The X-ray Lightcurve of WR 140 M. F. Corcoran1,2, A. M. T. Pollock2, K. Hamaguchi1,4, C. Russell5 1 CRESST/NASA-GSFC,Greenbelt,MD,USA 2 UniversitiesSpaceResearchAssociation,Columbia,MD,USA 1 1 3 EuropeanSpaceAgency,VillanuevadelaCan˜ada,Madrid,Spain 0 2 4 UniversityofMaryland,BaltimoreCounty,Baltimore,MD,USA n 5 UniversityofDelaware,Baltimore,MD,USA a J 7 ] R Publishedinproceedingsof S ”StellarWindsinInteraction”,editorsT.EversbergandJ.H.Knapen. . h Fullproceedingsvolumeisavailableonhttp://www.stsci.de/pdf/arrabida.pdf p - o r t s a [ Abstract: WR140isacanonicalmassive“collidingwind”binarysysteminwhichperiodically-varyingX-ray 1 emission is produced by the collision between the wind of the WC7 and O4-5 star components in the space v 2 between the two stars. We have obtained X-ray observations using the RXTE satellite observatory through 2 almostonecompleteorbitalcycleincludingtwoconsecutiveperiastronpassages. Wediscusstheresultsofthis 4 observingcampaign,andtheimplicationsoftheX-raydataforourunderstandingoftheorbitaldynamicsand 1 . thestellarmassloss. 1 0 1 1 1 Introduction: WR 140 as a “Canonical” System : v i X AlthoughmassivestarsarethemostimportantobjectsintheUniverseforgenerationofmetalsneeded r for the formation of rocky planets and biological entities, there remain many open questions about a theseobjects,especiallyconcerninghowtheyevolveanddie. Understandingthisevolutionaryprocess is largely a matter of understanding how these massive stars lose mass and angular momentum. The studyofmassivebinariesisonekeysinceinthesesystemsthefundamentalstellarparameterscanbe directlymeasured(atleastforthosefortuitouslysituated)andbecausewind-wind(orevenwind-star) collisions in these systems provide an in-situ measure of the radiatively driven process which is the majormeansofmassandangularmomentumlossformostofamassivestar’snaturallife. WR 140 is often termed a “canonical” colliding wind system by stellar astrophysicists. Presum- ably this is meant in the sense that WR 140 may be used to establish the rule of behaviour of such massive binaries. Physicists refer to canonical pairs of complementary variables and this is at least superficiallyappropriateforWR140. Acanonicalobjectalsosuggestsakindofsystemicuniqueness, andindeedWR140asasystemisnearlyunique(ifuniquenesscanbeapproachedindegrees): it’san Workshop“StellarWindsinInteraction”ConventodaArra´bida,2010May29-June2 Years 10 2002 2004 2006 2008 2010 2012 10 Conjunction (WC7 behind) Quadrature Conjunction (O behind) Quadrature PCU2 L1 net rate 8 8 Cycle 1 repeated Lx (1/D) Net PCU2 cts/s 64 Net PCU2 cts/s 64 2008122904 2 2009010600 2 0 2009012212 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 0 2.94 2.96 2.98 3.00 3.02 3.04 Phase Phase Figure 1: Left: The RXTE lightcurve of WR 140, December 2000 to June 2010. Right: Detailed view of the two minimum observed by RXTE. The data in grey represents is the RXTE data offset by one orbital period, adopted to be P =2897days. evolvedbinarysystemwithanunusuallylong(7.9year)period,anditpossessesnearlythehighestec- centricityofallknownstellarorbits(seeWilliamsetal. 1990,andR.Fahedetal.,theseproceedings). Itiswellbeyondthescopeofthecurrentdiscussiontotrytounderstandhowthisbinarycollapsedinto this peculiar state (stellar capture? explosive disintegration of a third component?). Nature, however, has kindly provided this laboratory in which the stellar separations and orbital velocities vary by an order of magnitude around the orbit. This allows the patient astronomer to measure the dependence of the state of the shock heated gas in the wind-wind collision on the local thermodynamic variables, and to look for the influence of subtle (and sometimes not so subtle) effects due to radiative trans- fer of photospheric photons winding their way through a complex interacting wind structure, or due to the sudden onset of fluid instabilities. All these things make WR 140 an object of intense inter- est for observational astronomers of all stripes, stellar evolutionists, hydrodynamicists, and plasma physicists. 2 X-ray Emission from Colliding Wind Binaries Like WR 140 TheX-raybandisextremelyusefulforthestudyofcollidingwindshocksingeneralandforthestudy of WR 140 in particular. In WR 140 the stellar winds have terminal speeds V ∼ 3000 km s−1. A ∞ collisionatthesespeedsproducesshock-heatedgasatatemperatureof 3 T ≈ µV2R−1 ≈ 1.51×105(V /100kms−1)2K ≈ 1.3×108K s 16 ∞ ∞ (where R is the gas constant and µ the mean molecular weight of the winds), suitably emitting in the X-ray band at energies of 1–10 kilo-electron volts (keV). Because a substantial portion of the stellar winds becomes involved in the shock, the amount of X-ray emission generated is substantial and observable with modern X-ray satellite observatories. WR 140 was first detected as an X-ray bright source on 1984 May 19 by the EXOSAT observatory (Pollock 1985), and has been observed many timessubsequentlybyX-rayobservatorieslikeROSAT,ASCA,XMM,CHANDRA,andRXTE. 3 RXTE Observations TheRossiX-rayTimingExplorer(RXTE,Bradtetal. 1993),launchedinDecember1995,isasatellite observatory that flexibly observes variable X-ray emitting cosmic objects on timescales of microsec- onds to years. The workhorse instrument, the Proportional Counter Array (PCA), is sensitive to 38 X-raysemittedinthe2–60keVband,whichincludesthe10–20millionKemissiongeneratedwithin WR 140’s wind-wind collision shock. The first observation of WR 140 by RXTE was obtained on 2000December9,just53daysbeforeperiastronpassage,andRXTEobservedthesystemaboutonce per week up to about 778 days past periastron passage. Observations resumed on 2005 March 8, just past apastron of the stellar orbits, and continue up through the time of this writing, at a variable cadenceofafewobservationspermonthtooneobservationperdayduringthemostrecentperiastron passagein2009January. Figure 2: Comparison of the orbital variation of the 2cm radio flux (from White & Becker 1995) and the 2–10 keV X-rayflux. Figure 1 shows the WR 140 RXTE lightcurve as a function of phase φ (where φ = 0 is periastron passage)andtimeinyears. ThegeneralcharacteristicisagradualincreaseinX-raybrightnessasthe starsapproachperiastron,followedbyarapiddeclinetoabriefminimumstate,aquickrecoveryfol- lowed by a gradual decline in brightness. It’s worth noting that there is no rapid (∼weeks) variations (“flaring”) of the X-ray brightness as seen in WR 140’s “sister star” η Car just prior to the start of its X-ray minimum. As shown in Fig. 1, the X-ray minima from the two orbits observed byRXTE agree very well with each other in terms of brightness variability, depth and duration. There may, however, be slight variations in the level of X-ray brightness between the two cycles. Figure 1 also shows a curvewhichrepresentsa1/D variation,whereD istheseparationbetweenthetwostars. NearX-ray minimumtheX-raybrightnesscurvedoesnotfollowa1/Dvariationasitwouldifthecollidingwind shock were adiabatic and there were no line-of-sight absorption variations. We also note that X-ray minimum apparently precedes O-star conjunction (i.e. the time in the orbit when the O star is behind the WR star) by about 6 days. Figure 2 compares the orbital variation of the 2cm radio flux with the 2–10keVX-rayflux. Ascanbeseenfromthisfigure,boththeradiofluxandX-rayfluxshowminima nearperiastronpassage/Ostarconjunctionwhenthecollidingwindshockisviewedthroughthethick wind of the WC7 star. The radio decline is much more gradual than the X-ray minimum and in fact just before periastron passage the radio emission is in its minimum state while the X-ray emission is still increasing. The X-ray recovery from minimum is much more rapid than the radio recovery. The radio emissionreaches a maximumintensity nearapastron when thecolliding wind shockmoves out frombehindtheradiophotosphereofthesystem. Changes in the X-ray spectrum occur in concert with the observed 2–10 keV flux variations. ThesespectralvariationshavebeenseenbyEXOSAT,ROSAT,andASCA,buttheRXTEobservations provideamorecompleteviewofhowthespectrumchangesaroundtheorbit. BecausetheRXTEPCA has rather poor energy resolution, we choose to characterise these spectral variations as changes in X-ray hardness rather than attempting to fully model the spectra. Figure 3 shows the hardness ratio variations as seen by RXTE, along with a detailed view of the changes near periastron passage. In 39 1.0 1.0 Conjunction (WC7 behind) Quadrature Conjunction (O behind) Quadrature Hardness Ratio (7.5-3 keV)/(7.5+3 keV) --0010....0505 More Absorbed Less Absorbed cpurerrveionut sc ycyclcele Hardness Ratio (7.5-3 keV)/(7.5+3 keV) --0010....0505 Less AbsorbedMore Absorbed cpurerrveionut sc ycyclcele 2.0 2.2 2.4 2.6 2.8 3.0 3.2 2.94 2.96 2.98 3.00 3.02 3.04 Phase Phase Figure3: WR140hardnessratio(HR)variationasmeasuredbytheRXTEPCA,whereHRisthefluxat7.5keVminus thefluxat3keVdividedbytheirsum. Left:CompleteHRcurve. Right:Detailedviewnearperiastronpassage. Thethick verticlelinejustafterφ=3.0marksthephasingoftheminimumX-raybrightness. this figure the data in grey are data from the previous cycle advanced by 1 period. In contrast to the X-ray flux variation, the X-ray hardness reaches an extremum at O-star conjunction, as expected if thehardnessmaximumsignalsthetimeofmaximumX-rayabsorption. 4 Modelling ThefirstcalculationsofthelevelofX-raystobeexpectedfromthewind-windshockincollidingwind binarieswerebyCherepashchuk(1976)andPrilutskii&Usov(1976). Recentmodellingeffortshave included numerical 3-D hydrodynamic simulations of the wind-wind interaction (Parkin & Pittard 2008,Okazakietal. 2008). Figure4showsadensityslicethroughtheorbitalplanefora3-Dsmooth particle hydrodynamics (SPH) simulation using the Okazaki SPH code (see, for example, Okazaki et al. 2008) for WR 140. Four snapshots are shown: near X-ray maximum; near periastron; near X-ray minimum; and near O-star conjunction (maximum X-ray hardness). The axes are marked in units of the orbital semi-major axis a. The black arrow in the first image shows the observer’s line of sight projectedontotheorbitalplane. Figure 4: Adensityslicethroughtheorbitalplanefora3-DSPHsimulationforWR140. Foursnapshotsareshown, fromlefttoright: nearX-raymaximum; nearperiastron; nearX-rayminimum; andnearO-starconjunction(maximum X-rayhardness). Theblackarrowinthefirstimageshowstheobserver’slineofsightprojectedontotheorbitalplane. X-raymaximumoccurswhentheprojectedlineofsightpassesnearthemiddleofthelowdensity region of the O star’s wind, channeled by the boundary of the colliding wind shock. As the stars re- volveintheirorbitthebowshockaroundtheOstartwistsbehindthecompanionandgetsincreasingly obscured by the wind of the WR star through X-ray minimum. At conjunction the X-ray hardness reaches a maximum; this suggests that only the highest-energy photons can penetrate through the thickwindoftheWC7staratthisphase. 40 Figure5: AnX-raylightcurvefromtheSPHmodelcomparedtotheRXTEfluxes. Figure5showsanattempttomodeltheX-raylightcurveusingtheSPHmodel. Themodelshown is somewhat artificial in that it assumes a mono-energetic point source of X-rays located at the stag- nationpointoftheflow. TheX-raysproducedbythispointsourceareabsorbedbythedistortedwind structureoftheWC7starwhichissimulatedbytheSPHmodel. AsseeninFig.5thisreproducesquite well the variation seen in the RXTE lightcurve. However, more realistic models under development, inwhichtheX-raysareproducedbyadistributionofhotgasextendedalongthewind-windcollision shock, and using a more realistic temperature distribution, do not reproduce the X-ray behaviour as wellasthepointsourcemodel. This“distributed-emission”modelisstillunderinvestigation. 5 Conclusion TheRXTEmonitoringofWR140confirmsthattheX-rayemissionvariesinapredictablewaywhich is mostly consistent over the two orbital cycles studied. Far from periastron the X-ray emission mostly follows a 1/D law expected from an adiabatic shock, though strong deviations are seen near periastron,notallofwhichcanbeexplainedbyabsorptioninthewindoftheWC7star. Matchingthe X-ray minimum from 2001 to the 2009 minimum suggests that the X-ray period is P = 2897 days, about 2 days shorter than the period proposed by Marchenko et al. (2003). The X-ray data confirm thatWR140isagoodlaboratoryforthestudyofthephenomenaassociatedwithstrongshocksinthe astrophysicalenvironment. Acknowledgements This research was supported through NASA cooperative agreement NNG06EO960A, and made use of the Astrophysics Data System and the HEASARC archive. We express our appreciation to the workshop organisers, and to the amateur astronomers for their dedication and herculean efforts to observeWR140duringits2009periastronpassage. References Bradt,H.V.,Rothschild,R.E.,Swank,J.H.,1993,A&AS,97,355 Cherepashchuk,A.M.,1976,SovietAstronomyLetters,2,138 Marchenko,S.V.,Moffat,A.F.J.,Ballereau,D.,etal.,2003,ApJ,596,1295-1304 Okazaki,A.T.,Owocki,S.P.,Russell,C.M.P.,Corcoran,M.F.,2008,MNRAS,388,L39-L43 41 Parkin,E.R.,Pittard,J.M.,2008,MNRAS,388,1047-1061 Pollock,A.M.T.,1985,SpaceScienceReviews,40,63 Prilutskii,O.F.,Usov,V.V.,1976,SovietAstronomy,20,2 Williams,P.M.,vanderHucht,K.A.Pollock,A.M.T.,Florkowski,D.R.,vanderWoerd,H.,Wamsteker,W.M., 1990,MNRAS,243,662 White,R.L.,Becker,R.H.,1995,ApJ,451,352 42