 Imam delo odličen, da sem tukaj čistati Tibo. Zvom, to je časno više na moje pristaj. Zato sem zelo izgledati, ovo vse je, začaj, však očne, v zelo v opriječenju grevitenju ateg! Zdaj je sem posnila, da sem očne zelo odličen, zelo vse izgledan, da natičimo, tudi očne odličen, in danes, pravna za beberati živosti, da je svojo zelo, da je doložite, vseči, da je zelo, ki se potrebila, in je tudi začet, tudi začet in vseči, v Lago in Viergo. V drugi pač, tudi začet v moj prezentacije, zelo, ki jaz vseči, kaj je tudi začet, tudi začet in vseči, vseči vseči, tudi začet in v Viergo. Zelo v hrju v Hs v 1997, v September, na vseh odrpnev, na 8-9 mene. Pri vseh sem vsegačila, da je vsegačila v kosmologi, da je vsegačila in vsegačila, da je vsegačila, da se vsegačila, da je vsegačila, da se vsegačila, da je vsegačila, z kolaboratorom Gabriela Veneziano, ki je nekaj tudi vzelo, ali si se je vsegačila. Vsegačila je, da sem vsegačila, v november, 1997, na konferenci, na gravitasno vedat analizično v Orsay. Tibo je to pripravil, nijakom konferenci, če je pripravila v vse mačna moment, ker je to najbolj izgav na gravitasno vedat z vršenje sistemov. Neliših njense na konferenci, nekaj neko bomo srečili, vsi nekaj v HKS, Kipton, Bernerschood, Patrick Brady, Brusal and Satya Prakash, in je to všel taj zelo vršen. ki se prijezati na 1. mnah, na HES, je zelo vsečen, da je zelo vsečen na neko, z tvoj tvoj, zelo vzelo v 2, na kosmike vseče, radiečne reakcije v kosmike, z njega vsečen na zelo, zelo, zelo, gravitationalne vseče. In potem, da je zelo vsečen na mnah, da je zelo vsečen Gabrielo Veneziano, ki je vsečen na Sabati, da je, da je vsečen na HES, zelo vsečen na probleme initijne konditijne v prvej banku. In se namočila imač pa krič na 1998. Zato sem več kratil na moj nobuk. Svoj prav sam je jednjo nobuk, kaj sem počala, kot sem stvarila vso diskussion kaj ječim z gravitajnimi z vrcičenimi sistemokliče. To je prav, in ko je z evoč ovolj naša vzgleda. Na vse v 1999 v 2002 je začal, in še pričo je začal v 2002. Virgo je z njihovo početne. Protože in sem jaz bila povodil, da in s Nihakim korek, da, if Black calls, in a netonje tkod 10, so solar masses, zebila v vse basne, našli bila sranjeil sračšiselnih vsega. Prečo je, da je vsega vsega kanale in nared капri na kajom in tako je bila vsega negacid, z njega lahko v 100R. To tekdu in spremuse poživaj vsega, in nared, da je to ležite. zelo se ideli nosili template, in in drugim alternacijom, to je kratko in več meroj čampoč, ne glasbo, nekaj rada na Romaničke relativitje, ne glasbo, pa na Romaničski relativitje. V srešelju dokonjo vzelo se, v mene v početkagi, nekiči podcat, da ne glasbo, Romanički relativitje in tvoj simulacij REDATOVSKI zelo chto komplicirje. Na sezami Romanički as Ljuk was showing before, at the time the binding energy and the gravitational wave energy flux were only known at very low order, 2 pn and 2.5 post Newtonian order. And these are the crucial ingredients because to get the gravitational wave frequency and the phase, you need to solve the balance equation, so you need to know these two ingredients here. And just to show you in a plot as a function of the velocity with respect to the speed of light, this was the 1 pn result and the 2 pn result at the time. And I added just the most recent result in the last 10 years. And this was the flux. And you can see these agreements in the region of velocities where we expect actually black hole systems to merge. I'm sorry that the plot doesn't show very well here. OK, so then after starting, as I said in December of 1998, in November 1998 we put out the first paper where we looked at the conservative part of the dynamics and we mapped the two body description in a one body one with a particular mapping of the energy. We'll go into the details of this in a few slides. And then we wanted to focus on the waveform. And at the same time in 1999 I actually moved to Caltech with a fellowship in the group of kick-torn. So we finished the paper when I was there. And this was the paper of the transition from inspirant to plunger with also the waveform including the merger and the ring down. OK, so now let me describe a little bit more the details of all this framework. So the main idea was that in a post-muttonian expansion you expand everything. The Hamiltonian, for example, is post-muttonian expanded including also the probe limit, the test body limit where on the other hand we know the solution, we know exactly the solution, Schwarzschild and Kerr. And so the idea was to not expand the test body limit but keep it exact. And then map the dynamics in the dynamics of a Schwarzschild-Kerr black hole deform where the deformation parameter was the finite mass ratio, this parameter nu. Some of the ideas of this approach were inspired by quantum field theory when describing the binding energy of comparable mass charge bodies. So we came up with doing, working in a coordinate invariant manner with the Hamilton-Jakobi formalism and with this map of the energy, the effective energy and the real energy, which was very close actually, I mean, was actually mapping to similar results obtained in the 17 by Bresenicison and Zengisten for the positronium. And if you look at scattering states, this actually expression here of the effective energy can be related to the Manderson variable s and is basically the symmetric function of the asymptotic momenta which reduces in the test mass limit to the energy of the mass m2 in the rest frame of m1. Ok, so we thought we had a better amytonian so just to recap. So one start from the amytonian expanded. The work we did in 1998 was a 2pn order. So this was the effective amytonian with the potential here deformed with dependence on nu now. This was then the nu amytonian that was resum, having the exact problem it incorporated. And now all the dynamics is condensed on these two coefficients here and for circular orbits is actually the a coefficient that plays the most important role. So this was the 2pn result, 3pn, the work by Damouli Janowski-Chef and then later 4pn, 5pn is unknown today. Ok, so now one has better hopefully amytonia to describe the conservative part of the dynamics then one can write the amyton equation with this amytonian and we need it to have the radiation reaction force. And in the first paper this was resum dallappade from work that Tivo have done with Satya Prakash and Bala Hayer and then one from the waveform one can compute the waveform on the trajectory and this was the waveform during the inspiring and the late inspire as you can see here this is the evolution of the gravitational wave frequency which is twice the orbital frequency and then at the end once the two black hole form merge, sorry, a new black hole form the black hole is ringing with quasi normal modes and the waveform Well, there came some ideas from the 70s papers including etcher paper here from Remorufini of 72 were pointing out that the quasi normal modes are excited at the light ring crossing if you were studying the case of a small body plunging into a black hole and so what we did was to just add a superposition of quasi normal modes and do this very quick transition up to the final frequency of the black hole ring in the least done quasi normal mode now of course at the time there was no numerical relativity so we didn't know given the masses of the inspire what would be the final mass and the final spin of the black hole so that we guessed we took the energy at the light ring and the angular momentum at the light ring and we did basically the mass and the spin and actually this value is 10% off of what the numerical relativity predicted few years later ok, so I want actually to emphasize again this simplicity of the merger waveform going back to the tesmas limit because I think this was very crucial to then build this model extend it to spin and then use result from numerical relativity later on so if you go in the tesbody limit and you look at the perturbation the gravitational perturbation you have the regio wheel at the real equation that I wrote here this is a potential the potential actually peaks at the light ring so again the idea followed was that ok there is this potential which peaks at the light ring if now you consider a body that spiral in and plunges and go into the black hole until the moment the body goes into the black hole emits gravitational radiation from the quadrupole formula but once it is inside a potential then the direct radiation is basically filtered by the barrier and the only thing that you can see is basically this space time vibrations which leak out from the potential which we are just representing as a superposition of quasi normal modes actually assuming that the linear approximation really really works starting from merger basically ok so then in 2005 there was the black hole of numerical relativity with transfratorios first this is the famous figure from his paper and then in november 2005 the group at nasa godard and the group of campanelli and lustro also got for the first time the result or for the merger and now there is another conference which played an important role for me because at the time this is the result of Baker and Campanelli was presented at this conference in november 2005 and I had moved back to cnrs but I was back actually in Maryland at that time I was invited at the conference and at the conference I talk about this other paper which I think is my last paper with tibo except for some other papers with other many more collaborators in RER collaboration etc so this was a paper that I did while I was at IAP with tibo and Yanbei Chen where we extended the UB for the first time to spin effects although not so accurate as we have it today but still spin effects even precessing and we produced the first waveforms with precession and spin effects which are shown here non-spinning, generic up and down means spin 50% of the maximum value a little bit this misalign with the angular momentum and in this paper we also had a plot showing the signal to noise ratio at 100 megaparsec for different value of the spin of the binary as a function of the total mass for initial LIGO and you can see again that the majority of the contribution in the SNR comes from large black hole systems ok, so then I started at that point that conference as I said was very important to me because I met Franz Petorius and Greg Cook and I was very eager to compare the results of the UB formalism so these were two figures taken from the paper with Franz his first simulation of the two black holes merging this is the distorted common apparent horizon the plot I will show you before of the frequency increasing and going to the quasi normal mode is now shown here for the numerical simulation this is the peak of the luminosity this point is the light ring of the final black hole 50% of the energy is just emitted in the last part and you can see that the transition is very quick and is very rapid although very energetic so then in this paper we compare with the UB formalism for the first time in dash is the UB for the spiral plung and then mergering down so there are some differences because this is an analytical approximate model but the main features are captured and this model has no information about numerical relativity it was augmented because in the meantime the 3pm calculation was done so was included the coefficients and also at 3.5pm there were the results for the flux so at that point started because the numerical relativity now became available all this work at the interface collaborating with the numerical relativity people and I had close by the people so I started a collaboration with them, calibrating the UB and R waveforms you see here an example where now result from numerical relativity are incorporated and this was the first model that was actually used in the first search for binary black hole by LIGO and VIRGO INITIA LIGO and VIRGO so we are before advanced LIGO here and the data were taken in these two periods actually this was only LIGO and here also we joined and there were two papers by the LIGO collaboration no detection this was before 2015 but upper limits were set on the merge event rates for binary black holes and I should emphasize this was work also done with great collaboration with the group of Satya Prakash in Cardiff so at that point given the result from numerical relativity we really needed to improve even more the model, the spin sector and here I want to emphasize a couple of things so when you do the mapping of the two spins, two bodies with mass and spin you have also some more freedom where you put the spins to the central object to the test mass and I want to show you two paths so first of all if you go to the care case you can consider a test spin in care space time and this was what we followed and developed when I was at Maryland with Enrico Barrause and on the other hand starting from the paper by Tebo and then followed by as you will see in a moment by collaborators and you can also consider still a test mass in care so you can introduce the spin effect in other ways with gyro magnetic functions in fact when you go from care to the comparable mass case then just symbolically you have two possible amiltonians so one that was developed by my group in Maryland and then in Germany as you will be in A and T, you will be in SMS as I said started from Tebo and then worked by Tebo, Pioto and Gerard Scheffer and then Alessandro Naga ok the other crucial thing important thing was the improvement of the radiation reaction in the first papers we use the radiation force radiation reaction force expressed in terms of the flux and then it came this paper by Tebo and Balajer and Alessandro Naga where they suggested to rewrite the force in terms of the modes, LM and they had a resummation of the modes in factorized form inspired again by result in the test body limit ok, so now I want to move more on work closer to Advanced LIGO and then the discovery and all the work on the inference inference studies so we needed also to move all the process of calibrating this waveform and I want just to show with this plot the fact how do we do that so we start from a model which is not calibrated and if you compare with numerical relativity there is some difference after let's say 60 gravitation away cycles then we include post-neutralian corrections in the model which are unknown today and we fit them or extract them from the numerical relativity results and we get a better agreement and then because the models are originally built for quasi-circular orbit and the last part when the two body planches the motion becomes go beyond quasi-circularity then we correct with some called non-quasi-circular corrections which are again inferred from the numerical results now for what concern the work that have been involved which is the SEO-BNR we built the templates that were used in the first second and third observing run of the Lagom-Vigo collaboration working closely with the simulating a string space time collaboration so this is an example where this what I explain on the left is repeated in the parameter space of the mass ratio and the combination of the spin of the binary for different points so you see result from the Teokoski code and then you extrapolate the model everywhere else and you validate it with some numerical relativity simulation that were not used for the calibration similar work has been done also for the TOB resum S so the template bank that Lagom-Vigo has been using starting from the first run in 2015 is illustrated here this is the projection in the masses M2 of the binary there are also the direction of the spins for masses larger than 3 solar masses of the order of 300,000 SEO-BNR templates have been used for the search for lower masses plain post Newtonian templates can be used because the signal to noise ratio in the merger is basically negligible so one used just in spiral waveforms from post Newtonian theory because I will show you now some highlights of the signs with the Lagom-Vigo detectors I have to introduce very briefly other two waveform models because you will see some plots with them one is called Inspira Merger in Down Phenomenologica is built completely in the frequency domain in closed form it's quite fast because it's a frequency domain and it's built by hybridizing in time domain an effective one body waveform at low frequency and the other one which is quite more recent is to do the retrial interpolation of the numerical relativity waveform because today we have more than we had many years ago and this is called NR-surrogate however although this is a very accurate model it is confined to the region in which we know numerical relativity simulations they are very time consuming to produce and typically they are only of the order of 20 orbits so we can only use them when the total mass of the binary is quite large ok, so conclusion you cannot see this plot we have many extensions of the models including harmonics, precession parameterized form to include deviation from general relativity eccentricity etc also in the other template families but let me start with the comparison with the use with the LIGO in Virgo so this waveform were used as I said in the table bank and also for inference study for the first detection that you see on the left since then LIGO in Virgo have discovered 55 binary black holes represented in this plot here this is the mass in solar masses including two binary neutron star and two neutron star black holes that the paper just came out in June so let me start on the left you see a visualization of this event 1908-14 actually it is not in numerical relativity simulation it would be too long to produce essentially an EOBNR waveform which got a visualization and now this event why it was puzzling it was puzzling because the secondary mass between a neutron star and a black hole as the mass between a neutron star and black hole and because of the asymmetry in the masses the signal is quite rich what you see here going from 4 seconds before merger you can still see the higher mode for example the L equal 4 m equal 4 and 1 second before merger the L equal 5, m equal 5 maybe you can see it better here which is just the last part of the evolution so let me simplify properly this event and in particular to nail down the mass of the secondary we really needed to have very accurate waveforms that include higher harmonics and precession and what you see in this plot which is the posterior distribution for the secondary mass is the comparison with the two waveforms EOBNR and Phenom and you can see that the posterior become quite tight if you include higher harmonics and precession and let me emphasize was very important for this event to understand the mass of the secondary because there is this puzzle is it a black hole or a neutron star the other example I wanted to give is 1905-21 this is a simulation now a numerical simulation produced by Neil Fisher in my group and and you can see the very large masses is a very short signal unfortunately you cannot see this plot in this room these are the plots of the masses and the projection of the spin on the orbital plane here sorry on the orbital plane here along the direction perpendicular to the orbital plane on the right and you can see that now also the NR-surrogate is included because this was such a high mass system that the method that interpolate the numerical relativity simulation could be used because the waveform was quite short anyway, until now we are not dominated by systematics even if there are differences between the models we are dominated by the statistical uncertainty which is set by the signal to noise ratio in the detector so let me now say a few words about the fact that until now, until maybe a year ago all the waveforms that have been used by LIGO and VRGO were eccentricity because the most promising sources were supposed to be quasi circular but as we go to more sensitive detectors in the next few years starting from next run next August it will be very important to include eccentricity so I want to show you just an extension of the model that we did recently with spins that are not precessing, aligned but including also the higher harmonics this is the comparison with numerical relativity I wonder if you can see this plot here with eccentricity 0.06 you can also produce with a larger eccentricity 0.8 for example in red this is an example of the dynamical capture and there have been also other work in the literature and I think Sebastiano will show something later in the afternoon ok now I want to say a few words about the extension to matter binary neutron starts how much time I have ok I can speak that's a couple of minutes ok so then I go very fast so I just wanted to say Sebastiano will talk also about this so I don't have of course we want to understand the question of state of neutron star in the core and here there is as Luc was saying a new parameter which is 0 for black holes which is the tidal deformability parameter so we have extended the EOBNR for model to tidal effects in this paper with Tanya and Jan with tidal effects the potential gets the term and becomes more attractive in our group we have focus actually on the description of dynamical tides instead of adiabatic tides in the sense we have included the possibility that because of the f-mode in the neutron star the tidal force if you go in the frequency domain could excite could have a frequency that could excite actually the f-mode of the neutron star so this is just shows actually this effect of the tidal deformability depending on time or frequency and also here all the work to compare to numerical relativity has been done and these waveforms have been used for the influence of properties with the two binary neutron stars that have been detected by Ligo and Virgo so the last thing I wanted to say was the neutron star black holes also here with Andrew Matas we have extended the model to neutron star black holes you can see here the comparison of the model with the binary black holes and because of the tidal disruption there is a difference at the very end and this neutron star black hole family was in particular used for understanding the tidal effects in the neutron star black hole that was actually detected by Ligo this is a visualization by team Dietrich sorry the simulation by team Dietrich for the second event here and because of the masses and the spin of the black hole not much of a disk is formed the neutron star is swallowed all as you can see here in the animation and again the waveform were used by the Ligo and Virgo collaboration to extract the masses for this event I will not show the test of general relativity you can ask me and I just go to the end ok this is just the summary so I hope I convince you to make precise prediction of the two body dynamics and gravitational radiation has been crucial to observe and learn identify events with Ligo and Virgo with numerical relativity has been also very important I didn't talk about many things about the effect of body formalism but I hope people will discuss this here during the week bright future of this field in the next few years with Ligo and Virgo but also the next decades on the ground and in space we need to continue to improve the waveform models because of the larger sensitivity that the detector will have in the future I want to thank my group on which many things have been based for this presentation and finally I want really to thank Tibo I had a wonderful adventure here at HHS 23, 24 years ago and I really want to thank you because you really raise my interest in the subject of gravitational waves from binary system and also for many fruitful and enjoyable discussions I had with you since then