changes the balance between growth and erosion (Zimmermann et al., 18 Dec 2025).

5. Techniques for resolving binary architecture across scales

Modern work on binary architecture combines classical orbital methods with high-angular-resolution interferometry, Gaia astrometry, spectroscopy, and eclipse photometry. The interferometric study of B-star multiplicity demonstrates the close-separation regime that is inaccessible to many other techniques: the sample of $e$62 B stars was observed with VLTI/PIONIER in the $e$63 band, with central wavelength approximately $e$64m, bandwidth approximately $e$65m, and spectral resolving power $e$66. Six baselines and three closure-phase triangles were recorded over projected baselines of approximately $e$67--$e$68 m, giving an angular resolution of roughly $e$69--$e$70 mas and an effective search domain of $e$71--$e$72 mas, or about $e$73--$e$74 au at the sample’s average distance of $e$75 pc (Frost et al., 5 May 2025).

Binary detection in this regime relies on simultaneous modeling of squared visibilities and closure phases. For a two-source model with flux ratio $e$76 and projected separation vector $e$77, the complex visibility is

$e$78

and the closure phase on a baseline triangle is

$e$79

Closure phase is zero for centrosymmetric brightness distributions and becomes non-zero in the presence of asymmetry, making it a strong multiplicity diagnostic (Frost et al., 5 May 2025).

The fitting pipeline uses PMOIRED with a grid-search engine inspired by CANDID, while detection limits follow the Absil et al. methodology. Typical $e$80 flux-ratio sensitivities range from approximately $e$81--$e$82 in favorable cases to approximately $e$83--$e$84 for the least favorable targets. In magnitude terms, $e$85 corresponds to $e$86 mag, and $e$87 to $e$88 mag. Bootstrapping is used for parameter uncertainties (Frost et al., 5 May 2025).

Completeness can then be quantified across period and mass-ratio space. Interferometric detectability exceeds $e$89 for periods of $e$90--$e$91 days and remains above $e$92 from $e$93 days to approximately $e$94 days, provided the $e$95-band flux ratio is at least $e$96. Spectroscopy is more complete at short periods: simulated $e$97-epoch campaigns with $e$98 km s$e$99 precision achieve more than $\omega$00 completeness for $\omega$01 d, with useful sensitivity out to about $\omega$02 d in favorable cases. The combined interferometric, spectroscopic, and eclipsing completeness is approximately $\omega$03 up to $\omega$04 d, and approximately $\omega$05 up to $\omega$06 yr, for $\omega$07 (Frost et al., 5 May 2025).

The standard summary statistics are

$\omega$08

with binomial uncertainty for MF and Poisson uncertainty for CF. These metrics are now routinely used to compare close interferometric samples, spectroscopic samples, and wider Gaia-based companion censuses (Frost et al., 5 May 2025).

6. B-star multiplicity, hierarchical systems, and case studies

Across observational scales, B-type stars are predominantly multiple. In the interferometric sample of $\omega$09 B stars, the multiplicity fraction within the interferometric range is $\omega$10, and the companion fraction is $\omega$11. When spectroscopic companions and Gaia-wide companions are added, the full-sample values become $\omega$12 and $\omega$13, while the spectroscopically complete sub-sample gives $\omega$14 and $\omega$15. The higher-order multiplicity fraction is approximately $\omega$16--$\omega$17, and many interferometric binaries become hierarchical triples when inner spectroscopic companions are included (Frost et al., 5 May 2025).

The larger Sco-Cen census extends this picture to $\omega$18 B-type primaries and $\omega$19 detected companions. The sample yields $\omega$20 binaries, $\omega$21 triples, $\omega$22 quadruples, and $\omega$23 quintuples, with overall single-star fraction $\omega$24 and companion frequency approximately $\omega$25. The single-star fraction depends strongly on primary mass: $\omega$26 for $\omega$27, versus $\omega$28 in the abstract and $\omega$29 in the details for lower-mass B stars. For companions with $\omega$30, the median semimajor axis is $\omega$31 au for primaries above $\omega$32, but $\omega$33 au below that threshold. Very wide companions with $\omega$34 au form a lower-mass population with median $\omega$35 and a companion-mass distribution centered near $\omega$36 (Gratton et al., 2023).

The mass-ratio distribution is not IMF-like. For massive primaries with $\omega$37 and $\omega$38 au, only $\omega$39 M-dwarf companions are found against $\omega$40 companions above $\omega$41, so M dwarfs comprise only approximately $\omega$42 of companions in this regime. The favored interpretation is disk fragmentation followed by selective mass accretion onto the secondary and inward migration, with more prolonged accretion episodes and stronger migration in higher-mass B systems (Gratton et al., 2023).

HD 155448 illustrates a higher-order B-star configuration with circumstellar structure. It is a comoving quintuple of B-type dwarfs near the ZAMS, with components A:B1V, B1:B6V, B2:B9V, C:B4Ve, and D:B8V. Standard ZAMS masses give a total of approximately $\omega$43--$\omega$44, with a central estimate near $\omega$45, and the system lies at mean distance approximately $\omega$46 kpc, with $\omega$47 kpc adopted for simple conversions. Relative to A, the projected separations are approximately $\omega$48 AU for B1, $\omega$49 AU for B2, $\omega$50 AU for C, and $\omega$51 AU for D; the B component itself is a binary with projected separation approximately $\omega$52 AU (Schuetz et al., 2011).

The C component of HD 155448 is an emission-line B4Ve star with strong H$\omega$53, H$\omega$54, [O I], [N II], and [S II] emission. The forbidden lines and H$\omega$55 are extended eastward by about $\omega$56--$\omega$57, with centroids at roughly $\omega$58 to $\omega$59 km s$\omega$60 relative to the star, while [O I] remains near systemic velocity and is less extended. Mid-infrared data separate silicate-dominated emission near the star from PAH-dominated emission in an asymmetric arc northeast of C, with PAH bands at $\omega$61, $\omega$62, and $\omega$63m. The favored interpretation is an outflow from C colliding with remnant interstellar material from the star-formation process (Schuetz et al., 2011).

Two Kepler SB2 systems show how binary architecture shapes stellar variability. KIC 4931738 is a B6 V + B8.5 V system with $\omega$64 d, $\omega$65, and $\omega$66. Its B6 V primary exhibits tidally excited g modes, with many frequencies consistent with $\omega$67. KIC 6352430 is a B7 V + F2.5 V system with $\omega$68 d, $\omega$69, and $\omega$70. Its B-type primary lies within the SPB instability strip and shows classical SPB g modes, while the unexpected power in the p-mode region is interpreted as nonlinear resonant mode excitation rather than tidal forcing (Pápics et al., 2013).

Taken together, these systems show that B-star bi-star configurations are rarely isolated binaries in a narrow sense. They are often hierarchical, sometimes quintuple, and frequently combine stellar multiplicity with discs, outflows, pulsations, or wide co-moving companions (Frost et al., 5 May 2025, Gratton et al., 2023).

7. Open problems, limitations, and broader implications

Several limitations recur across the literature. For binary-star central configurations, existence regions are known for the admitted coplanar four-body cases, but a comparable linear stability analysis has not yet been carried out. The future task is to compute eigenvalues of perturbations around Cases #1, #2, #4, and #6 and to determine how stability varies with substellar mass (Veras, 2016).

For polar circumbinary discs, the principal uncertainties concern optical depth, gas mass, detailed morphology, and long-term post-gas stability. In HD 98800, the inner rim is non-axisymmetric, possibly because of spiral or warped structure induced by the eccentric binary. Additional lines such as $\omega$71CO and C$\omega$72O are needed to constrain optical depth and excitation temperature, while the degree of gas-dust coupling and the long-term survival of planets after gas dispersal remain unresolved (Kennedy et al., 2019).

For S-type terrestrial planet formation in misaligned binaries, the current models are deliberately restricted. They begin after the gas phase, use an equal-mass binary, exclude giant planets, and treat in-simulation collisions as perfect merging, with Leinhardt-Stewart outcomes applied only in post-processing. The integrations run for $\omega$73 Myr, which is sufficient for early growth and dynamical friction but not for complete final assembly or definitive long-term stability classification (Zimmermann et al., 18 Dec 2025).

Multiplicity censuses also retain completeness limits. In the interferometric B-star survey, only a small number of companions with $\omega$74 or short periods may remain undetected, but the authors judge these corrections smaller than the statistical uncertainties of about $\omega$75--$\omega$76. In Sco-Cen, the census is nearly complete for stellar companions with $\omega$77 and $\omega$78 au, but completeness degrades below about $\omega$79 au for sub-solar companions. In HD 155448, more precise proper motions, radial velocities, and long-term astrometry are still needed to fully confirm the orbital architecture (Frost et al., 5 May 2025, Gratton et al., 2023, Schuetz et al., 2011).

A broader implication is that coplanarity is not the default end-state of all binary environments. Four-body central configurations tied to planetary mass cuts remain close to Trojan-like angles but are not exactly equilateral; circumbinary discs can relax to polar rather than coplanar states; and circumstellar S-type discs in inclined binaries can remain productive but collisionally harsher. On the stellar side, B stars are not merely binaries but frequently hierarchical multiples whose inner and outer scales span days to $\omega$80 au (Veras, 2016, Kennedy et al., 2019, Frost et al., 5 May 2025).

Bi-Star Configurations therefore form a unified research theme only when treated as a problem of geometry, hierarchy, and dissipation across scales. Whether the system consists of two stars and two substellar bodies in a rigidly rotating central configuration, a polar circumbinary planet-forming disc, a misaligned circumstellar embryo disc, or a compact B-star hierarchy with outer companions and circumstellar structure, the recurring question is the same: which geometries are dynamically admitted, which are observable, and which remain stable long enough to shape stars, discs, and planets.