1: \documentclass{mn2e}[3]

2: \usepackage{rotating}

3: \input psfig.sty

4: \begin{document}

5: 

6: \title[binary evolution \& blue stragglers]

7: {Binary coalescence from case A evolution -- mergers and blue stragglers}

8: \author[Chen and Han]{Xuefei Chen$^{1}$\thanks{

9: xuefeichen717@hotmail.com} and Zhanwen Han$^{1}$\\

10: $^1$National Astronomical Observatories/Yunnan Observatory, CAS, Kunming,

11: 650011, P.R.China}

12: \maketitle

13: 

14: \begin{abstract}

15: We constructed some main-sequence mergers from case A binary

16: evolution and studied their characteristics via Eggleton's stellar

17: evolution code. Both total mass and orbital angular momentum are

18: conservative in our binary evolutions.

19:  Assuming that the matter from the secondary homogeneously mixes with

20: the envelope of the primary and that no mass are lost from the

21: system during the merger process, we found that some mergers might

22: be on the left of the zero-age main sequence as defined by normal

23: surface composition (i.e helium content $Y=0.28$ with metallicity

24: $Z=0.02$ for Pop I) on a colour-magnitude diagram(CMD) because of

25: enhanced surface helium content. The study also shows that central

26: hydrogen content of the mergers is independent of mass. Our simple

27: models provide a possible way to explain a few blue stragglers

28: (BSs) observed on the left of zero-age main sequence in some

29: clusters, but the concentration toward the blue side of the main

30: sequence with decreasing mass predicted by Sandquist et al. will

31: not appear in our models. The products with little central

32: hydrogen in our models are probably subgiants when they are

33: formed, since the primaries in the progenitors also have little

34: central hydrogen and will likely leave the main sequence during

35: merger process. As a consequence, we fit the formula of magnitude

36: $M_{\rm v}$ and $B-V$ of the mergers when they return back to

37: thermal equilibrium with maximum error 0.29 and 0.037,

38: respectively.

39: 

40: Employing the consequences above, we performed Monte Carlo

41: simulations to examine our models in an old open cluster NGC 2682

42: and an intermediate-age cluster NGC 2660. Angular momentum loss

43: (AML) of low mass binaries is very important in NGC 2682 and its

44: effect was estimated in a simple way. In NGC 2682, binary mergers

45: from our models cover the region with high luminosity and those

46: from AML are located in the region with low luminosity, existing a

47: certain width. The BSs from AML are much more than those from our

48: models, indicating that AML of low mass binaries makes a major

49: contribution to BSs in this old cluster. Our models are

50: corresponding for several BSs in NGC 2660. At the region with the

51: most opportunity on the CMD, however, no BSs have been observed at

52: present. {\bf Our results are well-matched to the observations if

53: there is $\sim 0.5M_\odot$ of mass loss in the merger process, but

54: a physical mechanism for this much mass loss is a problem.}

55: 

56: \end{abstract}

57: 

58: \begin{keywords}

59: binaries:close -stars:evolution - blue stragglers

60: \end{keywords}

61: 

62: \section{Introduction}

63: Much evidence shows that primordial binaries make an important

64: contribution to blue stragglers (BSs) \cite{fer03,dpa04,map04}. At

65: present, a few BSs, i.e. F190, $\theta$ Car, have already been

66: confirmed to be in binaries by observations, and their formation

67: may be interpreted by mass transfer between the components of a

68: binary. Whereas in intermediate-age and old open and globular

69: clusters, the number of observed close binaries among well-studied

70: BSs is consistent with the hypothesis of binary coalescence. For

71: example, Mateo et al. \shortcite{mat90} made a comparison of the

72: number of close binaries with the total number of BSs in NGC 5466

73: and found that it is an acceptable claim that all non-eclipsing

74: BSs are formed as the result of mergers of the components in close

75: binaries, though the possibility of other mechanisms to produce

76: BSs cannot be ruled out due to the large uncertainties in their

77: analysis. Monte-Carlo simulations of binary stellar evolution

78: \cite{pol94} also show that binary coalescence may be an important

79: channel to form BSs in some clusters (e.g. with an age greater

80: than 40 Myr). Meanwhile, the arguments in theory show that W UMa

81: binaries (low-mass contact binaries) must eventually merge into a

82: single star \cite{web76,web85,ty87,mat90}. Observationally, the

83: lack of radial velocity variations for most BSs further indicates

84: that binary coalescence may be more important than mass transfer

85: for BS formation \cite{str93,pol94}. FK Comae stars are generally

86: considered to be direct evidence for binary coalescence

87: \cite{str93}. The smallest mass ratio of components among observed

88: W UMa systems to date is about 0.06. All of the above show that it

89: is important to study the remnants of close binaries. However the

90: merge process is complicated and the physics during the process is

91: still uncertain. Recently, Andronov, Pinsonneault \& Terndrup

92: \shortcite{apt06} studied the mergers of close primordial binaries

93: by employing the angular momentum loss rate inferred from the

94: spindown of open cluster stars. Their study shows that main

95: sequence mergers can account for the observed number of single BSs

96: in M67 and that such mergers are responsible for at least  one

97: third of the BSs in open clusters older than 1 Gyr. The physics of

98: mergers are limiting case treatments in the study of Andronov,

99: Pinsonneault \& Terndrup \shortcite{apt06}. Based on previous

100: studies of contact binaries and some assumptions, we construct a

101: series of merger models in this paper, to study the structure and

102: evolution of the models and show some comparisons with

103: observations.

104: 

105: Case A binary evolution has been well studied

106: by Nelson \& Eggleton \shortcite{nel01}.

107: They defined six major subtypes for the evolution

108: (AD, AR, AS, AE, AL and AN) and two rare cases (AG and AB).

109: Three of the subtypes (AD, AR, AS) lead the binary contact as both components

110: are main--sequence stars

111: and two cases (AE and AG) reach contact

112: with one or both components having left the main sequence.

113: As there is no description for weird objects except for

114: two merged main--sequence stars,

115: merger products (except for two main-sequences stars) are generally

116: assumed to have terminated their evolution \cite{pol94},

117: i.e. they have left the main sequence and cannot be recognized as BSs.

118: Here we are interested in the cases of two main-sequence stars,

119: i.e. cases AS, AR and AD.

120: If $t_{\rm dyn}$, $t_{\rm KH}$, $t_{\rm MS}$ represent the dynamic timescale,

121: thermal timescale and main sequence timescale of the primary

122: (the initial massive star,*1), respectively,

123: the following shows a simple definition of the three evolutionary cases:

124: AD--dynamic Roche lobe overflow (RLOF), $\dot{M}>M/t_{\rm dyn}$;

125: AR--rapid evolution to contact,  $\dot{M}>M/t_{\rm KH},

126: t_{\rm contact} -t_{\rm RLOF}(*1)<0.1t_{\rm MS}(*1)$;

127: AS--slow evolution to contact,

128: $t_{\rm contact} -t_{\rm RLOF}(*1)>0.1t_{\rm MS}(*1)$,

129: where $t_{\rm RLOF}$ and $t_{\rm contact}$ are the ages

130: at which RLOF begins and the binary comes into contact, respectively.

131: In case AD,

132: the core of the secondary spirals in quickly

133: and stays in the center of the merger.

134: The merger then has a chemical composition similar to that of the primary,

135: resembling the result of smoothed particle hydrodynamic calculations

136: \cite{lrs96,sill97,sill01}.

137: We therefore studied just the systems in cases AR and AS for this work.

138: 

139: \section {Assumptions}

140: Using the stellar evolution code devised

141: by Eggleton \shortcite{egg71,egg72,egg73},

142: which has been updated with the latest physics over the last three decades

143: \cite{han94,pol95,pol98},

144: we re-calculate the models of cases AS and AR with primary masses between

145: 0.89 and $2M_\odot$ until the systems become contact binaries.

146: The structures of the primaries and the compositions of the secondaries

147: are stored to construct the merger remnants.

148: 

149: Before the system comes into contact,

150: the accreting matter is assumed to be deposited onto

151: the surface of the secondary with zero falling velocity

152: and distributed homogeneously all over the outer layers.

153: The change of chemical composition on the secondary's surface caused by the

154: accreting matter is

155: \begin{equation}

156: {\partial X_i / \partial t }={(\partial M /\partial t)/[(\partial M /\partial t){\rm d}t+M_{\rm s}}] \cdot (X_{i{\rm a}}-X_{i{\rm s}}),

157: \end {equation}

158: where $\partial M /\partial t$ is the mass accretion rate,

159: $X_{i{\rm a}}$  and $X_{i{\rm s}}$ are element abundances

160: of the accreting matter and of the secondary's surface for species $i$,

161: respectively,

162: and $M_{\rm s}$ is the mass of the outermost layer of the secondary.

163: The value of $M_{\rm s}$ will change

164: with the moving of the non-Lagrangian mesh

165: as well as the chosen model resolution,

166: but it is so small ($\sim 10^{-9}-10^{-12} M_{\odot}$)

167: in comparison with

168: $(\partial M /\partial t){\rm d}t$ ($\sim 10^{-3}-10^{-5} M_{\odot}$)

169: during RLOF that we may ignore the effect of various $M_{\rm s}$

170: on element abundances.

171: Before and after RLOF, we get $\partial X_{\rm i}/\partial t =0$

172: from the equation,

173: which is reasonable in the absence of mixing \cite{ch04}.

174: 

175: The merger models are constructed based on the following assumptions:

176: (i) contact binaries with two main-sequence components coalesce finally and

177: the changes of structures of individual components

178: during coalescence are ignored;

179: (ii) the matter of the secondary is homogeneously mixed with

180: that of the primary beyond the core-envelope transition point,

181: which separates the core and the envelope of the mass donor;

182: (iii) the system mass is conserved.

183: 

184: Firstly, we present a brief discussion on these assumptions.

185: Webbink \shortcite{web76} studied the evolutionary fate of

186: low-mass contact binaries, and found that a system cannot sustain

187: its binary character beyond the limits set by marginal contact

188: evolution ($\mu =M_1/(M_1+M_2)=1.0$). He stated that a contact

189: binary will very likely coalesce as the primary is still on the

190: main sequence in a real system. Up to now, it is widely believed

191: that case AD probably leads to common envelope, spiral-in, and

192: coalescence on quite a short timescale. The final consequences of

193: AS and AR are not very clear, but Eggleton \shortcite{egg00}

194: pointed out that systems undergoing AR or AS evolution may

195: maintain a shallow contact (perhaps intermittently) as the mass

196: ratio becomes more extreme, and finally coalesce. Recent study on

197: W UMa (Li, Zhang \& Han, 2005) also shows that these systems will

198: be eventually coalescence. The merged timescale, i.e. the time

199: from a binary contact to coalescence, is important here. If it is

200: too long, the structures of both components will change remarkably

201: and the system may have not completed coalescence within the

202: cluster age. There are many {\bf conflicting estimates} for the

203: timescale, however, from observations and theoretical models of

204: the merger process. {\bf Early observational estimates range from}

205: $10^7$--$10^8$ yr in various environments \cite{van79,eggen89}.

206: The following study explored the average age about $5 \times 10^8$

207: yr \cite{van94,dry02}. Bilir et al. \shortcite{bil05} pointed out

208: that the age difference between field contact binaries and

209: chromospherically active binaries, 1.61 Gyr, is likely an upper

210: limit for the contact stage by assuming an equilibrium in the

211: Galaxy, whereas the study of W UMa by Li, Han \& Zhang

212: \shortcite{lhz04} suggested a much longer timescale, about 7 Gyr.

213: We adopt the empirically estimated values in this paper {\bf (i.e

214: from $5 \times 10^7$ to $1 \times 10^9$ yrs)} and ignore the

215: changes of structure of individual components during merger

216: process. For low-mass contact binaries, the common envelope is

217: convective \cite{web77}, and the matter in it is thus homogeneous.

218: If a system mimics shallow contact during coalescence, it is

219: reasonable to assume that the matter of the secondary mixes with

220: the envelope homogeneously. Van't Veer \shortcite{van97} found

221: that the mass loss from the system during coalescence is at a rate

222: of about $2 \times 10^{-10}M_\odot{\rm yr}^{-1}$ by observations.

223: If we consider that the coalescence time is $5 \times 10^8$ yr in

224: a binary, only $0.1M_\odot$ is lost from the system as the binary

225: finally becomes a single star. We then roughly assume that the

226: mass is conservative during coalescence. However mass loss might

227: be an important way to carry orbital angular momentum away from

228: the binary in this process.

229: 

230: Secondly, we discuss the choice of the core-envelope transition point

231: which separates the core and the envelope in the primary.

232: Many characteristics of the merger are relevant to the choice,

233: e.g. the chemical composition in the envelope,

234: evolutionary track on Hertzsprung-Russel diagram,

235: and some observational characteristics.

236: Unfortunately,

237: one cannot find the core-envelope transition point in a main-sequence

238: star as easily as in evolved stars

239: because the density profile,

240: as well as many other thermodynamic quantities

241: (entropy, pressure, temperature etc.), is smooth and does

242: not have a deep gradient for main-sequence stars.

243: Chen \& Han \shortcite{ch05} studied

244: the influences of core-envelope transition point

245: on the mergers of contact binaries with two main-sequence components.

246: They found that

247: one may ignore the effects which result from different choices of

248: the transition point on colours and magnitudes of the merger

249: if it is outside the nuclear reaction region of the primary,

250: which is commonly considered

251: as the nearest boundary of the secondary reaches in cases AS and AR.

252: In this paper, the core-envelope transition is determined as the point

253: within which the core produces 99 per cent of total luminosity.

254: This choice is generally outside the nuclear reaction regions

255: and has little effect on the final results.

256: 

257: Finally the merger remnant is constructed as follows:

258: it has the total mass of the system

259: and a chemical composition within $M_{\rm 1c}$

260: similar to the core of the primary.

261: The chemical composition in the envelope of the merger is given by

262: \begin{equation}

263: X_i=(M_{i2}+M_{i1\rm b})/(M_2+m_{\rm b}),

264: \end{equation}

265: where $M_{i2}$ and $M_{i1\rm b}$ denote the total masses of species $i$

266: of the secondary and of the primary's envelope, respectively.

267: $m_{\rm b}$ is the envelope mass of the primary.

268: There might be a region in which

269: the helium abundance is less than that of the outer region.

270: The matter in this region then has a lower mean molecular weight than that

271: in the outer region.

272: This results in secular instability and thermohaline mixing \cite{kip80,ulr72}.

273: We include it as a diffusion

274: process in our code \cite{ch04}.

275: 

276: In the models of Nelson \& Eggleton \shortcite{nel01}, both total

277: mass and angular momentum are conservative. It was mentioned by

278: the authors, however, that these assumptions were only reasonable

279: for a restricted range of intermediate masses, i.e spectra from

280: about G0 to B1 and luminosity class III-V. Observationally, some

281: low mass binaries with late-type components show clear signs of

282: magnetic activity, which indicates that the systems evolve by way

283: of a scenario implying angular momentum loss (AML) by magnetic

284: braking \cite{mes84}. Magnetized stellar winds probably do not

285: carry off much mass, but they are rich in angular momentum because

286: of magnetic linkage to the binaries. For close binaries, rotation

287: is expected to synchronize with orbital period, so AML is at the

288: expense of the orbital angular momentum, resulting in orbital

289: decaying and the components approaching each other. A detached

290: binaries, then, may become contact and finally coalesce at or

291: before the cluster age \cite{ste95}. There are a number of

292: subjects including the treatment of AML

293: \cite{lhz04,ste06,mk06,dek06}. For simplicity, the conservative

294: assumption is also adopted in our binary evolutions. In old

295: clusters, however, AML of low mass binaries is very important and

296: {\bf we estimate its importance in another way (see section 4.2).}

297: 

298: \section{Evolutionary Results}

299: A set of binaries undergoing AS and AR evolution from Nelson \&

300: Eggleton \shortcite{nel01} are choosen to study the

301: characteristics of the merger products and their connections with

302: blue stragglers. Table A1 gives the initial parameters of the

303: binary systems, their RLOF information and the structures and

304: evolutionary consequences of the mergers. The first three columns

305: contain the initial mass of the primary $M_{\rm 1i}$, the initial

306: mass ratio $q_{\rm i}$ (the primary to the secondary) and the

307: initial orbital period $P_{\rm i}$ in logarithmic, where $P_{\rm

308: ZAMS}$ is given by (Nelson \& Eggleton, 2001)

309: \begin{equation}

310: P_{\rm ZAMS} \approx $$0.19M_{\rm 1i}+0.47M_{\rm 1i}^{2.33}\over 1+1.18M_{\rm 1i}^2$$.

311: \end{equation}

312: The fourth and fifth columns are the ages at which

313: Roche lobe overflow begins ($t_{\rm RLOF}$)

314: and the binary comes into contact ($t_{\rm contact}$) in our calculation.

315: The next three columns show some system parameters at $t_{\rm contact}$,

316: i.e. the mass of the primary $M_{\rm 1}$,

317: the mass of the secondary $M_{\rm 2}$

318: and the orbital period $P_{\rm contact}$.

319: The remaining columns present the evolutionary results of the mergers:

320: the lifetime on the main sequence ($t_{\rm MS}$),

321: the central hydrogen mass fraction of the merger as constructed

322: ($X_{\rm Hcc}$) and after adjustment ($X_{\rm Hcm}$),

323: surface abundances for the elements H ($X_{\rm Hs}$), He ($X_{\rm Hes}$)

324: and the ratio of C/N at the surface ($(C/N)_{\rm s}$).

325: 

326: Figure. \ref{CMD} shows the location of the mergers

327: on a colour-magnitude diagram (CMD)

328: when the central hydrogen mass fraction reaches its maximum,

329: at which we consider that

330: the merger returns to thermal equilibrium and begins normal evolution.

331: In a real case, the thermal equilibrium point of the constructed models

332: are probably not just at the maximum of central hydrogen mass,

333: but the divergence should be very small and have no influence on the results.

334: `.' and `$\times $' in the figure represent the mergers

335: from AS and AR, respectively.

336: Open circles show the possible BSs produced from our models

337: in an old open cluster M67

338: and the dashed one shows zero-age main sequence (ZAMS) for

339: $(Z,Y)=(0.02, 0.28)$,

340: where $Z$ is metallicity and $Y$ is helium content.

341: In figure \ref{CMD} we see that most of the mergers

342: are located in the main sequence

343: while a few are to the left of ZAMS.

344: Surface chemical composition and central hydrogen content are both

345: responsible for placing objects to the left of ZAMS.

346: Generally the mergers have a larger helium content than 0.28

347: because the matter from the secondary, including some helium-rich matter

348: produced via nuclear reaction, homogeneously mixes

349: in the envelope of the primary.

350: They are thus bluer than stars of the same mass

351: and of the same central hydrogen mass fraction with $Y=0.28$.

352: Furthermore, the mergers may be on the left of ZAMS

353: if central H contents close to 0.7, initial H mass fraction for Pop I.

354: 

355: Sandquist, Bolte \& Hernquist \shortcite{sbh97} argued that

356: a fainter BS should have a less massive progenitor,

357: and therefore has a lower helium content in the core.

358: This means that

359: BSs should spend an increasing amount of time

360: near the ZAMS with decreasing mass,

361: which will lead to a definite concentration

362: toward the blue side of the main sequence in the region of low mass BSs.

363: In Fig. \ref{XHD}

364: we present the distribution of central hydrogen mass fraction at its

365: maximum $X_{\rm Hcm}$ with the mass of the merger.

366: No evidence in the figure shows that

367: lower mass mergers should have higher hydrogen mass fractions

368: in the center (say, more close to ZAMS).

369: In our models, there are at least two factors affecting

370: the central hydrogen mass fraction of a merger, i.e.

371: the evolutionary degree of the parent stars

372: when the system comes into contact and

373: the development of the central convective region of the merger

374: before normal evolution.

375: The former is relevant to the mass of the progenitor

376: and their contact ages.

377: If we simply assume that all BSs in a cluster are formed at a same time

378: (which is obviously not true),

379: larger (more massive) progenitors have less hydrogen content in the center,

380: but their mergers may develop larger convective regions in the centers,

381: involving more of the H-rich matter

382: from around the nuclear region of the primary in the center of the merger.

383: As a consequence, the enhancement of central hydrogen content

384: might be comparable to, or larger than, the less massive mergers.

385: On the other hand, if a fainter BS is produced earlier than a brighter one --

386: the formation time of a BS is strongly dependent on

387: some other system's initial parameters as well as its progenitors' mass.

388: The fainter BS has then evolved for a relatively longer time and may

389: have less hydrogen in the center than the larger one.

390: Therefore the blue concentration with mass decreasing predicted by

391: Sandquist, Bolte \& Hernquist \shortcite{sbh97} may be not true in

392: a real case.

393: In fact, we have not found the concentration observationally,

394: whereas there is a certain width of BSs sequence.

395: 

396: We obtained some models with very little central hydrogen content,

397: i.e $X_{\rm Hcm}$ very close to zero. Seen in Fig. \ref{XHD}, all

398: these objects are from the AS channel. Long-time RLOFs of their

399: progenitors lead the primaries be near the termination of main

400: sequence when the systems being contact. Detailed calculations for

401: these mergers show that their main-sequence timescales are in

402: order of $10^8$ yrs. Such unusually long lifetimes are relevant to

403: the development of central convective core. {\bf Comparison to the

404: primaries of the progenitors, the products have higher masses and

405: will develop larger convective core masses in the following

406: evolutions, leading central hydrogen increase. On the other hand,

407: nuclear reactions will consume some hydrogen in the core. The

408: maximum of central hydrogen mass fraction is the equibibrium point

409: at which the hydrogen involved in the core is comparable to that

410: exhausted by nuclear reactions. After that, the consumed hydrogen

411: is more than that involved in the core, but the growth of

412: convective core adds fresh fuels and extends the lifetime in this

413: phase.} For a $1.6M_\odot$ star with a stable convective core, it

414: is just about $2\times10^7$ yr when hydrogen mass fraction ranges

415: from 0.005 to 0.0000, while about $10^8$ yr in these models.

416: 

417: We should pay more attention to the mergers with little central

418: hydrogen content. The triangles in figure 1 show the positions of

419: these objects. From Table A1, we see that the primaries are very

420: close to the exhausted of hydrogen. So it is very likely that the

421: primaries have left the main sequence as eventually coalescence,

422: and the mergers are subgiants and much redder than that shown in

423: Fig. 1. Meanwhile, the development of convective core and the

424: merger process are probably synchronous. After coalescence, the

425: convective core develops more quickly than that in our models.

426: Sandquist, Bolte \& Hernquist \shortcite{sbh97} once suggested

427: that low-mass BSs (near the turnoff of a cluster) in M67 might be

428: subgiant stars and a very small central hydrogen content in these

429: objects may remarkably shorten their MS lifetimes and increase the

430: chance of them being in the subgiant region. This may explain the

431: observed spread in colour of low-mass BSs. These mergers with

432: little central hydrogen in our models are similar to the objects

433: mentioned by Sandquist, Bolte \& Hernquist \shortcite{sbh97}, but

434: they may be up to $2.7M_\odot$, not just staying in the region

435: around the turnoff in M67.

436: 

437: \begin{figure}

438: \centerline{\psfig{figure=CMD.ps,width=8cm,angle=270}}

439: \caption{The locations of the mergers as they reach their

440: maximum hydrogen mass fraction on a colour-magnitude diagram.

441: `.' and `$\times $' represent the mergers from AS and AR, respectively.

442: Open circles show the merge models from Table 1 and the triangles are the

443: objects with little central hydrogen content.}

444: \label{CMD}

445: \end{figure}

446: 

447: \begin{figure}

448: \centerline{\psfig{figure=XHD.ps,width=8cm,angle=270}}

449: \caption{The maximum of central hydrogen mass fraction

450: for all the merger models.}

451: \label{XHD}

452: \end{figure}

453: 

454: Figures.\ref{mv} and \ref{bv} show how

455: the magnitude in the V band, $M_{\rm V}$,

456: and the colour, $B-V$, change with the merger mass $M_{\rm BS}$

457: at $X_{\rm Hcm}$.

458: $M_{\rm V}$ is calculated by

459: \begin{equation}

460: M_{\rm V} = M_{\rm bol}-BC, M_{\rm bol}=4.75-2.5\times {\rm log}(L/L_{\odot}),

461: \end{equation}

462: where $BC$ and $B-V$ are obtained by linear interpolation

463: from the BaSel-2.0 model \cite{lcb97,lcb98}.

464: 

465: As seen in the figures,

466: $M_{\rm V}$ and $B-V$ are affected

467: not only by the total mass, but also by the initial orbital period.

468: For ${\rm log}P_{\rm i}/P_{\rm ZAMS}=0.1$,

469: Figs. \ref{mv2} and \ref{b-v2} present $M_{\rm V}$ vs $M_{\rm BS}$ and

470:  $B-V$ vs $M_{\rm BS}$, respectively.

471: The solid lines are fitted by

472: (two models marked with star are ruled out

473: because they are outliers):

474: 

475: \begin{equation}

476: M_{\rm v}=$$-10.93+22.33M-2.65M^{2.5}\over -10+10.76M$$

477: \end{equation}

478: 

479: \begin{equation}

480: (B-V)_0=$$47.89+9.75M-46.99M^{0.5}\over -10+16.07M$$

481: \end{equation}

482: The maximum errors for the two equations are 0.19 and 0.019,

483: respectively.

484: 

485: Considering the initial orbital period, we get

486: \begin{equation}

487: M_{\rm v}=$$15.87+19.58M-3.01M^{2.5}\over -10+16.22M$$-1.5({\rm log}P_{\rm i}/P_{\rm ZAMS}$$-0.05)

488: \end{equation}

489: 

490: \begin{equation}

491: (B-V)_0=$$102+21.73M-101.5M^{0.5}\over -10+27.44M$$+$$623({\rm log}P_{\rm i}/P_{\rm ZAMS}-0.05)^{2} \over 10+62.4M^{3}$$

492: \end{equation}

493: The maximum errors for the two equations are 0.29 and 0.037, respectively.

494: The four models marked with stars in Figs.\ref{mv} and \ref{bv}

495: are also ruled out in the fitting for the same reason as above.

496: The distribution of error for eqs.(6) and (7), from detailed calculation,

497: are shown in Figs.\ref{mverr} and \ref{bverr}.

498: 

499: \begin{figure}

500: \centerline{\psfig{figure=mv.ps,width=8cm,angle=270}}

501: \caption{Visual magnitude changes with the mass of the merger.

502: The dots and open crosses are for AS and AR, respectively.

503: The stars represent the ones which are ruled out in eqs.(7) and (8).}

504: \label{mv}

505: \end{figure}

506: 

507: \begin{figure}

508: \centerline{\psfig{figure=bv.ps,width=8cm,angle=270}}

509: \caption{Colour index $B-V$ changes with the mass of the merger.

510: The dots and open crosses are for AS and AR, respectively.}

511: \label{bv}

512: \end{figure}

513: 

514: \begin{figure}

515: \centerline{\psfig{figure=mv2.ps,width=8cm,angle=270}}

516: \caption{Visual magnitude changes with the mass of the merger as

517: ${\rm log}(P_{\rm i}/P_{\rm ZAMS})=0.10$ .

518: The dots and open crosses are for AS and AR, respectively.

519: The stars represent the ones which are ruled out in fitting eqs.(5).

520: The solid line shows the fitting curve via equation(5).}

521: \label{mv2}

522: \end{figure}

523: 

524: \begin{figure}

525: \centerline{\psfig{figure=b-v2.ps,width=8cm,angle=270}}

526: \caption{Colour index $B-V$ changes with the mass of the merger

527: as ${\rm log}(P_{\rm i}/P_{\rm ZAMS})=0.10$.

528: The dots and open crosses are for AS and AR, respectively.

529: The stars represent the ones which are ruled out in fitting eqs.(6).

530: The solid line shows the fitting curve via equation(6).}

531: \label{b-v2}

532: \end{figure}

533: 

534: \begin{figure}

535: \centerline{\psfig{figure=mverr.ps,width=8cm,angle=270}}

536: \caption{Visual magnitude error from equation (6) and the calculations.}

537: \label{mverr}

538: \end{figure}

539: 

540: \begin{figure}

541: \centerline{\psfig{figure=bverr.ps,width=8cm,angle=270}}

542: \caption{Colour index $B-V$ error  from equation (7) and the calculations.}

543: \label{bverr}

544: \end{figure}

545: 

546: Initial parameters of the four systems ruled out in eqs.(6) and (7)

547: are ($M_{\rm 1i}, {\rm log}q_{\rm i},{\rm log}(P_{\rm 0}/P_{\rm ZAMS})$)=

548: (0.89, 0.2, 0.1), (0.89, 0.25, 0.1), (1.26, 0.15, 0.15) and (1.41, 0.25, 0..2).

549: From Table A1, we see that all four of these mergers

550: have central hydrogen content close to zero,

551: i.e. 0.0007, 0.0020, 0.0028 and 0.0007, respectively.

552: This might be the reason that they deviate from most of our models

553: 

554: The evolutionary tracks of the mergers with $M=1.97M_{\odot}$

555: (($M_{\rm 1i}, {\rm log}q_{\rm i}$)=(1.41, 0.40))

556: are shown in Fig. \ref{197CMD}.

557: Two stars with $Y=0.28$ and $M=2.00$ and $1.97M_{\odot}$ are also presented

558: in the figure (dotted lines) for convenient comparison.

559: In the figure,

560: we can hardly distinguish the difference from various initial orbital period

561: except for the main-sequence life, $t_{\rm MS}$, of the mergers.

562: Though $t_{\rm MS}$ cannot be directly observed,

563: it may affect observations, i.e.

564: together with $t_{\rm contact}$ of the progenitor system,

565: it is crucial for whether or not the merger can be observed or not by now.

566: At the same time, various initial orbital periods result in

567: different luminosity functions for mergers --

568: a longer initial orbital period leads a shorter $t_{\rm MS}$,

569: but during the shorter $t_{\rm MS}$,

570: the merger mainly stays in the relatively high luminosity region

571: (Fig. \ref{197CMD}).

572: 

573: \begin{figure}

574: \centerline{\psfig{figure=197CMD.ps,width=8cm,angle=270}}

575: \caption{Evolutionary tracks of the mergers after adjustment.

576: ($M_{\rm 1i}, {\rm log}q_{\rm i}$)=(1.41, 0.40) for the solid lines and

577: a, b, c, d, e represent

578: ${\rm log}P_{\rm i}/P_{\rm ZAMS}=0.10,0.15, 0.20, 0.25,0.30$, respectively.

579: The dotted lines show the evolution of two stars with

580: surface helium content $Y$=0.28 for $M_{\rm 1i} =1.97$ and $2.00M_{\odot}$.}

581: \label{197CMD}

582: \end{figure}

583: 

584: 

585: \begin{figure}

586: \centerline{\psfig{figure=tms.ps,width=8cm,angle=270}}

587: \caption{The timescale of the mergers staying on the main sequences.

588: As $t_{\rm cc}=5 \times 10^8$ yr, the open circles show the models fulfilled

589: $t_{\rm contact}+t_{\rm cc}\le 4.5\times 10^9$ yr and

590: $t_{\rm contact}+t_{\rm cc}+t_{\rm MS}\ge 3.5\times 10^9$ yr.

591: The squares and stars are for

592: $t_{\rm contact}+t_{\rm cc}\le 2.4\times 10^9$ yr and

593: $t_{\rm contact}+t_{\rm cc}+t_{\rm MS}\ge 1.4\times 10^9$ yr and

594: $t_{\rm contact}+t_{\rm cc}\le 1.0\times 10^9$ yr and

595: $t_{\rm contact}+t_{\rm cc}+t_{\rm MS}\ge 0.9\times 10^9$ yr, respectively.}

596: \label{tms}

597: \end{figure}

598: 

599: \section{the mergers and BSs in open clusters}

600: If a BS in a cluster is the merger remnant of binary coalescence

601: \footnote{It just means AS and AR evolutionary channels in this section},

602: then there are some constrains on the progenitor and on the merger itself.

603: Firstly the total mass of the progenitor should be larger than

604: the turnoff of the cluster $M_{\rm to}$.

605: Secondly, the contact age of the progenitor $t_{\rm contact}$

606: should be appropriate.

607: It cannot be so short that the merger has terminated its

608: evolution on the main sequence, indicating

609: $t_{\rm contact}+t_{\rm cc}+t_{\rm MS}\ge t_{\rm cluster}$.

610: Here $t_{\rm cc}$ is the timescale from binary contact to final merger.

611: $t_{\rm contact}$ also cannot be too long

612: to complete its merger process at $t_{\rm contact}$, indicating

613: $t_{\rm contact}+t_{\rm cc}\le t_{\rm cluster}$.

614: If a binary has already come into contact

615: but has not completed the mergering process

616: (i.e. a W UMa system), it can also be considered as a BS \cite{str93},

617: e.g. S1036 and S1282 in M67.

618: The initial parameter space for W UMa systems may be obtained by

619: $t_{\rm contact}+t_{\rm cc}\ge t_{\rm cluster} \ge t_{\rm contact}$.

620: 

621: \begin{table}

622:  \begin{minipage}{85mm}

623:  \caption{Characteristics of some open clusters (i.e. the age $t$

624: and metallicity $Z$)

625:  and the numbers of BSs ($N_{\rm BS}$), W UMa

626: systems ($N_{\rm W UMa}$) and stars on the main sequence to two

627: magnitudes below the turnoff ($N_{\rm 2}$) in them. Most of the

628: information is from Rucinski(1998) for $N_{\rm W UMa}$, Ahumada \&

629: Lapasset (2007) for $N_{\rm BS}$ and $N_{\rm 2}$, and Xin \&

630: Deng(2005) for $t$ and $Z$. The data not from the references above

631: are marked with 1, 2, where 1 -- Mochejska et al. (2004); 2 --

632: Kafka et al.(2004). The stars means that we have not found related

633: reports.}

634:  \label{tab1}

635:    \begin{tabular}{lccccc}

636: \hline

637: ID & $t (Gyr)$ & $Z$ & $N_{\rm bs}$ & $N_{\rm W UMa}$ & $N_{\rm 2}$\\

638: \hline

639: Be 33    & 0.7 & 0.005 & 2  & 1 & 270\\

640: Tom 2    & 1   & 0.009 & 17 & 4 & 440\\

641: NGC 2243 & 1.1 & 0.007 & 9  & 2 & 120\\

642: NGC 2158 & 1.2 & 0.006 & 40 & $6^1$ & 700\\

643: NGC 2660 & 1.2 & 0.02  & 8  & $0^*$ & 150\\

644: NGC 6939 & 1.6 & 0.02  & 5  & $6^2$ & 180\\

645: NGC 3680 & 1.6 & 0.026 & 1  & $0^*$ & 30\\

646: NGC 752  & 1.7 & 0.014 & 1  & 1 & 25\\

647: NGC 7789 & 2.0 & 0.016 & 22 & 5 & 130\\

648: NGC 2682 & 4.0 & 0.02  & 30 & 3 & 175\\

649: NGC 188  & 7.0 & 0.024 & 24 & 7 & 185\\

650: NGC 6791 & 7.2 & 0.039 & 75 & 4 & 800\\

651: Be 39    & 8.0 & 0.01  & 43 & 9 & 600\\

652: \hline

653: \label{oc}

654: \end{tabular}

655: \end{minipage}

656: \end{table}

657: 

658: Figure \ref{tms} presents the lives of the mergers on the main

659: sequence. We see that some low-mass BSs (i. e. $M \le

660: 2.0M_{\odot}$) may exist for about $10^9$ yr, which is long enough

661: to be observed. In most cases, it is in order of $10^8$ yr, which

662: is similar to that of W UMa stars from observations, and therefore

663: we may roughly estimate the contribution to BSs from AS and AR via

664: the number of W UMa systems in a cluster. The estimation, however,

665: is not absolutely since both of the two timescales have wide

666: ranges and large uncertainties, and we cannot rule out {\bf other

667: methods for creating W UMa systems} except for AS and AR. Table 1

668: presents characteristics of some open clusters and the numbers of

669: BSs and W UMa systems in them from observations. We see that

670: $N_{\rm BS}>>N_{\rm W UMa}$ in old clusters (i.e $t \ge 2.0$Gyr)

671: and $N_{\rm BS} \sim N_{\rm W UMa}$ in half of the left clusters,

672: indicating that our models (binary coalescence from AS and AR) are

673: not important  for the produce of BSs in old open clusters, while

674: likely play a critical role in some younger open clusters. In old

675: open clusters, where stellar collisions may be ignored because of

676: low stellar density, AML of low mass binaries is possibly

677: dominated in producing BSs, since the time is long enough for

678: binaries with initial orbital period about 2 d evolving from

679: detached to contact by AML and the mergers may be more massive

680: than the turnoff. Meanwhile, from initial mass functions which

681: have been presented, most stars are concentrated on low mass.

682: Since the individual components almost have not evolved before

683: contact, their mergers from this way have much longer timescales

684: on the main sequences.

685: 

686: \subsection{Binary Samples}

687: To investigate BSs resulting from binary coalescence,

688: we have performed a Monte Carlo simulation

689: where a sample of $10^6$ binaries are considered

690: (very wide binaries are actually single stars)

691: including BSs originated from AS and AR evolution channels.

692: A single starburst is assumed in the simulation,

693: i.e. all the stars have the same age and metallicity ($Z=0.02$).

694: The initial mass function (IMF) of the primary,

695: the initial mass ratio distribution

696: and the distribution of initial orbital separation are as follows:

697: 

698: i) the IMF of Miller \& Scalo \shortcite{ms79} is used

699: and the primary mass is generated from the formula of

700: Eggleton, Fitchett \& Tout \shortcite{egg89}:

701: \begin{equation}

702: M_{\rm 1}=$$0.19X\over (1-X)^{0.75}+0.032(1-X)^{1/4}$$

703: \end{equation}

704: where $X$ is a random number uniformly distributed between 0 and

705: 1. The mass ranges from 0.1 to 100$M_\odot$.

706: 

707: ii)the mass ratio distribution is quite controversial and,

708: for simplity, we only consider a

709: constant mass ratio distribution \cite{maz92}.

710: \begin{equation}

711: n(q)=1,  0\le q \le 1

712: \end{equation}

713: where $q=M_2/M_1$.

714: 

715: iii)We assume that all stars are members of binary systems

716: and the distribution of separations  is constant in log$a$

717: ($a$ is separation).

718: \begin{equation}

719: an(a)=\left\{

720: \begin{array}{ll}

721: \alpha_{\rm sep}(a/a_0)^m, &a \le a_0\\

722: \alpha_{\rm sep}, & a_0<a<a_1\\

723: \end{array}

724:    \right.

725: \end{equation}

726: where $\alpha =0.070, a_0=10R_{\odot},a_1=5.75\times 10^6R_{\odot}=0.13pc$ and

727: $m=1.2$. This distribution gives an equal number of wide binary systems per

728: logarithmic interval and 50 per cent of systems with orbital periods of

729: less than 100 yr.

730: 

731: \subsection{NGC 2682}

732: Some studies show that the metallicity of NGC 2682 is a little different

733: from the solar \cite{car96,fan96},

734: while some other studies concur that it is virtually indistinguishable from

735: solar \cite{hob91,fri93}.

736: So we select this cluster as the first sample to examine our models.

737: The distance modulus $m-M=9.55$

738: and the reddening $E(B-V)=0.022$ \cite{car96,fan96}

739: as we translate the theory results to observations.

740: 

741: Previous studies on the age of this cluster showed several discrepancies.

742: It may range from 3.2{\underline +}0.4 Gyr \cite{bb03} to 6.0 Gyr \cite{jp94}.

743: The study of VandenBerg \& Stetson \shortcite{vs04} derived an age of 4.0 Gyr.

744: In the N-body model of this cluster \cite{hur05},

745: the authors investigated the behaviour around 4 Gyr.

746: As mentioned in section 2,

747: it is also a great uncertainty for the timescale $t_{\rm cc}$.

748: Table 2 and Table 3 present initial parameter spaces

749: from different considerations for both BSs and W UMa systems in our grid.

750: In Table 2,

751: we fixed the cluster age at $t_{\rm cluster}=3.8 \times 10^9$ yr

752: (corresponding to the turnoff of $1.26M_\odot$)

753: and varied $t_{\rm cc}$, i.e  $t_{\rm cc}=1 \times 10^8$ yr,

754: $t_{\rm cc}=5 \times 10^8$ yr and $t_{\rm cc}=1 \times 10^9$ yr.

755: In Table 3, we set $t_{\rm cc}=5 \times 10^8$ yr

756: while the age of the cluster has a width from 3.2 to 4.2 Gyr.

757: then the conditions for valid parameter space to form BSs via case AS or AR

758: are

759: $t_{\rm contact}+t_{\rm cc}\le 4.2\times 10^9$ yr,

760: $t_{\rm contact}+t_{\rm cc}+t_{\rm MS}\ge 3.2\times 10^9$ yr,

761: and the constrains for W UMa systems in the cluster are

762: $t_{\rm contact}\le 4.2\times 10^9$ yr and

763: $t_{\rm contact}+t_{\rm cc}\ge 3.2\times 10^9$ yr.

764: 

765: In Table \ref{1},

766: we see that $t_{\rm cc}$ will remarkably affect on initial parameter space

767: for both of the two kinds of objects.

768: For example, the initial parameter space is much larger for

769: $t_{\rm cc}=1 \times 10^9$ yr

770: than those with other $t_{\rm cc}$.

771: We should bear in mind that, however,

772: the long $t_{\rm cc}$ may be unreasonable from both of observations

773: and theories (see section 2).

774: The long $t_{\rm cc}$ may delay the appearance of the mergers

775: and shorten their timescales on the main sequence,

776: since individual evolution during mergering process are ignored.

777: Especially for the primaries with a very littler hydrogen  in the center

778: at the system contact,

779: the mergers may have never been on the main sequence.

780: 

781: From Table \ref{2}, we see that,

782: because of the existence of an age range,

783: some models might be either mergers or W UMa systems.

784: For large uncertainties of the age of the cluster,

785: the initial parameter defined this way

786: include almost all the models in Table \ref{1}.

787: 

788: %\setcounter{table}{1}

789: \begin{table}

790:  \begin{minipage}{70mm}

791:  \caption{Initial parameters for the mergers and W UMa systems in the grid for the old open cluster M67. We take $t_{\rm cluster}=3.84\times 10^9 {\rm yr}$, corresponding turn-off is $1.26M_\odot$.}

792:  \label{tab2}

793:    \begin{tabular}{cccc}

794: \hline

795: &$t_{\rm cc}=1.0\times 10^8$ yr&\\

796: \hline

797:  &$\scriptstyle M_{\rm 1i}$ & $\scriptstyle {\rm log}q_{\rm i}$ &

798:    $\scriptstyle {\rm log}P_{\rm i}$\\

799: \hline

800: the mergers &  1.26 & 0.15 & 0.1\\

801: &&&\\

802: W UMa       &  1.26 & 0.05 & 0.2\\

803: \hline

804: &$t_{\rm cc}=5.0\times 10^8$ yr&&\\

805: \hline

806: the mergers &  1.26&  0.25&  0.05\\

807:             &  1.41&  0.15&  0.2\\

808: &&&\\

809: W UMa       &  1.26&  0.05&  0.2\\

810:             &  1.26&  0.15&  0.1\\

811: \hline

812: &$t_{\rm cc}=1.0\times 10^9$ yr&&\\

813: \hline

814: the mergers&1.12&  0.05&  0.05\\

815:            &1.12&  0.3 &  0.05\\

816:            &1.26&  0.05&  0.15\\

817:            &1.26&  0.1 &  0.1\\

818:            &1.26&  0.25&  0.05\\

819:            &1.26&  0.35&  0.15\\

820:            &1.26&  0.4 & 0.15\\

821:            &1.41&  0.2 & 0.15\\

822:            &1.41&  0.3 & 0.15\\

823: &&&\\

824: W UMa      &1.26&  0.05&  0.2\\

825:            &1.26&  0.15&  0.1\\

826:            &1.41&  0.15&  0.2\\

827:            &1.41&  0.25&  0.15\\

828: \hline

829: \label{1}

830: \end{tabular}

831: \end{minipage}

832: \end{table}

833: 

834: %\setcounter{table}{2}

835: \begin{table}

836:  \begin{minipage}{70mm}

837:  \caption{Initial parameters for the mergers and W UMa systems in the grid

838: for the old open cluster M67

839: ($t=3.2-4.2$ Gyr and $t_{\rm cc}=5\times 10^8$ yr). }

840:  \label{tab2}

841:    \begin{tabular}{ccccccccc}

842: \hline

843: & the mergers &&&&& W UMa&&\\

844: \hline

845:  $\scriptstyle M_{\rm 1i}$ & $\scriptstyle {\rm log}q_{\rm i}$ &

846:    $\scriptstyle {\rm log}P_{\rm i}$&&&

847: $\scriptstyle M_{\rm 1i}$ & $\scriptstyle {\rm log}q_{\rm i}$ &

848:    $\scriptstyle {\rm log}P_{\rm i}$\\

849: \hline

850:   1.12&  0.05&  0.05&&&   1.12&  0.1&  0.05\\

851:   1.12&  0.3&  0.05&&&  \\

852:   1.26&  0.05&  0.15&&&  1.26&  0.05&  0.15\\

853:   1.26&  0.1&  0.1&&&  1.26&  0.05&  0.2\\

854:   1.26&  0.15& 0.1&&& 1.26&  0.1&  0.1\\

855:   1.26&  0.25&  0.05&&& 1.26&  0.15&  0.1\\

856:   1.26&  0.35&  0.15&&& 1.26&  0.3&  0.1\\

857:   1.26&  0.4&  0.15&&&  \\

858:   1.41&  0.05&  0.25&&& 1.41&  0.15&  0.2\\

859:   1.41&  0.15&  0.2&&&  1.41&  0.2&  0.15\\

860:   1.41&  0.2&  0.15&&&  1.41&  0.25&  0.15\\

861:   1.41&  0.25&  0.15&&& 1.41&  0.25&  0.2\\

862:   1.41&  0.3&  0.15&&&   1.41&  0.3&  0.2\\

863:   1.41&  0.35&  0.3&&& \\

864:   1.41&  0.4&  0.3&&& \\

865: 

866: \hline

867: \label{2}

868: \end{tabular}

869: \end{minipage}

870: \end{table}

871: 

872: Some models with $M_{\rm 1i}=1.41M_\odot$ appear in Table \ref{1}.

873: We checked the evolutionary details of them in Table A1

874: and find that

875: the system

876: ($M_{\rm 1i}, {\rm log}q_{\rm i},{\rm log}(P_{\rm 0}/P_{\rm ZAMS})$)=

877: (1.41, 0.15, 0.2) begins RLOF at  1.4 Gyr and becomes contact at 3.3 Gyr.

878: As $t_{\rm cc}=5\times10^8$ yrs, the merger is formed at 3.8 Gyr with

879: $X_{\rm Hcm}=0.0003$ after self-adjustment

880: and will leave the main sequence after 0.45 Gyr.

881: As discussed in section 3,

882: the merger of this system is very likely to be a subgiant when it formed.

883: For other binaries with $M_{\rm 1i}=1.41M_\odot$ in Table \ref{1},

884: though the primary is not near to zero,

885: relatively long contact timescales ($1\times10^9$)

886: lead the mergers be formed at about 3.5 to 3.8 Gyr.

887: So they are still on the main sequence at the cluster age.

888: 

889: Now we consider another case

890: which is related to the BSs in a binary BS+MS in this cluster.

891: It means that the secondary is a BS

892: before the system being contact.

893: The parent binary for this case should fulfill the conditions as follows:

894: (a) $t_{\rm contact} > t_{\rm cluster}> t_{\rm RLOF}$;

895: (b) at $t_{\rm cluster}$, $M_{\rm 2}> M_{\rm to}$.

896: Only one model is found in our results to possibly produce BS+MS in M67.

897: The initial parameters for the model are

898: $(M_{\rm 1i},{\rm log}q_{\rm i}, {\rm log}P_{\rm i})=(1.41,0.25,0.2)$.

899: At the age $3.82 \times 10^9$ yr, the secondary is $1.26M_\odot$

900: and the system comes into contact at $4\times 10^9$ yr

901: with an orbital period of 0.5276 d.

902: It means that the BS+MS may exist at least for $2 \times 10^8$yr.

903: A system like this should be near the turn-off if only the secondary (BS)

904: is considered. However no observed BSs in this region are located in a binary

905: like this. The evolved component may contribute to some luminosity

906: \cite{tian06}.

907: 

908: \begin{table}

909:  \begin{minipage}{70mm}

910:  \caption{Consequences of Monte Carlo simulation at different conditions

911: for binaries undergoing AS and AR evolutionary channels in NGC 2682 (M67).}

912: % \label{}

913:    \begin{tabular}{cccc}

914: \hline

915: & conditions & $N_{\rm BS}$&$N_{\rm W UMa}$\\

916: \hline

917: case 1 & $t_{\rm cc}=1 \times 10^8$yr & 5 &3 \\

918:        & $t_{\rm cluster}=4\times 10^9$ yr&&\\

919: \hline

920: case 2 & $t_{\rm cc}=5 \times 10^8$yr & 5 & 10\\

921:        & $t_{\rm cluster}=4 \times 10^9$ yr&&\\

922: \hline

923: case 3 & $t_{\rm cc}=1 \times 10^9$yr & 14& 15\\

924:        & $t_{\rm cluster}=4 \times 10^9$ yr&&\\

925: \hline

926: case 4 & $t_{\rm cc}=1 \times 10^8$yr & 20&13 \\

927:        & $t_{\rm cluster}=3.4-4.2\times 10^9$ yr&&\\

928: \hline

929: case 5 & $t_{\rm cc}=5 \times 10^8$yr & 25&19\\

930:        & $t_{\rm cluster}=3.4-4.2 \times 10^9$ yr&&\\

931: \hline

932: case 6 & $t_{\rm cc}=1 \times 10^9$yr & 38 & 36 \\

933:        & $t_{\rm cluster}=3.4-4.2 \times 10^9$ yr&&\\

934: \hline

935: \label{3}

936: \end{tabular}

937: \end{minipage}

938: \end{table}

939: 

940: 

941: \begin{table}

942:  \begin{minipage}{70mm}

943:  \caption{AML Consequences for NGC 2682.}

944: % \label{}

945:    \begin{tabular}{cccc}

946: \hline

947: & conditions & $N_{\rm BS}$&$N_{\rm W UMa}$\\

948: \hline

949: case 1 & $t_{\rm cc}=1 \times 10^8$yr & 855 &14 \\

950:        & $t_{\rm cluster}=4\times 10^9$ yr&&\\

951: \hline

952: case 2 & $t_{\rm cc}=5 \times 10^8$yr & 964 & 84\\

953:        & $t_{\rm cluster}=4 \times 10^9$ yr&&\\

954: \hline

955: case 3 & $t_{\rm cc}=1 \times 10^9$yr & 1079& 204\\

956:        & $t_{\rm cluster}=4 \times 10^9$ yr&&\\

957: \hline

958: case 4 & $t_{\rm cc}=0 \times 10^8$yr & 824 &0 \\

959:        & $t_{\rm cluster}=4 \times 10^9$ yr&&\\

960: \hline

961: case 5 & $t_{\rm cc}=5 \times 10^8$yr & 996&120\\

962:        & $t_{\rm cluster}=3.5 \times 10^9$ yr&&\\

963: \hline

964: case 6 & $t_{\rm cc}=5 \times 10^8$yr & 1194 & 109 \\

965:        & $t_{\rm cluster}=4.5\times 10^9$ yr&&\\

966: \hline

967: \label{aml}

968: \end{tabular}

969: \end{minipage}

970: \end{table}

971: 

972: \begin{figure*}

973: \centerline{\psfig{figure=montec.ps,width=16cm,angle=0}}

974: \caption{Monte-Carlo simulation results for M67 at different conditions

975: (Table \ref{3}.) `$\times $' represents the models from grid calculation,

976: open circles show the models in Table \ref{1} and the dots give the

977: possible BSs. The observed BSs from  Sandquist \& Shetrone (2003)

978: are also plotted in the figure.}

979: \label{montec}

980: \end{figure*}

981: 

982: By interpolating from Table A1,

983: we get different numbers of BS ($N_{\rm BS}$) under various conditions

984: (Table \ref{3}).

985: The location of the produced BSs on a CMD are presented in Fig.\ref{montec}.

986: 

987: As mentioned before, AML of low mass binaries probably makes a

988: major contribution to BSs in this old open cluster. For low mass

989: binaries, the individual components have almost not evolved (very

990: close to ZAMS) before contact, and therefore their mergers may be

991: replaced with ZAMS models with a mass of the parent binaries.

992: After coalescence, however, the mergers are much more massive and

993: their evolutions cannot be negected agian. To examine the effect

994: of AML in this cluster, we simply deal with low-mass binaries

995: ($M_{\rm 1i} <1M_\odot$) as follows:

996: 

997: (1) From the binary sample, we found out some binaries with

998: $M_{\rm 1i}>1/2M_{\rm to}$ and with the orbital period less than

999: 0.5 d at the cluster age by AML. The lower mass limit $1/2M_{\rm

1000: to}$ is sure that the mergers are probably larger than the

1001: turnoff, and the constrain on the orbital period leads the

1002: binaries possibly be contact before or at the cluster age. A

1003: semi-empirical formula for the orbital period variation is adopted

1004: here \cite{ste06}:

1005: \begin{equation}

1006: $${\rm d} P_{\rm orb} \over {\rm d} t$$ = -(2.6+1.3) \times 10^{-10}P_{\rm orb}^{-1/3}e^{-0.2P_{\rm orb}}

1007: \end{equation}

1008: where $P_{\rm orb}$ in days and time in years. For very short

1009: orbital periods the exponential factor is close to unity and

1010: varies very little during the subsequent evolution of the orbital

1011: period of the binaries. So it is ignored in this paper.

1012: 

1013: (2)Find out the time, $t_{\rm RLOF}$, at which the primary is full

1014: of its Roche lobe for the selected binaries from step 1. A rapid

1015: binary evolution code written by Hurley, Pols \& Tout (2002) is

1016: employed here and the Roche lobe of the primary $R_{\rm cr1}$ is

1017: calculated by \cite{egg83}

1018: \begin{equation}

1019: R_{\rm cr1}/A=$$ 0.49q^{2/3} \over 0.6q^{2/3}+{\rm

1020: ln}(1+q^{1/3})$$ ,

1021: \end{equation}

1022: where A is the separation and $q=M_{\rm 1}/M_{\rm 2}$

1023: 

1024: (3)Evolve the mergers to the cluster age. The mergers are ZAMS

1025: models instead with a mass of the systems. The starting age is

1026: $t_{\rm RLOF}$, since the systems will {\bf reach} contact very

1027: quickly from the onset of RLOF because of AML. Different

1028: timescales $t_{\rm cc }$ are adopted here. Some of the binaries

1029: from step 1 have left the main sequence, but most of them remain

1030: on the main sequence.

1031: 

1032: Table 5 gives the results from different cluster age assumptions

1033: and different $t_{\rm cc}$, and the magnitudes and colors for the

1034: mergers from these cases are also presented in Fig. 11.

1035: 

1036: In Fig.\ref{montec}, we see that, the products from binary

1037: evolution (AS and AR in the paper) and those from AML cover

1038: different regions. The mergers from conservative evolution are

1039: located in the region of high luminosity, i.e far away from the

1040: the turnoff, and the remnants from AML occupy the region with low

1041: luminosity and have a scatter on the color. The figure also

1042: indicates that AML of low-mass binaries is much more important in

1043: this old open cluster. In some cases (case 3 to 6), the produced

1044: BSs from conservative evolutionary appears a certain width with

1045: decreasing mass. Another phenomena worth noticing is in case 6,

1046: the produced BSs result extends very close to F81, the observed

1047: brightest BSs in this cluster. Further study shows that the lower

1048: limit of the cluster age may be 3.8 Gyr for $t_{\rm

1049: cc}=1\times10^9$ yr to result in the produced BSs extending to the

1050: position of F81.

1051: 

1052: As shown in Table 4, we only obtained 38 BSs from a $10^6$ binary

1053: sample in the widest condition. NGC 2682 is an open cluster and

1054: has a much less stars than $10^6$. In Hurley et al.

1055: \shortcite{hur05}, 12000 single stars and 12000 binaries

1056: ($N=24000$) are adopted in the best fitting model. Multipying this

1057: factor, we just got 1 BS with high luminosity. There are 3 BSs

1058: from the observations, however, located in the region covered by

1059: the mergers from AS and AR. The number of stars initially in this

1060: cluster is important here, but it is difficult for us to estimate

1061: it because of the incompleteness of observations as determing the

1062: current mass $M$ and the uncertainties of mass loss history of the

1063: cluster when converting $M$ to an initial mass $M_{\rm 0}$. IMF

1064: may also affect the results. We adopt the field single star IMF

1065: for the primaries and the lower mass limit is $0.1M_\odot$ in the

1066: binary sample, however, M67 is an old open cluster and rich in

1067: binaries, and then it possibly has different IMF as well as

1068: different lower mass limit as described by Chabrier

1069: \shortcite{cha03}.

1070: 

1071: In the region with low luminosity, i.e the region covered by the

1072: mergers of AML, the numbers shown in Table 5 are enough to account

1073: for BSs in this region. Dynamically unstable mass transfer (AD) is

1074: also possible to form BSs in this region, since the binaries

1075: undergoing AD have larger initial mass ratios in general, leading

1076: the smaller mergers. According to the result of Nelson \& Eggleton

1077: (2001), we roughly estimate the number of binaries undergoing AD

1078: evolution with initial primary mass between 0.89 and $1.26M_\odot$

1079: as $N=10^6$. It is about 41--63 and the mass of the binaries are

1080: less than  $1.36M_{\rm to}$, indicating that AD may produce BSs

1081: near the turnoff indeed, but the contribution is {\bf much smaller

1082: than that of AML in low-mass binaries.}

1083: 

1084: In  Hurley et al.\shortcite{hur05},

1085: the authors obtained seven BSs from case A mass transfer leading to coalescence

1086: by unperturbed evolution.

1087: The age of the mergers in that paper were calculated based on the assumption

1088: that core hydrogen burning proceeds uniformly and

1089: that the end of the main sequence is reached when  10 per cent of the total

1090: hydrogen has been burnt.

1091: These assumptions for calculating the MS lives of the mergers

1092: are questionable as they mentioned in the paper.

1093: According to our calculations,

1094: almost all of the mergers have relatively short main-sequence lives,

1095: i. e. less than $10^9$ yr (see Fig.\ref{tms}).

1096: It means that five of the seven BSs are likely to leave the main sequence

1097: and no longer be observed as BSs at 4 Gyr.

1098: The range of age for M67  is likely to improve the result.

1099: 

1100: \subsection{NGC 2660}

1101: \begin{table*}

1102:  \begin{minipage}{150mm}

1103:  \caption{Consequences of Monte Carlo simulations from different assumptions

1104: for binaries undergoing AS and AR evolutionary channel in NGC 2660.}

1105: % \label{}

1106:    \begin{tabular}{cccccc}

1107: \hline

1108: & conditions & $N_{\rm BS}$&$N_{\rm W UMa}$& $N_{\rm BS}$&$N_{\rm W UMa}$ \\

1109: &&$(M_{\rm L}=0.8M_\odot)$&$(M_{\rm L}=0.8M_\odot)$&$(M_{\rm L}=0.1M_\odot)$&$(M_{\rm L}=0.1M_\odot)$\\

1110: \hline

1111: case 1 & $t_{\rm cluster}=1.2 \times 10^9$, $t_{\rm cc}=5 \times 10^7$  yr & 268 & 59 & 70 & 11 \\

1112: \hline

1113: case 2 & $t_{\rm cluster}=1.2 \times 10^9$, $t_{\rm cc}=1 \times 10^8$  yr &284&117&71 & 21\\

1114: \hline

1115: case 3 & $t_{\rm cluster}=1.2 \times 10^9$, $t_{\rm cc}=5 \times 10^8$  yr &296&631&72& 126\\

1116: \hline

1117: case 4 & $t_{\rm cluster}=1.2 \times 10^9$, $t_{\rm cc}=1 \times 10^9$  yr &87&1038&21&220\\

1118: \hline

1119: case 5 & $t_{\rm cluster}=0.9-1.2 \times 10^9$, $t_{\rm cc}=5 \times 10^7$  yr &693&454&141&87\\

1120: \hline

1121: case 6 & $t_{\rm cluster}=0.9-1.2 \times 10^9$, $t_{\rm cc}=1 \times 10^8$  yr &703&525&149&106\\

1122: \hline

1123: case 7 & $t_{\rm cluster}=0.9-1.2 \times 10^9$, $t_{\rm cc}=5 \times 10^8$  yr &445&934&102&194\\

1124: \hline

1125: case 8 & $t_{\rm cluster}=0.9-1.2 \times 10^9$, $t_{\rm cc}=1 \times 10^9$  yr &87&1125&21&241\\

1126: \hline

1127: \label{2660t}

1128: \end{tabular}

1129: \end{minipage}

1130: \end{table*}

1131: 

1132: NGC 2660 is an intermediate-age open cluster with a solar

1133: metallicity. The early study by Hartwick \& Hesser (1973) showed

1134: the following properties for this cluster: $E(B-V)=0.38$,

1135: $(m-M)_{\rm 0}=12.3 \pm 0.3$, age $\sim 1.2$ Gyr, metallicity

1136: similar to the Hyades, and high possibility of membership for the

1137: N-type carbon star. There are some uncertainties for the

1138: determination of the cluster age, e.g. Lynga (1987) cited 1.6 Gyr,

1139: Jane \& Phelps (1994) gave an age of 0.9 Gyr, while Carrio \&

1140: Chiosi (1994) derived 0.7 Gyr. The latest report on this cluster

1141: is Sandrelli et al. (1999). It was shown that, metallicity about

1142: solar, $(m-M)_{\rm 0}=12.1 -12.3$, $E(B-V)=0.37-0.42$, age $ \le

1143: 1Gyr$, with a fraction of binaries of about 30 per cent.

1144: 

1145: According to the new catalogue of blue stragglers in open clusters

1146: \cite{al07}, there are 8 BSs in this cluster while the cluster age

1147: is ${\rm log}t=9.03$ (about 1.1 Gyr). The turnoff is around

1148: $2M_\odot$ for this age, indicating that AML in low-mass binaries

1149: has little influence on the product of BSs in this cluster. Table

1150: 6 presents our simulation results from different assumptions of

1151: the cluster age and $t_{\rm cc}$ by interpolating in Table A1.

1152: 

1153: Though we show the simulation results for $t_{\rm cc}=5 \times

1154: 10^8$ and $t_{\rm cc}=1 \times 10^9$ yr in Table \ref{2660t}, we

1155: should bear in mind that NGC 2660 is an intermediate-age cluster

1156: and a long $t_{\rm cc}$ is unreasonable for it from both

1157: observations and our models. Such a long  $t_{\rm cc}$ as $5

1158: \times 10^8$ or $1 \times 10^9$ yr will inevitably cause large

1159: differences of both components of a binary in our models, and

1160: therefore a shorter $t_{\rm cc}$, namely, $t_{\rm cc}=1 \times

1161: 10^8$ and

1162:  $t_{\rm cc}=5 \times 10^7$ yr,

1163: is more appropriate for this cluster.

1164: From Table \ref{2660t}

1165: we see that BS number is not sensitive to $t_{\rm cc}$ in this cluster,

1166: but the contact systems are very sensitive to $t_{\rm cc}$,

1167: since a longer $t_{\rm cc}$

1168: indicates a larger parameter space for these systems.

1169: Contact binaries exist in the cluster for all the cases in Table \ref{2660t},

1170: but it is less than the BSs number produced from AS and AR

1171: when $t_{\rm cc}=1 \times 10^8$ and $t_{\rm cc}=5 \times 10^7$ yr,

1172: especially in case 1,

1173: it is just of 1/6 to 1/5 of that of BSs.

1174: 

1175: There are 150 stars on the main sequence to two magnitudes below the turnoff

1176: in NGC 2660 and 175 in NGC 2682,

1177: as shown in Ahumada \& Lapasset \shortcite{al07}.

1178: From this comparison,

1179: our models may produce the necessary BSs in

1180: NGC 2660 when $M_{\rm L}=0.8M_\odot$.

1181: If $M_{\rm L}=0.1M_\odot$ for the binary sample, however,

1182: the produced BSs number from AS and AR is just 1/5 to 1/4 of that from

1183: $M_{\rm L}=0.8M_\odot$.

1184: Normalization is necessary here to estimate the BSs birthrate from our models

1185: in this cluster

1186: and the same problem mentioned in section 4.2 appears here again.

1187: Nevertheless,

1188: our models may explain several BSs in this intermediate age cluster.

1189: 

1190: Figure \ref{2660} shows some examples from Monte Carlo simulation on CMD.

1191: The observed BSs \cite{al07} and an isochrone ($t=1.2 Gyr$)

1192: are also plotted in the figure.

1193: Several BSs are located in the region of our models from AS and AR.

1194: In the region with the most chance from our models $M_{\rm v} \approx 1.0$,

1195: however,

1196: there are no BSs observed at present.

1197: The observed peak appears about $M_{\rm v} \approx 2.0$, one magnitude

1198: lower than the theoretical value.

1199: Mass loss during merger process may shorten the discrepancies,

1200: but we should find out an appropriate mechanism to explain

1201: such a large mass loss about $0.5M_\odot$,

1202: which is much larger than that from observations

1203: and from smooth particle hydrodynamic simulations.

1204: 

1205: Though the BSs number (or birthrate) from our models is not

1206: sensitive to $t_{\rm cc}$, the initial parameter space will change

1207: with  $t_{\rm cc}$, resulting in some differences for the mergers.

1208: In the upper right panel, we show the results from case 1 to case

1209: 3 with $M_{\rm L}=0.1M_\odot$. As seen in this figure, the mergers

1210: from a long $t_{\rm cc}$, i.e. $5\times 10^8$ yr, extend to a

1211: higher luminosity and concentrated on the ZAMS, but small

1212: differences appear between $t_{\rm cc}=1\times 10^8$ yr and

1213: $t_{\rm cc}=5\times 10^7$ yr.

1214: 

1215: {\bf We have not considered the effect of AML in this cluster,

1216: since we simply assume that the conservation assumptions are

1217: reasonable for stars with spectra from G0 (about $1.05M_\odot$) to

1218: B1 according to Nelson \& Eggleton \shortcite{nel01}. In fact, a

1219: main sequence star can generate a magnetic field as long as it has

1220: a sufficiently thick convective envelop, i.e. with a spectrum

1221: later than F8. In the study of Andronov, Pinsonneault \& Terndrup

1222: \shortcite{apt06}, the mass threshold for AML is between 1.2 and

1223: $1.4M_\odot$, then the total mass of the merger might be up to 2.4

1224: and $2.8M_\odot$, above the turnoff of this cluster. So it could

1225: be possible that AML is also impacting the blue straggler

1226: population in this intermediate-age cluster. This provides another

1227: possible explanation for the magnitude offset from the models and

1228: observations.}

1229: 

1230: \begin{figure*}

1231: \centerline{\psfig{figure=2660.ps,width=16cm,angle=0}}

1232: \caption{Monte-Carlo simulation results for NGC 2660 from

1233: different assumptions (Table \ref{2660t}). The dots show the blue

1234: stragglers listed by Ahumada \& Lapasset (2007), and the error

1235: bars are from Sandrelli et a. (1999), i.e. $m-M=12.1-12.3$,

1236: $E(B-V)=0.37-0.42$. The stars present the positions of BSs when

1237: $m-M=12.2$, $E(B-V)=0.31$ as shown in Ahumada \& Lapasset (2007).

1238: In the upper right panel, the circles, dots and triangles are for

1239: $t_{\rm cc}$ equal to $5 \times 10^7$, $1 \times 10^8$ and $5

1240: \times 10^8$, respectively. } \label{2660}

1241: \end{figure*}

1242: 

1243: \section{DISCUSSIONS AND CONCLUSIONS}

1244: In Sect.4, we notice that the timescale from contact to complete

1245: coalescence, $t_{\rm cc}$, strongly affects the initial parameter

1246: space of primordial binaries which eventually produce single BSs

1247: in a cluster. On the other hand, there are {\bf some conflicting

1248: estimates} for $t_{\rm cc}$ from observations and theoretical

1249: models. {\bf In this paper we adopt empirical values, i.e. $t_{\rm

1250: cc}$ is short in comparison to the evolution timescale of both

1251: components in a binary, and ignore the changes during the merger

1252: process. In this section, we will first discuss the consequences

1253: of a long $t_{\rm cc}$. Because of evolution of both components

1254: during the merger process, the primaries have lower central

1255: hydrogen content and the matter from the secondaries have larger

1256: He content. The former results in a redder colour for the mergers

1257: while the latter makes the mergers bluer. So the final positions

1258: of the mergers are possibly similar to those shown in this paper,

1259: except that the primaries have left the main sequence at final

1260: coalescence. This case will appear in the mergers with little

1261: central hydrogen. For example, a star with $2M_\odot$ may evolve

1262: from ZAMS to exhausted of central hydrogen in $10^9$ yr, and then

1263: none of the mergers from binaries with primary' masses larger than

1264: $2M_\odot$ will be on the main sequence if $t_{\rm cc}= 1 \times

1265: 10^9$ yr. For the primaries with very little hydrogen in the

1266: center at contact, the mergers may never be on the main sequence

1267: even in the cases of short $t_{\rm cc}$. The long $t_{\rm cc}$

1268: also delays the appearance of the mergers and shortens their

1269: timescales on the main sequence. The latter has not been exactly

1270: expressed in our models, and therefore we just see that the

1271: mergers from a long $t_{\rm cc}$ have larger luminosities as shown

1272: in section 4.}

1273: 

1274: {\bf In our binary evolutions, we have not included AML, which

1275: exists in low-mass binaries and may be the main course making the

1276: binaries change from detached to contact and finally coalesce,

1277: resulting in a large contribution to BSs in old clusters, e.g NGC

1278: 2682. In young and intermediate-age clusters, however, AML has

1279: little contribution to the birthrate of BSs, since (a) the time is

1280: not long enough for binaries to go from detached to contact and

1281: (b) the mass of the mergers is probably less than the turnoff of

1282: the cluster even though their parents may coalesce in the cluster

1283: age. So, we simply estimated the effect of AML in NGC 2682 while

1284: negecting it in NGC 2660.}

1285: 

1286: The mass loss during the merger process can also affect our

1287: result, mainly the location on the CMD of the products. As shown

1288: in NGC 2660, no BSs have been observed in the region with the most

1289: opportunity from our models. Because of mass loss, the mergers

1290: will be fainter than those given in the paper. However the

1291: faintness will be slight since the mass loss is not vast from both

1292: observations and smooth particle hydrodynamic simulations

1293: \cite{lrs96,sill97,sill01}. The lost mass may carry some angular

1294: momentum out from the parent binary. By analyzing the BS spectra

1295: from Hubble Space Telescope (there is an apparent continuum

1296: deficit on the short-wavelength side of Balmer discontinuity ), De

1297: Marco et al. \shortcite{dm04} argued that some BSs might be

1298: surrounded by a circumstellar disk. However, Porter \& Townsend

1299: \shortcite{pt05} showed that the flux deficits may be attributed

1300: wholly to rapid rotation. The rotation rates needed are of the

1301: order of those found in the study of De Macro et al.

1302: \shortcite{dm05}. Whether the flux deficits shortward of the

1303: Balmer jump are induced by a circumstellar disk or rapid rotation,

1304: it provides a possible explanation for the orbital angular

1305: momentum of the system after coalescence. Such a large mass loss

1306: as shown in NGC 2660 (about $0.5 M_\odot$), however, is a problem

1307: and should be explained reasonably in physics.

1308: 

1309: Based on some assumptions, we studied the mergers of close

1310: binaries from AS and AR evolution by detailed evolutionary

1311: calculations. The products from our models may stay on the left of

1312: the ZAMS and have no central concentration with decreasing mass.

1313: Because of the {\bf development} of the convective core, the

1314: mergers with little central hydrogen (less than 0.01) in our

1315: models have unusually long timescales on the main sequence ($\sim

1316: 10^8$ yrs). These objects are probably subgiants as they are

1317: formed, since the primaries in the progenitors also have little

1318: central hydrogen and may have left the main sequence during merger

1319: process.

1320: 

1321: The mergers from our models stay on the main sequence for a

1322: timescale in order of $10^8$ yrs. Some low-mass mergers may stay

1323: on the MS for about $10^9$ yrs. The timescale is similar to that

1324: of W UMa stars from observations, and therefore we may roughly

1325: estimate the contribution to BSs from AS and AR via the number of

1326: W UMa systems in a cluster. The estimation, however, is not

1327: absolutely since both of the two timescales have wide ranges and

1328: large uncertainties, and we cannot rule out other methods for

1329: creating W UMa systems except for AS and AR. Comparison to

1330: observations indicates that our models (binary coalescence from AS

1331: and AR) are not important for the produce of BSs in old open

1332: clusters, while likely play a critical role in some younger open

1333: clusters.

1334: 

1335: We performed Monte Carlo simulations to examine our models in an

1336: old open cluster NGC 2682 and in an intermediate-age cluster NGC

1337: 2660. The effect of AML was estimated in NGC 2682 in a simple way,

1338: where the mergers are replaced with ZAMS models. In NGC 2682,

1339: binary mergers from our models cover the region with high

1340: luminosity and those from AML are located in the region with low

1341: luminosity, existing a certain width. The BSs from AML are much

1342: more than those from our models, indicating that AML of low mass

1343: binaries makes a major contribution to BSs in this cluster. Our

1344: models are corresponding for several BSs in NGC 2660. In the

1345: region with the most opportunity on CMD, however, no BSs have been

1346: observed. {\bf Our results are well-matched to the observations if

1347: there are $\sim 0.5M_\odot$ of mass loss in the merger process,

1348: but a physical mechanism for this much mass loss is a problem.}

1349: 

1350: %For a given binary with $M_{\rm 1i}$, $q_{\rm i}$ and $P_{\rm i}$,

1351: %if it undergoes the AS or AR evolution,

1352: %we may get information on its RLOF,

1353: %the structures and evolutionary consequences of the mergers

1354: %by interpolating in Table A1.

1355: %Meanwhile $M_{\rm v}$ and $B-V$ at maximum central hydrogen content

1356: %may be obtained from eqs.(6) and (7).

1357: %Hence, we may roughly estimate that the binary evolves to a BS or not

1358: %in a cluster and the character of the merger,

1359: %and calculate its luminosity function.

1360: 

1361: 

1362: \section{ACKNOWLEDGMENTS}

1363: The authors thank R. S. Pokorny for his improvement in language.

1364: This work is supported by

1365: the Chinese National Science Foundation (Grant Nos. 06GJ061001 and 10433030),

1366: the Yunnan Natural Science Foundation (Grant No. 2004A0022Q) and

1367: the Chinese Academy of Sciences (Grant No. O6YQ011001).

1368: 

1369: \begin{thebibliography}{}

1370: \bibitem[\protect\citename{Ahumada \& Lapasset }2007]{al07}

1371: Ahumada J. A., Lapasset E., 2007, A\&A, 463, 789

1372: \bibitem[\protect\citename{Andronov, Pinsonneault \& Terndrup}2006]{apt06}

1373: Andronov N., Pinsonneault M. H., Terndrup D. M., 2006, ApJ, 646, 1160

1374: \bibitem[\protect\citename{Bilir et al.}2005]{bil05}

1375: Bilir S., Karatas Y., Demircan O., Eker Z., 2005, MNRAS, 357, 497

1376: \bibitem[\protect\citename{Bonatto \& Bica }2003]{bb03}

1377: Bonatto Ch., Bica E., 2003, A\&A, 405, 525

1378: \bibitem[\protect\citename{Carraro et al. }1996]{car96}

1379: Carraro G., Girardi L., Bressan A., Chiosi C., 1996, A\&A, 305, 849

1380:  \bibitem[\protect\citename{Chabrier }2003]{cha03}

1381: Chabrier G., 2003, PASP, 115, 763

1382:  \bibitem[\protect\citename{Chen \& Han }2004]{ch04}

1383: Chen X., Han Z., 2004, MNRAS, 355, 1182

1384:  \bibitem[\protect\citename{Chen \& Han }2005]{ch05}

1385: Chen X., Han Z., 2005, ChJAA, 5, 65

1386:  \bibitem[\protect\citename{Davies, Piotto \& Angeli }2004]{dpa04}

1387: Davies M. B., Piotto G., De Angeli F., 2004, MNRAS, 349, 129

1388:  \bibitem[\protect\citename{De Marco et al. }2004]{dm04}

1389: De Macro, O., Lanz, T., Oueellette, J. A., Zurek, D., Shara, M. M., 2004, ApJ, 606, L151

1390:  \bibitem[\protect\citename{De Marco et al. }2005]{dm05}

1391: De Macro, O., Shara, M. M., Zurek, D., L Oueellette, J. A., Lanz, T.,Saffer, R. A., Sepinsky, J. F., 2005, ApJ, 632, 894

1392:  \bibitem[\protect\citename{Demircan et al.} 2006]{dek06}

1393: Demircan O., Eker Z., Karatas Y. et al., 2006, MNRAS, 366, 2511

1394:  \bibitem[\protect\citename{Dryomova \& Svechnikov} 2002]{dry02}

1395: Dryomova G. N., Svechnikov M. A., 2002, Ap, 45, 158

1396:  \bibitem[\protect\citename{Eggen \& Iben }1989]{eggen89}

1397:   Eggen O. J., Iben I. Jr., 1989, AJ, 97, 431

1398:  \bibitem[\protect\citename{Eggleton }1971]{egg71}

1399:   Eggleton P.P., 1971, MNRAS, 151, 351

1400:  \bibitem[\protect\citename{Eggleton }1972]{egg72}

1401:   Eggleton P.P., 1972, MNRAS, 156, 361

1402:  \bibitem[\protect\citename{Eggleton }1973]{egg73}

1403:   Eggleton P.P., 1973, MNRAS, 163, 179

1404:  \bibitem[\protect\citename{Eggleton }1983]{egg83}

1405:   Eggleton P.P., 1983, ApJ, 268, 368

1406:  \bibitem[\protect\citename{Eggleton, Fitchett \& Tout }1989]{egg89}

1407:   Eggleton P.P., Fitchett M. J., Tout C. A., 1989, ApJ, 347, 998

1408:  \bibitem[\protect\citename{Eggleton }2000]{egg00}

1409:   Eggleton P.P., 2000, NewAR, 44, 111

1410:  \bibitem[\protect\citename{Fan et al. }1996]{fan96}

1411:   Fan X. et al., 1996, AJ, 112, 628

1412:  \bibitem[\protect\citename{Ferraro et al. }2003]{fer03}

1413: Ferraro F. R., Sills A., Rood R. T., Paltrinieri B., Buonanno R., 2003, ApJ, 588, 464

1414:  \bibitem[\protect\citename{Friel \& Janes }1993]{fri93}

1415:   Friel E. D., Janes K. A., 1993, A\&A, 267, 75

1416:  \bibitem[\protect\citename{Han et al.\ }1994]{han94}

1417:   Han Z., Podsiadlowski Ph., Eggleton P.P., 1994, MNRAS, 270, 121

1418:  \bibitem[\protect\citename{Hobbs \& Thorburn }1991]{hob91}

1419:   Hobbs L. M., Thorburn J. A., 1991, AJ, 102, 1071

1420:  \bibitem[\protect\citename{Hurley et al.\ }2005]{hur05}

1421:   Hurley J. R., Pols O. R., Aarseth S. J., Tout C. A., 2005, MNRAS, 363, 293

1422:  \bibitem[\protect\citename{Hurley, Tout \& Pols }2002]{hur02}

1423:   Hurley J. R., Tout C. A., Pols O. R., 2002, MNRAS, 329,897

1424:  \bibitem[\protect\citename{Iben \& Livio }1993]{il93}

1425: Iben, Jr. Ic., Livio, M., 1993, PASP, 105, 1373

1426:  \bibitem[\protect\citename{Janes \& Phelps }1994]{jp94}

1427: Janes K. A., Phelps R. L., 1994, AJ, 108, 1773

1428:  \bibitem[\protect\citename{Kafka et al. }2004]{kaf04}

1429: Kafka S., Gibbs, D. G., Henden A. A., et al. 2004, AJ, 127, 1622

1430:  \bibitem[\protect\citename{Kippenhanhn, Ruschenplatt \& Thomas}1980]{kip80}

1431:   Kippenhanhn, R., Ruschenplatt, G., Thomas, H.-C.,1980, A\&A, 91, 175

1432:  \bibitem[\protect\citename{Lejeune, Cuisinier \& Buesel} 1997]{lcb97}

1433: Lejeune T., Cuisinier F., Buesel R., 1997, A\&AS, 125, 229

1434:  \bibitem[\protect\citename{Lejeune, Cuisinier \& Buesel} 1998]{lcb98}

1435: Lejeune T., Cuisinier F., Buesel R., 1998, A\&AS, 130, 65

1436:  \bibitem[\protect\citename{Li, Han \& Zhang} 2004]{lhz04}

1437: Lifang Li., Zhanwen Han, Fenghui Zhang, 2004, MNRAS, 355, 1383

1438:  \bibitem[\protect\citename{Li, Han \& Zhang} 2005]{lhz051}

1439: Lifang Li., Zhanwen Han, Fenghui Zhang, 2005, MNRAS, 360, 272

1440:  \bibitem[\protect\citename{Lombardi, Rasio \& Shapiro}1996]{lrs96}

1441: Lombardi, J. C. Jr, Rasio, F. A., Shapiro, S. L., 1996, ApJ, 468, 797

1442:  \bibitem[\protect\citename{Maceroni \& van't Veer}1996]{mv96}

1443: Maceroni C., van't Veer F., 1996, A\&A, 311, 523

1444:  \bibitem[\protect\citename{Mapelli et al.}2004]{map04}

1445: Mapelli M., Sigurdsson S., Colpi M. et al. 2004, ApJ, 605, L29

1446:  \bibitem[\protect\citename{Mateo et al.}1990]{mat90}

1447: Mateo M., Harris H., Nemec J., Olszewski E., 1990, AJ, 100, 469

1448:  \bibitem[\protect\citename{Mazeh et al.}1992]{maz92}

1449: Mazeh T., Goldberg D., Duquennoy A., Mayor M., 1992, ApJ, 401, 265

1450:  \bibitem[\protect\citename{Mestel }1984]{mes84}

1451: Mestel L., 1984, in Baliunas S. L., Hartmann L., eds., Cool Stars, Stellar Systems, and the Sun. Springer-Verlag, Berlin, p.49

1452:  \bibitem[\protect\citename{Micheal \& Kevin } 2006]{mk06}

1453: Micheal P., Kevin P. W., 2006, ApJL, 641, 137

1454:  \bibitem[\protect\citename{Miller \& Scalo }1979]{ms79}

1455: Miller G. E., Scalo J. M., 1979, ApJS, 41, 513

1456:  \bibitem[\protect\citename{Mochejska et al. }2004]{moc04}

1457: Mochejska B. J., Stanek K. Z., Sasselov D. D. et al.  2004, AJ, 128, 312

1458:  \bibitem[\protect\citename{Nelson \& Eggleton }2001]{nel01}

1459: Nelson C. A., Eggleton P. P., 2001, ApJ, 552, 664

1460:  \bibitem[\protect\citename{Pols et al.\ }1998]{pol98}

1461:   Pols O.R., Schr\"oder K.-P., Hurley J.R., Tout C.A., Eggleton P.P.,

1462:   1998, MNRAS, 298, 525

1463:  \bibitem[\protect\citename{Pols et al.\ }1995]{pol95}

1464:   Pols O.R., Tout C.A., Eggleton P.P., Han Z., 1995, MNRAS, 274, 964

1465:  \bibitem[\protect\citename{Pols \& Marinus }1994]{pol94}

1466: Pols O. R., Marnus M., 1994, A\&A, 288, 475

1467:  \bibitem[\protect\citename{Porter \& Townsend }2005]{pt05}

1468: Porter J. M., Townsend, R. H. D., 2005, ApJ, 623, L129

1469:  \bibitem[\protect\citename{Rucinski }1998]{ruc98}

1470: Rucinski S. M., 1998, AJ, 116, 2998

1471: \bibitem[\protect\citename{Sandquist, Bolte \& Hernquist }1997]{sbh97}

1472: Sandquist E. L., Bolte M., Hernquist L., 1997, ApJ, 477, 335

1473:  \bibitem[\protect\citename{Sandquist \& Shetrone }2003]{ss03}

1474: Sandquist E. L., Shetrone M. D., 2003, AJ, 125, 2187

1475:  \bibitem[\protect\citename{Sandrelli et al. }1999]{san99}

1476: Sandrelli S., Bragaglia A., Tosi M., Marconi G., 1999, MNRAS, 309, 739

1477:  \bibitem[\protect\citename{Sills \& Lombardi }1997]{sill97}

1478: Sills A., Lombaridi, J. C. Jr., 1997. ApJ, 105, 1081

1479:  \bibitem[\protect\citename{Sills et al. }2001]{sill01}

1480: Sills A. et al. 2001. ApJ, 105, 1081

1481:  \bibitem[\protect\citename{Stepien }1995]{ste95}

1482: Stepien K., 1995, MNRAS, 274, 1019

1483:  \bibitem[\protect\citename{Stepien }2006]{ste06}

1484: Stepien K., 2006, AcA, 56, 199

1485:  \bibitem[\protect\citename{Stryker }1993]{str93}

1486: Stryker L. L., 1993. PASP, 105, 1081

1487:  \bibitem[\protect\citename{Tian et al. }2006]{tian06}

1488: Tian B., Deng L., Han Z., Zhang X., 2006, A\&A, 455, 247

1489:  \bibitem[\protect\citename{Tutukov \& Yungelson }1987]{ty87}

1490: Tutukov A., Yungelson L., Comm. Astrophys. C., 1987, 12, 51

1491:  \bibitem[\protect\citename{Ulrich }1972]{ulr72}

1492:   Ulrich, R.,K., 1972,ApJ, 172, 165

1493:  \bibitem[\protect\citename{VandenBerg \& Stetson }2004]{vs04}

1494: VandenBerg  D. A., Stetson P. B., 2004, PASP, 116, 1012

1495:  \bibitem[\protect\citename{van't Veer }1979]{van79}

1496: van't Veer F., 1979, A\&A, 80, 287

1497:  \bibitem[\protect\citename{van't Veer }1994]{van94}

1498: van't Veer F., 1994, Mem.S.A.It. 65, 105

1499:  \bibitem[\protect\citename{van't Veer }1997]{van97}

1500: van't Veer F., 1997, In: Kam-Ching Leung, eds., ASP Conf. Ser. Vol.130, the Third Pacific Rim Conference on Recent Development on Binary Star Reseach, p.57

1501:  \bibitem[\protect\citename{Webbink }1976]{web76}

1502: Webbink R. F., 1976, ApJ, 209, 829

1503:  \bibitem[\protect\citename{Webbink }1985]{web85}

1504:  \bibitem[\protect\citename{Webbink }1976]{web77}

1505: Webbink R. F., 1977, ApJ, 215, 851

1506:  \bibitem[\protect\citename{Webbink }1985]{web85}

1507: Webbink R. F., 1985,Stellar evolution and binaries, In Interacting Binary Stars, ed. J.Pringle and R. Wade (Cambridge: Cambridge University Press), p.39

1508:  \bibitem[\protect\citename{Xin \& Deng }2005]{xin05}

1509: Xin Y., Deng L., 2005, ApJ, 619,824

1510: \end{thebibliography}

1511: 

1512: 

1513: \appendix

1514: \section{Evolutionary results in the paper}

1515: We choose a set of binaries undergoing AS and AR evolution

1516: from Nelson \& Eggleton \shortcite{nel01}

1517: to study the characteristics of the merger products and their

1518: connections with blue stragglers.

1519: Table A1 gives the initial parameters of the binary systems,

1520: their RLOF information and

1521: the structures and evolutionary consequences of the mergers.

1522: The first three columns contain the initial mass of the primary

1523: $M_{\rm 1i}$,

1524: the initial mass ratio $q_{\rm i}$ (the primary to the secondary)

1525: and the initial orbital period $P_{\rm i}$ in logarithmic, where

1526: $P_{\rm ZAMS}$ is given by (Nelson \& Eggleton, 2001)

1527: \begin{equation}

1528: P_{\rm ZAMS} \approx $$0.19M_{\rm 1i}+0.47M_{\rm 1i}^{2.33}\over 1+1.18M_{\rm 1i}^2$$.

1529: \end{equation}

1530: The fourth and fifth columns are the ages at which

1531: Roche lobe overflow begins ($t_{\rm RLOF}$)

1532: and the binary comes into contact ($t_{\rm contact}$) in our calculation.

1533: The next three columns show some system parameters at $t_{\rm contact}$,

1534: i.e. the mass of the primary $M_{\rm 1}$,

1535: the mass of the secondary $M_{\rm 2}$

1536: and the orbital period $P_{\rm contact}$.

1537: The remaining columns present the evolutionary results of the mergers:

1538: the lifetime on the main sequence ($t_{\rm MS}$),

1539: the central hydrogen mass fraction of the merger as constructed

1540: ($X_{\rm Hcc}$) and after adjustment ($X_{\rm Hcm}$),

1541: surface abundances for the elements H ($X_{\rm Hs}$), He ($X_{\rm Hes}$)

1542: and the ratio of C/N at the surface ($(C/N)_{\rm s}$).

1543: 

1544: 

1545: %\include{ATable1}

1546: 

1547: \end{document}

1548: