Radial Velocity Studies of Close Binary Stars.XIII
Slavek M. Rucinski, Theodor Pribulla, Stefan W. Mochnacki, Evgenij Liokumovich, Wenxian Lu, Heide DeBond, Archie de Ridder, Toomas Karmo, Matt Rock, J.R. Thomson, Waldemar Ogloza, Krysztof Kaminski, Piotr Ligeza
aa r X i v : . [ a s t r o - ph ] M a y Radial Velocity Studies of Close Binary Stars. XIII Slavek M. Rucinski, Theodor Pribulla , Stefan W. Mochnacki, Evgenij Liokumovich,Wenxian Lu , Heide DeBond, Archie de Ridder, Toomas Karmo, Matt Rock, J.R. Thomson David Dunlap Observatory, University of TorontoP.O. Box 360, Richmond Hill, Ontario, Canada L4C 4Y6 (rucinski,mochnacki,debond,ridder,karmo)@astro.utoronto.ca,[email protected]
Waldemar Og loza
Mt. Suhora Observatory of the Pedagogical Universityul. Podchora˙zych 2, 30–084 Cracow, Poland [email protected]
Krysztof Kaminski, Piotr Ligeza
Adam Mickiewicz University Observatory, S loneczna 36, 60–286 Pozna´n, Poland [email protected],[email protected]
ABSTRACT
Radial-velocity measurements and sine-curve fits to the orbital radial velocity vari-ations are presented for ten close binary systems: EG Cep, V1191 Cyg, V1003 Her,BD+7 ◦ ◦ q = 0 . Subject headings: stars: close binaries - stars: eclipsing binaries – stars: variable stars Astronomical Institute, Slovak Academy of Sciences, 059 60 Tatransk´a Lomnica, Slovakia Current address: Department of Geography and Meteorology and Department of Physics and Astronomy, Val-paraiso University, Valparaiso, IN 46383, U.S.A. e-mail: [email protected] Based on the data obtained at the David Dunlap Observatory, University of Toronto.
1. INTRODUCTION
This paper is a continuation of the series of papers (Papers I – XII) of radial-velocity studiesof close binary stars and presents data for the twelfth group of ten close binary stars observed atthe David Dunlap Observatory. For full references to the previous papers, see the last paper byPribulla et al. (2007, Paper XII); for technical details and conventions, for preliminary estimatesof uncertainties, and for a description of the broadening functions (BFs) technique, see the interimsummary paper by Rucinski (2002, Paper VII). The DDO studies use the efficient program ofPych (2004) for removal of cosmic rays from 2-D images.All data used in the present paper were obtained using the broadening functions extracted fromthe region of the Mg I triplet at 5184 ˚A, as in most of the previous papers. In August 2005, a new2160 lines/mm grating was acquired to replace the previously most frequently used 1800 lines/mmgrating which after many years of use lost its efficiency. The new grating markedly improved qualityof the observed spectra and of the resulting BFs. Of the reported results, the older grating was usedfor 1997 observations of V357 Peg, V1123 Tau and V1128 Tau (this star was also briefly observedin 2003); all three stars were later re-observed using the new grating. The radial velocity (hereafterRV) observations reported in this paper have been collected between October 1997 and August2007. The ranges of dates for individual systems can be found in Table 1. A few additional, lowquality spectra of PY Vir, not listed in Table 1, were taken on February 19, 2008.Selection of the targets in our program remains quasi-random: At a given time, we observe afew dozen close binary systems with periods usually shorter than one day, brighter than 10 – 11magnitude and with declinations > − ◦ ; we publish the results in groups of ten systems as soonas reasonable orbital elements are obtained from measurements evenly distributed in orbital phase.For this paper, we selected mostly targets from among fainter HIPPARCOS discoveries. None ofthe present targets has a spectroscopic orbit published. V1003 Her and PY Vir have only relativelypoor, ground-based photometric data while V1191 Cyg is a rather neglected but – as we describe– a very interesting contact binary.The radial velocities for the short period binaries reported in this paper were determined byfitting the double rotational profiles, as explained in Pribulla et al. (2006). Similarly as in ourprevious papers dealing with multiple systems (here the cases of BD+7 ◦ B − V ) color indices usually taken from the Tycho-2 catalog (Høg et al. 2000) and thephotometric estimates of the spectral types using the relations of Bessell (1979). In this paper wealso made use of infrared colors determined from 2 µ All Sky Survey (Skrutskie et al. 2006, 2MASS).Especially useful is the J − K color index, which is monotonically rising from the early spectraltypes to about M0V (Cox 2000). This infrared color is affected relatively less by the interstellar 3 –absorption than B − V . Parallaxes cited throughout the paper were adopted from the new reductionof the HIPPARCOS data (van Leeuwen 2007) which supersede the original reductions (ESA 1997).This paper is structured in a way similar to that of previous papers, in that most of the datafor the observed binaries are in two tables consisting of the RV measurements in Table 1 and of theirpreliminary sine-curve solutions in Table 2. Radial velocities and the corresponding spectroscopicorbits for all ten systems are shown in phase diagrams in Figures 1 – 3. The measured RV’s arelisted in Table 1. Only data with full weights were used to compute the orbits; we discardedpoorer observations. Table 2 contains also our new spectral classifications of the program objects.Section 2 of the paper contains summaries of previous studies for individual systems and commentson the new data. Examples of BFs of individual systems extracted from spectra observed close toquadratures are shown in Fig. 4.The data in Table 2 are organized in the same manner as in the previous papers of this series.In addition to the parameters of spectroscopic orbits, the table provides information about therelation between the spectroscopically observed upper conjunction of the more massive component, T (not necessarily the primary eclipse) and the recent photometric determinations of the primaryminimum in the form of the O − C deviations for the number of elapsed periods E . The referenceephemerides were taken from various sources: For V1003 Her, we doubled the HIPPARCOS periodand shifted the instant of the maximum by 0 . P ; for BD+7 ◦ survey (Pojmanski 2002); for V407 Peg, from Maciejewski et al. (2002) and for PY Vir, fromWils & Dworak (2003). For the rest of the systems, the ephemerides given in the on-line versionof “An Atlas O-C diagrams of eclipsing binary stars” (Kreiner 2004) were adopted. Because theon-line ephemerides are frequently updated, we give those used for the computation of the O − C residuals in the comments to Table 2 (the status as of December 2007). The deeper eclipse in W-type contact binary systems corresponds to the lower conjunction of the more massive component;in such cases the epoch in Table 2 is a half-integer number.
2. RESULTS FOR INDIVIDUAL SYSTEMS2.1. EG Cep
The variability of EG Cep was discovered by Strohmeier (1958). Photoelectric light curvesand their solutions have been presented by several investigators; for references see Erdem et al.(2005). The system shows a β Lyrae-type light curve with minima about 1.0 and 0.3 mag deep.Etzel & Olson (1993) determined the projected rotational velocity of the primary as 146 ± − . A photometric analysis of the system was performed by Kaluzny & Semeniuk (1984) who http://archive.princeton.edu/ ∼ asas/ q ph = 0 . − .
50 and a semi-detached configuration forthe system. Chochol et al. (1998) discussed the long-term period change in the system and alsoarrived at the semi-detached configuration with the less massive component filling its Roche lobeand the mass ratio q ph = 0.47. The new analysis of Erdem et al. (2005) led to a similar mass ratio.No spectroscopic orbit of the system has been published so far.In view of discrepancies between the spectroscopic and photometric mass ratios previouslynoted in our DDO series, it is surprising, but also encouraging, that all photometric investigationshave led to a mass ratio close to our spectroscopic value, q sp = 0.464(5). This is in spite of the semi-detached configuration which has weaker photometric constraints than the contact configuration inthe light curve solution. The agreement must result from the high orbital inclination angle, ≈ ◦ (Chochol et al. 1998) and the presence of the total eclipses.The intrinsic color of the system was estimated at about ( B − V ) = 0.197 (Kaluzny & Semeniuk1984) indicating the spectral type of the primary component of A7. The Tycho-2 color index B − V = 0 .
24 is consistent with our A7V classification and requires a slight reddening. The J − K = 0 . HIPPARCOS astrometricmeasurements.
The variability of the contact binary V1191 Cyg (GSC 3159-1512, V max = 10.82, V min = 11.15)was detected by Mayer (1965) while observing the nearby star V1187 Cyg. Since then the systemwas rather neglected with the only thorough photometric analysis of Pribulla et al. (2005), whofound that the orbital period of the system increases at a record rapid (for contact binaries) rateof ∆ P/P = 2.12 10 − year − . The system is totally eclipsing so the geometric elements derivedthrough a light curve solution, i = 80.4 ◦ , q = 0.094 and f = 0.46 were well defined in the solutionof Pribulla et al. (2005). The deeper minimum is flat, hence the system was classified as a W-type contact binary. The authors determined ( B − V ) = 0 .
62 (uncorrected for the interstellarabsorption). No spectroscopic study of the system has been published yet.We found V1191 Cyg to be a rather difficult spectroscopic target due to its relative faintness,the short orbital period of P = 0 . q sp = 0 . ± . J − K = 0 . B − V = 0 . ± .
10 is too uncertain to draw any firm conclusions.
The variability of this system was found by the
HIPPARCOS satellite. It is described thereas a periodic variable of an unspecified variability type with the period P = 0 . HIPPARCOS value. Analysis of photometric observations of the binary is complicated by the ratherlow amplitude, ∆ V = 0 .
09. In the ASAS-3 survey (Pojmanski 2002), the system appears with anundefined variability type and the best period of 21.846 days; however, the same observations,when phased with the double of the
HIPPARCOS period, show a light curve with the minima ofsimilar depths, a feature which is typical for contact binaries. Rather noisy observations can befound in the NSVS database (NSVS 11074663, http://skydot.lanl.gov/nsvs/nsvs.php). Except forthose fragmentary observations, no other photometric or spectroscopic observations of V1003 Herare available.Our spectroscopic observations show that V1003 Her is indeed a close binary, most likely ofthe W UMa type, but – because of the very small photometric amplitude – a full description andclassification of the system is impossible at this point and would require a high-precision photometry.The projected total mass of the system, ( M + M ) sin i = 0 . ± .
009 M ⊙ , the low photometricamplitude, the early spectral type (A7) and the absence of any third light, all indicate a very lowinclination angle. The HIPPARCOS parallax, π = 4 . ± .
66 mas is of limited use because ofits low relative accuracy. Our spectral type estimate, A7V, is consistent with the 2MASS infraredcolor, J − K = 0.234, but the Tycho-2 color index B − V = 0 .
38 indicates a substantial reddeningin the direction to V1003 Her, E ( B − V ) ≈ . ◦ Variability of BD+7 ◦ HJ D = 2 452 383 .
92 +0 . × E for the primary minima. No photometric or spectroscopic observations of the systemhave been published yet. Simultaneous photometry during our spectroscopic observations has ledto determination of one light minimum at HJ D = 2 454 188 . ◦ L / ( L + L ) = 0 .
50 6 –at the brightness maximum of the eclipsing pair; its signature is narrow with V sin i ≤
15 km s − ,which is close to the resolution of our spectroscopy. After approximating the third peak by aGaussian and its removal from the BFs, we obtained well-defined BFs of the eclipsing pair. Theradial velocity of the third component, RV = − . ± . − was found to be close to thecenter of mass velocity of the eclipsing pair, V = − . ± . − , hence the third componentappears to be a physical member of the system. The small difference of velocities is, very probably,a result of a slow mutual revolution. The system is not listed in the WDS catalogue as a previouslyrecognized visual double (Mason et al. 2001).The parallax of the system is unknown, but its absolute visual magnitude can be estimatedusing calibration of Rucinski & Duerbeck (1997) and the K2 spectral type (( B − V ) = 0 .
74) at M V = 4 .
84. A correction of the maximum visual magnitude of the combined system, V = 9 .
89, forthe contribution of the third component, results in V = 10 .
33 and then the distance d = 97 pc.By combining the proper motion of BD+7 ◦ − . The 2MASS color of BD+7 ◦ J − K = 0 . B − V = 0 . The variability of V357 Peg (HD222994) was discovered during the
HIPPARCOS mission. Thesystem was correctly classified as a contact binary. The first photometry of V357 Peg, publishedby Yasarsoy et al. (2000), shows a typical W UMa-type light curve. The photometric amplitude isabout 0.48 mag so the eclipses cannot be far from total (assuming our q sp = 0 . V max = 9 . q = 0 .
401 which is moderately large for this type. The total projected mass ofthe system ( M + M ) sin i = 2 . M ⊙ is probably not far from the true total mass. Our spectrataken in August and September 2005 showed a dark photospheric spot on the secondary componentwhich became visible just after the secondary minimum (Fig. 5). In addition to the main seriesof observations in 2005, V357 Peg had been shortly observed in 1997. These observations do notshow any indication of the photospheric spots. All radial velocities are listed in Table 1, but the1997 data were not used in orbital solution given in Table 2 to avoid any influence of the possiblyimprecisely-known or variable period. An orbit based on the 1997 data ( V = − . − , K = 93 . − , K = 234 . − , T = 2 , , . π = 6 . ± .
56 mas, is too imprecise to determinereliably its absolute magnitude. Our estimate of the spectral type, F2V, agrees with both, the 7 –2MASS color, J − K = 0 . B − V = 0 . V407 Peg (BD+14 ◦ V max = 9.28) was found to be a variable during the Semi-AutomaticVariability Search program at the Piwnice Observatory close to Toru´n, Poland (Maciejewski et al.2002). The authors presented a moderate-precision BV photometry and the first ephemeris for theprimary minima HJ D = 2 452 558 . . × E . The light curve of V407 Peg was that ofa contact binary.Later Maciejewski & Ligeza (2004) published 13 radial-velocity determinations based on spec-tra taken at the David Dunlap Observatory and processed using the BF formalism. Unfortunately,about a half of the available spectra were taken close to the orbital conjunctions making the derivedspectroscopic elements ( V = 22 . ± . − , K = 54 . ± . − and K = 233 . ± . − ) rather uncertain and in fact very different from our orbit which is based on 63 spectra.The most substantial difference is in the 8.7% larger sum of the semi-amplitudes resulting in the29% larger total (projected) mass.The star was not included in the HIPPARCOS mission. Its J − K color in the 2MASS catalogue(Skrutskie et al. 2006) is 0.181 corresponding to the F1 spectral type, which is consistent with the( B − V ) = 0 .
35 determined by Maciejewski et al. (2002). Our spectral classification is F0V.
The variability of V1123 Tau (visual double WDS 03350+1743) was discovered during the
HIPPARCOS satellite mission. In the
HIPPARCOS
Variable Stars Annex, the star is classi-fied as a β Lyrae-type eclipsing binary with the ephemeris for the primary minimum:
HJ D =2 458 500 . . × E . The binary is accompanied by a fainter companion ( ρ = 4 . ′′ , θ =136 ◦ and ∆ V = 1 . HIPPARCOS trigonometric parallax, π = 6 . ± . B − V ) = 0 . U filter the full amplitude of the light curve was 0.413mag, it was only 0.352 mag in the R filter.The light of the visual companion was partially entering the spectrograph slit during periods of 8 –the poor seeing; it was however marginally visible in the extracted BFs with the relative contributionnot larger than typically 0.02 – 0.03. Its radial velocity, V = 29 km s − was constant during ourobserving run and was fairly close to the center-of-mass velocity of eclipsing pair, V = 25 . ± . − .In addition to the recent observations reported here (September 2005 – March 2007), V1123 Tauwas also observed in 2002 using the older CCD chip (see previous papers of this series) and the1800 lines/mm grating. The resulting radial velocities, listed in Table 1, were not used in orbitalsolution given in Table 2 due to the long interval between the two datasets. The orbit based onthe 2002 data ( V = 24.6 km s − , K = 69.3 km s − , K = 258.2 km s − , T = 2 , , . V1128 Tau (visual double WDS 03495+1255) is another
HIPPARCOS discovery. Originallyit was classified as a β Lyrae-type eclipsing binary with the 0.3043732 days orbital period. Theeclipsing pair forms a relatively wide visual double with BD+12 ◦ θ = 196 ◦ was not entering our spectrograph slit which ispermanently oriented in the E-W direction. The high-precision photometry of Tas et al. (2003)showed that the system is a totally eclipsing contact binary, with the totality lasting about 16minutes. A subsequent light curve modeling led to q ph = 0 .
48 and the high orbital inclination angleof i = 85 degrees. The light curve asymmetry, with the maximum following the primary minimumbeing brighter, was interpreted in terms of a cool spot on the cooler component.Our spectroscopic mass ratio, q sp = 0 . V = − − , K = 127.1 km s − , K = 243.4 km s − , T = 2 , , . HIPPARCOS parallax, 9 – π = 1 . ± .
27 mas, rather uncertain (see Pribulla & Rucinski (2006)). There is a disparitybetween J − K = 0 . The photometric variability of HH UMa was discovered by the
HIPPARCOS mission whereit was classified as a periodic variable with P = 0 . HIPPARCOS period. The ground-based observations of Pribulla et al. (2003) supported thecontact binary nature of HH UMa; they also gave an improved ephemeris:
HJ D = 2 452 368 . . × E . Due to the partial eclipses, the mass ratio could not be reliably determined andwas estimated as q ph ≈ . − . q sp = 0 . M + M ) sin i = 0 . M ⊙ , together with low amplitude of thelight curve, 0.17 mag, support the low inclination angle, as found in the photometric analysis.The trigonometric parallax of the system, π = 4 . ± .
78 mas, is too imprecise to draw anyconclusions on its absolute magnitude. The 2MASS color of the system, J − K = 0 . B − V = 0 .
50 corresponds tothe F7V spectral type.
PY Vir (GSC 4961-667) was found on Stardial images to be a variable of the W UMa type(Wils & Dworak 2003). The system is also known as an X-ray source (1RXS J131032.4-040934).No photometric or spectroscopic analysis of PY Vir has been published yet. A minimum observedphotometrically in parallel with the spectroscopic observations (
HJ D = 2 554 201 . V = − . ± . − ; this velocity is ratherdifferent from the systemic velocity of the contact binary, V = − .
66 km s − . A few additional,low quality spectra of PY Vir taken in poor-weather conditions on February 19, 2008 (not listedin Table 1) give V = − . ± . − (at the mean HJD = 2 454 515.883) so that we have anindication of a change in V . The velocities of the close binary from these additional observationsare not precise enough to see if the motion of the third component is reflected in systemic radialvelocity of the eclipsing pair. The third component may be a binary itself; otherwise, the velocity 10 –changes may result from its low mass and a relatively fast orbital motion in a tight triple system.We consider the latter possibility as a relatively probable and more exciting one, but its verificationwould require observations over several seasons.PY Vir has not been known to be a multiple system; it is also not listed in the WDS catalogue.A lunar occultation of the system in July 1984 did not reveal presence of any visual companion(Evans et al. 1985), hence the separation of components is probably very small. Moreover, thelight contribution of the third component is fairly small, only about L / ( L + L ) = 0 .
08, at themaximum light of the eclipsing pair.The 2MASS color of the system, J − K = 0 . B − V ) color is rather uncertainto draw any conclusions, B − V = 0 . ± .
3. SUMMARY
With the new ten short-period binaries, this paper brings the number of the systems studied atthe David Dunlap Observatory to 120. Almost all systems of this group have been rather neglectedand little has been known about them.The highlights of this series are: (1) the triple systems BD+7 ◦ ◦ REFERENCES
Bessell, M.S. 1979, PASP, 91, 589Chochol, D., Pribulla, T., Rovithis-Livaniou, H., Rovithis, P., Kranidiotis, A. 1998, Contrib. Astron.Obs. Skalnat´e Pleso, 28, 51Cox, A. 2000, Allen’s Astrophysical Quantities, Springer VerlagDuerbeck, H.W. 1997, Inf. Bull. Variable Stars., No. 4513Erdem, A., Budding, E., Demircan, O., Degirmenci, ¨O., G¨ulmen, ¨O., Sezer, C. 2005, Astron.Nachrichten, 326, 332European Space Agency. 1997. The
HIPPARCOS and
Tycho
Catalogues (ESA SP-1200)(Noor-dwijk: ESA) (HIP)Etzel, P.B., Olson, E.C. 1993, AJ, 106, 1200Evans, D.S., Edwards, D.A., Frueh, M., McWilliam, A., Sandmann, W. H. 1985, AJ, 90, 2360Høg, E., Fabricius, C., Makarov, V.V., Urban, S., Corbin, T., Wycoff, G., Bastian, U.,Schwekendiek, P., & Wicenec, A. 2000, A&A, 355L, 27Kaluzny, J., Semeniuk, I. 1984, Acta Astronomica, 34, 433Kreiner, J.M. 2004, Acta Astron., 54, 207van Leeuwen, F. 2007,
HIPPARCOS , the New Reduction of the Raw Data, Springer, 2007Maciejewski, G., Karska, A., Niedzielski, A. 2002, Inf. Bull. Variable Stars, No. 5343Maciejewski, G., Ligeza, P. 2004, Inf. Bull. Variable Stars, No. 5504Mason, B.D., Wycoff, G.L., Hartkopf, W.I., Douglass, G.G., & Worley, C.E. 2001, AJ, 122, 3466(WDS)Mayer, P. 1965, Bull. Astron. Obs. Czechosl., 16, 255.¨Ozdarcan, O., Topcu, A.T., Evren, S., Tas, G. 2006, Inf. Bull. Variable Stars, No. 5688 12 –Pojmanski, G. 2002, Acta Astr., 52, 397Pribulla, T., Parimucha, ˇS., Chochol, D., Vaˇnko, M. 2003, Inf. Bull. Variable Stars, No. 5414Pribulla, T., Vaˇnko, M., Chochol, D., Parimucha, ˇS., Baludansk´y, D. 2005, Ap&SS, 296, 281Pribulla, T., Rucinski, S.M. 2006, AJ, 131, 2986Pribulla, T., Rucinski, S.M., Lu, Wenxian, Mochnacki, S.W., Conidis, G., Blake, R.M., DeBond,H., Thomson, J.R., Pych, W., Ogloza, W., Siwak, M. 2006, AJ, 132, 769 (Paper XI)Pribulla, T., Rucinski, S.M., Conidis, G., DeBond, H., Thomson, J.R., Gazeas, K., Ogloza, W.2007, AJ, 133, 1977 (Paper XII)Pych, W. 2004, PASP, 116, 148Rucinski, S.M. 2000, AJ, 120, 319Rucinski, S.M. 2002, AJ, 124, 1746 (Paper VII)Rucinski, S.M., & Duerbeck, H.W. 1997, PASP, 109, 1340Rucinski, S.M., Pych, W., Ogloza, W., DeBond, H. Thomson, J.R., Mochnacki, S.W.,Capobianco,C.C. Conidis, G., & Rogoziecki, P. 2005, AJ, 130, 767Skrutskie, M.F., Cutri, R.M., Stiening, R. et al. 2006, AJ, 131, 1163Strohmeier, W. 1958, Kl. Ver¨off. Bamberg, 21, 22Tas G., Evren, S., Cakirli, O., Ibanoglu, C 2003, A&A, 411, 161Yasarsoy, B., Sipahi, E., Keskin, V. 2000, Inf. Bull. Variable Stars, No. 4866Wils, P., Dworak, S.W. 2003, Inf. Bull. Variable Stars, No. 5425
This preprint was prepared with the AAS L A TEX macros v5.2.
13 –Captions to figures:Fig. 1.— Radial velocities of the systems EG Cep, V1191 Cyg, V1003 Her and BD+7 ◦ ◦ ◦ V and V , as listed in Table 1, respectively. The component eclipsed at the minimum correspondingto T (as given in Table 2) is the one which shows negative velocities for the phase interval 0 . − . ◦ −
500 to +500 km s − .Fig. 5.— The broadening functions (BFs) of V357 Peg determined from observations betweenAugust 25 and September 6, 2005. The original BFs were binned into 0.02 phase intervals andsmoothed by convolution with a Gaussian profile ( σ = 0 .
02 in phase). The dark feature driftingthrough the secondary component profile is very probably a large photospheric spot. Its signatureappears to be variable, especially around the phase 0.75 implying that the spot may have changedits position during the two weeks of our observations. 14 – -300-200-1000100200300 -300-200-1000100200300 V r ( k m / s ) EG Cep
V1191 Cyg
Phase V r ( k m / s ) V1003 Her
Phase
BD+7 3142
Fig. 1.— Radial velocities of the systems EG Cep, V1191 Cyg, V1003 Her and BD+7 ◦ ◦ ◦ V and V , as listed in Table 1, respectively. The component eclipsed at the minimum correspondingto T (as given in Table 2) is the one which shows negative velocities for the phase interval 0 . − . -300-200-1000100200300 -300-200-1000100200300 V r ( k m / s ) V r ( k m / s ) V357 Peg
V407 Peg
Phase
V1123 Tau
Phase
V1128 Tau
Fig. 2.— The same as for Figure 1, but for V357 Peg, V407 Peg, V1123 Tau, and V1128 Tau. Allfour systems are contact binaries. V1123 Tau and V1128 Tau are members of relatively wide visualbinaries. -300-200-1000100200300
Phase V r ( k m / s ) HH UMa
Phase
PY Vir
Fig. 3.— The same as for Figures 1 and 2, for the systems HH UMa, and PY Vir. Both are typicalcontact binaries. 16 – V r (km/s) V r (km/s) Fig. 4.— The broadening functions (BFs) for all ten systems of this group, selected for phases closeto 0.25 or 0.75. The phases are marked by numbers in individual panels. The third star features inthe BFs of the contact binaries BD+7 ◦ −
500 to +500 km s − . 17 –Fig. 5.— The broadening functions (BFs) of V357 Peg determined from observations betweenAugust 25 and September 6, 2005. The original BFs were binned into 0.02 phase intervals andsmoothed by convolution with a Gaussian profile ( σ = 0 .
02 in phase). The dark feature driftingthrough the secondary component profile is very probably a large photospheric spot. Its signatureappears to be variable, especially around the phase 0.75 implying that the spot may have changedits position during the two weeks of our observations. 18 –Table 1. DDO radial velocity observations (the full table is available only in electronic form)
HJD–2,400,000 V W V W Phase[km s − ] [km s − ]54251.6402 − .
75 1.00 190.61 1.00 0.291954251.6582 − .
70 1.00 183.19 0.00 0.325054251.6736 − .
90 1.00 162.99 0.00 0.353354251.6902 − .
37 1.00 167.77 0.00 0.383754251.7059 0.00 0.00 0.00 0.00 0.412654251.7222 0.00 0.00 0.00 0.00 0.442554251.8037 0.00 0.00 0.00 0.00 0.592154251.8199 48.14 1.00 − .
99 0.00 0.621854251.8363 53.42 1.00 − .
35 0.00 0.652054251.8527 72.41 1.00 − .
16 0.00 0.6820Note. — The table gives the radial velocities used in the paper. Thefirst 10 rows of the table for the first program star, EG Cep, are shown.Observations leading to entirely inseparable broadening function peaksare given a zero weight; these observations may be eventually used inmore extensive modeling of broadening functions. The zero weightwas also assigned to observations of marginally visible peaks of thesecondary component (mainly in the case of EG Cep). The RV’s des-ignated as V correspond to the more massive component; it was alwaysthe component eclipsed during the minimum at the epoch T (this doesnot always correspond to the deeper minimum and photometric phase0.0). The phases listed in the last column correspond to T and periodsgiven in Table 2. For V1191 Cyg, where we used phase-smoothed BFs,the heliocentric Julian dates are omitted.
19 –Table 2. Spectroscopic orbital elements
Name Type Other names V K ǫ T – 2,400,000 P (days) q Sp. type K ǫ ( O − C )(d) [E] ( M + M ) sin i EG Cep EB HD194089 − . ◦
790 238.72(1.30) 8.67 +0 . − . . − . ◦ − . ◦ − . − . − . ◦ . ◦ . . ◦
579 25 . . ◦ − . − . ◦ . − . − . ◦ − . q ≤
1. The standard errors of the circularsolutions in the table are expressed in units of last decimal places quoted; they are given in parentheses after each value. The center-of-massvelocities ( V ), the velocity amplitudes ( K i ) and the standard unit-weight errors of the solutions ( ǫ ) are all expressed in km s − . Thespectroscopically determined moments of primary or secondary minima are given by T ; the corresponding ( O − C ) deviations (in days)have been calculated from the available prediction on primary minimum, as given in the text, using the assumed periods and the numberof epochs given by [E]. The values of ( M + M ) sin i are in the solar mass units.Ephemerides ( HJD min – 2,400,000 + period in days) used for the computation of the ( O − C ) residuals:EG Cep: 52500.5214 + 0.54462228V1191 Cyg: 52500.2507 + 0.3133874V1003 Her: 48500.43 + 0.493322BD+7 ◦◦