and IC 4651 (Mermilliod et al. 1995, paper IV). Our approach of the red giants in the colour-magnitude diagram, photometric systems.

Size: px

Start display at page:

Download "and IC 4651 (Mermilliod et al. 1995, paper IV). Our approach of the red giants in the colour-magnitude diagram, photometric systems."

Russell Rogers
6 years ago
Views:

1 Your thesaurus codes are: , individual: NGC 2099, , , AND ASTROPHYSICS Red giants in open clusters. V. NGC 2099?,?? J.-C. Mermilliod 1, G. Huestamendia 2, G. del Rio 2, and M. Mayor 3 1 Institut d'astronomie de l'universite de Lausanne, CH-1290 Chavannes-des-Bois, Switzerland 2 Observatorio Astronomico de Madrid (IGN), Alfonso XII, 3, Madrid, Spain 3 Observatoire de Geneve, CH-1290 Sauverny, Switzerland Received: July 5; accepted: July 5 Abstract. 474 radial-velocity observations and new U BV photoelectric data for 55 and 20 red giants respectively, in the eld of the intermediate-age open cluster NGC 2099 are presented and analysed for membership and duplicity. The membership of 35 red giants has been conrmed, 16 spectroscopic binaries have been discovered and 11 orbits have been determined. The cluster binaryfrequency lower limit is slightly above the average (9/35 = 26%). The radial distribution of the red giants shows a sharp limit between members and non-members. No mass segregation is observed between the binary and single red giants. The latest isochrones from the Geneva and Padova groups reproduce very well the observed morphology in the colour-magnitude diagram for log t = 8.65 and z = 0.02, with (m? M) = and E(B? V ) = Key words: cluster: open { individual: NGC 2099 { radial velocity { star: red giant { binary: spectroscopic { photometry: UBV 1. Introduction This fth paper in a series devoted to the study of red giants in open clusters by means of accurate radial-velocity data presents CORAVEL observations and new U BV photoeletric data in the intermediate-age and rich open cluster NGC Earlier papers discussed 93 red giants in the ve intermediate-age clusters NGC 2360, 2423, 5822, 6811 and IC 4756 (Mermilliod & Mayor 1990, paper III) and Send oprint requests to: J.-C. Mermilliod? Based on observations made at the Haute-Provence Observatory (France) and the Observatorio Astronomico de Madrid (IGN) at Calar Alto (Spain)?? The individual radial velocities are available in electronic form: see the Editorial in A&A 1994, Vol. 280, no. 3, p. E1 dealt with 35 stars in the two older clusters NGC 3680 and IC 4651 (Mermilliod et al. 1995, paper IV). Our approach is to identify non-member stars and discover spectroscopic binaries through accurate, long-term CORAVEL radial-velocity observations. In order to study the distribution of the red giants in the colour-magnitude diagram, we also present new photometric data in the Geneva or U BV photometric systems. NGC 2099 = C ( = 5 h 49:1, = , B1950) is an interesting, populous intermediate-age open cluster which has not been much observed since the 1960s. The colour-magnitude diagram based on the data from major photographic U BV photometric studies (Hoag et al. 1961; Brosterhus 1963; West 1967) shows a rather broad main sequence and the published proper motions (Meurers 1960; Jeerys 1962; Upgren 1966) do not provide a clear separation of the cluster members from the eld stars. Robin (1982) obtained U BV electronographic data in an outer eld of the cluster. The cluster parameters from Lynga 's (1987) fth catalogue are m-m = 11.63, E(B-V) = 0.31 and log t = NGC 2099 presents a well populated clump of red giants which makes it an interesting object to determine the red giant membership, search for spectroscopic binaries and try to determine the structure of the clump. Glushkova & Rastorguev (1991) published 14 radial-velocity observations of 9 red giants in NGC 2099 and their results are in good agreement with those presented here. 2. Observations 2.1. The sample denition Red stars brighter than V = 12. and redder than B? V = 0.50 were selected for radial-velocity observation, mostly from the U BV photographic data. The colour-magnitude diagram based on these data shows a prominent clump which contains most programme stars. The rather blue

2 ple a few stars that are supposed to be composite binaries because they fall more or less in the middle of the Hertzsprung gap. The survey made by van Zeipel & Lindgren (1921) extends a little further out from the cluster centre than the photographic surveys and candidate red giants were selected from the visual magnitudes (m vis ) and colours (C) from their paper. The sample contains a total of 55 stars. The numerous identications attributed to the red giants by eight dierent authors are summarized in Table 1. The numbering system of van Zeipel & Lindgren (1921) has been adopted in this study and is presented in column 1 (vz). The successive columns 2 to 8 contain the star numbers attributed by \N": Nordlund (1909), "G": Giebeler (1914), "J": Joy (1915), "M": Meurers (1960), "Bros": Brosterhus (1963), "Bron": Bronnikova (1958) and "West": West (1967) Radial-velocity measurements As part of the general programme relating to evolved stars in open clusters, at least 5 radial velocity measurements were obtained for each member star with the radial velocity scanner CORAVEL (Baranne et al. 1979) installed on the 1m Swiss telescope at the Haute Provence Observatory (OHP, France) from 1978 to As a rule integration times range from 10 to 20 minutes depending on the night quality and seeing. The radial velocities were corrected for zero-point dierences to place them in the system dened by the southern CORAVEL. The 474 individual observations are available in electronic form from the Strasbourg anonymous ftp account (see the Editorial in A&A 280, no 3, 1994). The mean radial velocities are displayed in Table 3 and columns 6 to 11 contain successively the mean radial velocities and errors in [km s?1 ], the number of measurements (N), the time span (T) of the observations [days], the probability P( 2 ) that the mean error merely reects random observational errors for a constant star, and remarks on binarity and membership Photometric data During an observing run in November 1988 we obtained U BV photoelectric observations for 20 red giants with a single channel pulse counting mode photoelectric photometer attached to the 1.52m reector of the Observatorio Astronomico de Madrid (IGN) at Calar Alto (Spain). The dry-ice cooled photomultiplier and lter combinations used were described by Huestamendia et al. (1991). Standard stars were taken from the list of Neckel & Chini (1980). Counting errors were kept under the 1% level. The reduction procedure took into account possible dependence of the residuals on observing time, azimuth, hour angle and colour, and applied, where necessary, the corresponding corrections. Standard-star rms were below 0.02 magnitude. The new photometric data, mostly based Table 1. Cross-identications vz N G J M Bros Bron West A F C N J Q K

3 4 of table 3 (source #1). The other sources of U BV photoelectric photometry are: (2) Coleman (1982), (3) West (1967), (4) Jennens & Helfer (1975), (5) Brosterhus (1963). Sources (6) and (7) refer to the photographic data from West (1967) and Brosterhus (1963) respectively, corrected for the mean dierences given in Table 2 computed in the sense our photometry minus the photographic data. For the stars without U BV data, only the magnitudes (m vis ) from van Zeipel & Lindgren (1921) have been tabulated (source 8). This concerns mostly non-members and it does not aect the aspect of the colour-magnitude diagram. Table 2. Mean dierences UBV pe { pg V (B-V) (U-B) N* Source Mean di Brosterhus Mean di West elements are given in Table 4. A few stars deserve special comments. vz 92 has a close companion and it is dicult to know where the automatic guiding sets the entrance slit. The scatter in the data may be simply due to the eect of the companion. vz 255 has a wide dip and the individual errors are large. Twenty-two observations were done but no period could be found. The last, better integrated, observations do not show the same large scatter. vz 1006 and 1866, both non-members, have three velocities in very close agreement and one observation about 2 km s?1 apart. Their binary nature seems very unlikely, although P( 2 ) = and is noted \SB??" in the remarks. vz 412 presents a very small mass function (0.08) and four of the ve binary members with an orbit have a value of M 2min / M 1 less than 0.5. The problem of the distribution of the secondary mass as a function of the primary mass will be discussed in a forthcoming paper collecting the published orbits and additional material to be published. 3. Results 3.1. Spectroscopic binaries The membership of a spectroscopic binary is dicult to determine until an orbit has been obtained or the observations are numerous enough to decide whether the mean velocity and the range of variation are compatible with membership. The presence of 20 to 25% of spectroscopic binaries among the red giants in open clusters requires numerous observations to determine the orbits. In NGC 2099, fourteen spectroscopic binaries, of which only 9 are really cluster members were discovered. Eleven orbits, some of them being still preliminary, have been determined. Less observations were done on the non-member binaries when their non-membership was recognized. The orbital elements are given in Table 4. Some errors may seem rather large. They result partly from the length of some periods and partly from the somewhat small amplitudes, for vz 412, 685, 966, for example. The radialvelocity curves for the ve member binaries (vz 49, 149, 412, 485, 966) are displayed in Figs. 1, 2, 3, 4, and 5. Two additional stars (vz 12 and 748) show clear radial-velocity variations, but the periods will be very long, i.e. much greater than 5000 d, and quite eccentric. The radial velocities are plotted in function of the Julian dates of observations in Fig. 6 and 7 respectively. The resulting binary frequency is 26% (= 9/35). Seven spectroscopic binaries were discovered among the non-member stars, and six orbits determined. The Fig. 1. Radial-velocity curve for vz Mean radial velocity and membership The distribution of the mean radial velocities for the 55 stars observed (Fig. 8) shows a pronounced maximum around 7.5 km s?1 and a clear separation of the members and non-members. Thirty-ve stars have been considered as cluster members and the others are marked with NM in Table 3. The mean velocity for NGC 2099 based on 30 non-variable stars and spectrocopic binaries with an orbit is km s?1. The r.m.s dispersion around the mean is 0.92 km s?1. This high value is nearly twice that given for M67 (0.48 km s?1 ) by Mathieu (1985). It may

4 Table 3. New UBV photometry and radial velocity observations vz V B-V U-B S V r N T P( 2 ) Rem SB SB, orbit SB SB, orbit SB, orbit SB SB, orbit NM SB, orbit, NM NM SB, orbit, NM SB SB, orbit, NM SB, orbit, NM SB, orbit, NM NM SB, orbit SB??, NM NM NM NM NM SB, orbit, NM NM NM NM SB??, NM NM NM

5 Table 4. Orbital elements of eleven spectroscopic binaries in NGC 2099 vz P T e! K f(m) a sini M2min (O? C) n obs [d] [HJD] [km s?1 ] [ ] [km s?1 ] [M ] [Gm] [M ] (xed) Fig. 2. Radial-velocity curve for vz 149. Fig. 3. Radial-velocity curve for vz 412. be due to yet undiscovered binaries with long periods and low amplitudes, but it certainly results for a large fraction from the true internal velocity dispersion since NGC 2099 is signicantly more massive than most clusters we have studied in previous papers. The mass for NGC 2099 published by Bruch & Sanders (1983) (531 M ) is grossly underestimated. The upper main-sequence stars have a mass around M and their estimated mass would imply that there are only about 200 stars in NGC 2099, which is probably too small by a factor of ten.

6 Fig. 4. Radial-velocity curve for vz 485. Fig. 6. Radial-velocity curve for vz 12. Fig. 5. Radial-velocity curve for vz 966. Fig. 7. Radial-velocity curve for vz Radial distribution of the red giants Artjukhina & Kholopov (1964) studied the structure of the open cluster NGC 2099 by the star count method. They obtained a nucleus radius of 10 0, an inner corona radius of 30 0 and an outer corona extending out to The area covered by the survey of van Zeipel & Lindgren (1921) is 44 0 x44 0. It therefore does not extend far enough to completely include the inner corona of the cluster. The area distribution of the red giant members (lled circles) and non-member stars (open circles) is displayed in Fig. 9. The scale is in minutes of arc. Thirty-one red giant members are included in a distance of 10 0 from the centre, and 34 are closer than Most red giants are therefore contained in the cluster nucleus and the present sample probably contains a very large fraction of the red giant population of the cluster. NGC 2099 was included in a search for mass segregation in clusters of various ages made by Raboud & Mermilliod (1994). It belongs to the younger group, and shows no dierence in the radial distribution of binaries and single member stars. 4. The colour-magnitude diagram The second aim of this work on red giants in open clusters is to study the morphology of the red giant locus in the colour-magnitude diagram to test the predictions of stellar evolutionary models. The colour-magnitude diagram plotted in Fig. 10 represents a large improvement

7 and turno region. From the (U? B; B? V ) diagram, a colour excess E(B? V ) = 0.29 has been determined and ZAMS tting gave a distance modulus m? M = The V magnitudes and (B? V ) colours of the red giants have been corrected by and mag. respectively to take into account the dierential reddening of red stars with respect to main-sequence stars. vz 391 (V = 10.88, B-V = 1.51) falls slightly to the red of the other stars and would also lie to the right of the isochrone when adjusted to t the clump position. Its radial velocity is very close to the cluster mean velocity and the U BV photoelectric data show also a very good agreement between the three sources. Therefore the observational data and evidence for membership are not questionable. Fig. 8. Radial-velocity distribution of the 55 stars observed. An isochrone from Schaller et al. (1992) with log t = 8.65 has been plotted over the observed points (Fig. 10). This age reproduces quite well the upper main-sequence morphology. The second sequence is the same isochrone moved upward by 0.75 mag. to simulate the binary limit. The two curves dene the main-sequence envelope and give a very good t to the observed narrowing. The predicted position of the red giants for the core-helium burning phase is slightly too red as compared with the observed points. However the observed red giant distribution is compatible with the expected distribution of the stars during the helium-burning phase, on the left part of the loop, although no structure can be discerned. Two isochrones from Bertelli et al. (1994) for z = 0.02 and log t = 8.60 and 8.70 have been plotted over the same colour-magnitude diagram (Fig. 11). The isochrones also reproduce quite well the upper main-sequence morphology. The isochrone for log t = 8.60 nicely predicts the magnitude of the clump, but is also slightly too red. Fig. 9. Projected distribution of the members and non-members. North is up and East is left. The scale unit is [arcmin]. with respect to the diagrams based on photographic data, but still shows some scatter. The main reason is probably related to the diculty of properly measuring the stars in such a crowded eld with a high sky background due to the fainter stars. The various sources of U BV photoelectric data present a generally good agreement, but dierences up to 0.05 mag. nevertheless exist. The existing photographic data for NGC 2099 are not precise enough to really constrain the ZAMS and isochrone tting. Furthermore the red giant B-V colours show signicant dierences that let them fall quite outside from the isochrone. Therefore the isochrones have been tted to the available photoelectric data, which, in spite of The isochrones from both groups are very similar on the main sequence, but closer examination of the red giant region shows some dierences in the detailed morphology of the loop representing the helium-burning phase. For example, the relative position of the bluer part of the loop with respect to the bottom of the red giant branch or the relative position of the faintest point after the red giant tip with respect to the termination of the main sequence are dierent. Models from the Padova group (Fig. 11) seem to reproduce better the observed characteristics. Unfortunately, there are few stars in the subgiant portion of the diagram to help deciding which grid of models oers the more appropriate description. Either the red giant tracks need a little correction to reproduce the observed red giant position in the colour magnitude diagram, or the metallicity of the cluster is slightly less than solar and the red giant colours are bluer. However NGC 2099 is not decient enough to justify the use of isochrones with z =

8 Fig. 10. Colour-magnitude diagram with the stars in NGC 2099 measured photoelectrically. Filled squares stand for cluster members, and open square for non-members. The isochrone is from Schaller et al. (1992), for z = 0.02 and log t = The second, incomplete, curve is the same isochrone displaced by 0.75 mag. to mark the binary limit. The adopted cluster parameters are: m-m = and E(B-V) = Note the very good agreement of the isochrone with the main-sequence morphology. The observations of CORAVEL radial velocities and new U BV photoelectric photometry of red giants in the eld of NGC 2099 have permitted to perform a clear selection of 35 certain members and to discover 16 spectroscopic binaries, among which 9 only are cluster members. Eleven orbits have already been determined. The comparisons of the observed morphology of the red giants in the colour-magnitude diagram with the theoretical isochrones for z=0.02 show a good agreement with isochrone for log t = 8.65 from Schaller et al. (1992) and for log t = 8.60 and 8.70 from Bertelli et al. (1994). The cluster parameters derived from the t (m-m = and E(B-V) = 0.29) are close to the values given by Lynga in the fth edition of his catalogue (Lynga 1987). This cluster is very interesting for comparison with evolutionary models as well as for studying the structure of populous open clusters and the eect of mass segregation. It would therefore be important to get well calibrated CCD photometry to improve the main sequence denition. That resulting from the photographic photometry is much too broad to provide constraints strong enough to test the models. Acknowledgements. We are grateful to the referee, Dr G. Bertelli, for his valuable remarks on the evaluation of the colour-magnitude diagram. The work of JCM and MM has been supported by continuous grants from the Swiss National Foundation for Scientic Research (FNRS). GH and GR have been partially supported by Spanish Grant DGICYT PB Fig. 11. Same colour-magnitude diagram as in Fig. 10. Two isochrones from Bertelli et al. (1994), for z = 0.02, log t = 8.60 and The adopted cluster parameters are: m-m = and E(B-V) = Symbols as in Fig. 10. References Artjukhina N.M., Kholopov P.N. 1964, Soviet AJ 7, 840 Baranne A., Mayor M., Poncet J.-L. 1979, Vistas in Astron. 23, 279 Bertelli G., Bressan A., Chiosi C., Fagotto F., Nasi E. 1994, A&AS 106, 275 Bronnikova N.M. 1958, Tr. Glavn. Astr. Obs. Pulkove 72, 77 Brosterhus E. 1963, Astron. Ahb. Hamb. Sternw. VII no 2 Bruch A., Sanders W.L. 1983, A&A 121, 237 Coleman L.A. 1982, AJ 87, 369 Giebeler H. 1914, Vero. Univ. Sternw. Bonn no 12 Glushkova E.V., Rastorguev A.S. 1991, Sov. Astr. Lett. 17, 13 Hoag A.A., Johnson H.L., Iriarte B., Mitchell R.I., Hallam K.L., Sharpless S. 1961, Publ. US. Nav. Obs. XVII, 347, part VII Huebner W.F., Merts A.L., Magee N.H., Argo M.F. 1977, Los Alamos Scientic Laboratory Reports LA-6760-M Huestamendia G., del Rio G., Mermilliod J.-C. 1991, A&AS 87, 153 Iglesias C.A., Rogers F.J., Wilson B.G. 1992, ApJ 397, 717 Jeerys III W.H. 1962, AJ 67, 532 Jennens P.A., Helfer H.L. 1975, MNRAS 172, 681 Joy A.H. 1915, AJ.29, 101 Lynga G. 1987, Fifth catalogue of cluster parameters (Strasbourg)

9 Symp. no 113, Eds. J. Goodman & P.Hut, (Reidel), p. 427 Mermilliod J.-C., Andersen J., Nordstrom B., Mayor M. 1995, A&A (in press) (paper IV) Mermilliod J.-C., Mayor M. 1990, A&A 237, 61 (paper III) Meurers J. 1960, Vero. Univ. Sternw. Bonn no 53 Neckel Th., Chini R. 1980, A&AS 39, 411 Nordlund J.O. 1909, Arkiv. Mat. Fis. 5 no 17 Raboud D., Mermilliod J.-C. 1994, A&A 289, 121 Robin A. 1982, A&AS 50, 251 Schaller G., Schaerer D., Meynet G., Maeder A. 1992, A&A 96, 269 Upgren A.R. 1966, AJ 71, 736 van Zeipel H., Lindgren J. 1921, Kungl. Sven. Vet. Handl. 61 no 15 West F.R. 1967, ApJS 14, 359 This article was processed by the author using Springer-Verlag LATEX A&A style le L-AA version 3.

Red giants in open clusters

Astron. Astrophys. 324, 91 96 (1997) ASTRONOMY AND ASTROPHYSICS Red giants in open clusters VII. Melotte 71, J.-C. Mermilliod 1, J.J. Clariá 2, J. Andersen 3, and M. Mayor 4 1 Institut d Astronomie de