Galaxy mergers can occur when two (or more) galaxies collide. They are the most violent type of galaxy interaction. The gravitational interactions between galaxies and the friction between the gas and dust have major effects on the galaxies involved. The exact effects of such mergers depend on a wide variety of parameters such as collision angles, speeds, and relative size/composition, and are currently an extremely active area of research. Galaxy mergers are important because the merger rate is fundamental measurement of galaxy evolution. The merger rate also provides astronomers with clues about how galaxies bulked up over time. Galaxy mergers can occur when two (or more) galaxies collide. They are the most violent type of galaxy interaction. The gravitational interactions between galaxies and the friction between the gas and dust have major effects on the galaxies involved. The exact effects of such mergers depend on a wide variety of parameters such as collision angles, speeds, and relative size/composition, and are currently an extremely active area of research. Galaxy mergers are important because the merger rate is fundamental measurement of galaxy evolution. The merger rate also provides astronomers with clues about how galaxies bulked up over time. During the merger, stars and dark matter in each galaxy become affected by the approaching galaxy. Toward the late stages of the merger, the gravitational potential (i.e. the shape of the galaxy) begins changing so quickly that star orbits are greatly affected, and lose any memory of their previous orbit. This process is called violent relaxation. Thus if two disk galaxies collide, they begin with their stars in an orderly rotation in the plane of the disk. During the merger, the ordered motion is transformed into random energy. The resultant galaxy is dominated by stars that orbit the galaxy in a complex, and random, web of orbits. This is what we see in elliptical galaxies, stars on random unordered orbits. Mergers are also locations of extreme amounts of star formation. The star formation rate (SFR) during a major merger can reach thousands of solar masses worth of new stars each year, depending on the gas content of each galaxy and its redshift. Typical merger SFRs are less than 100 new solar masses per year. This is large compared to our Galaxy, which makes only a few (~2) new stars each year. Though stars almost never get close enough to actually collide in galaxy mergers, giant molecular clouds rapidly fall to the center of the galaxy where they collide with other molecular clouds. These collisions then induce condensations of these clouds into new stars. We can see this phenomenon in merging galaxies in the nearby universe. Yet, this process was more pronounced during the mergers that formed most elliptical galaxies we see today, which likely occurred 1–10 billion years ago, when there was much more gas (and thus more molecular clouds) in galaxies. Also, away from the center of the galaxy gas clouds will run into each other producing shocks which stimulate the formation of new stars in gas clouds. The result of all this violence is that galaxies tend to have little gas available to form new stars after they merge. Thus if a galaxy is involved in a major merger, and then a few billion years pass, the galaxy will have very few young stars (see Stellar evolution) left. This is what we see in today's elliptical galaxies, very little molecular gas and very few young stars. It is thought that this is because elliptical galaxies are the end products of major mergers which use up the majority of gas during the merger, and thus further star formation after the merger is quenched. Galaxy mergers can be simulated in computers, to learn more about galaxy formation. Galaxy pairs initially of any morphological type can be followed, taking into account all gravitational forces, and also the hydrodynamics and dissipation of the interstellar gas, the star formation out of the gas, and the energy and mass released back in the interstellar medium by supernovae. Such a library of galaxy merger simulations can be found on the GALMER website. A study led by Jennifer Lotz of the Space Telescope Science Institute in Baltimore, Maryland created computer simulations in order to better understand images taken by the Hubble Telescope. Lotz's team tried to account for a broad range of merger possibilities, from a pair of galaxies with equal masses joining together to an interaction between a giant galaxy and a puny one. The team also analyzed different orbits for the galaxies, possible collision impacts, and how galaxies were oriented to each other. In all, the group came up with 57 different merger scenarios and studied the mergers from 10 different viewing angles. One of the largest galaxy mergers ever observed consisted of four elliptical galaxies in the cluster CL0958+4702. It may form one of the largest galaxies in the Universe. Galaxy mergers can be classified into distinct groups due to the properties of the merging galaxies, such as their number, their comparative size and their gas richness. One study found that large galaxies merged with each other on average once over the past 9 billion years. Small galaxies were coalescing with large galaxies more frequently. Note that the Milky Way and the Andromeda Galaxy are thought to collide in about 4.5 billion years. The merging of these galaxies would classify as major as they have similar sizes. The result would therefore be an elliptical galaxy. In the standard cosmological model, any single galaxy is expected to have formed from a few or many successive mergers of dark matter haloes, in which gas cools and forms stars at the centres of the haloes, becoming the optically visible objects historically identified as galaxies during the twentieth century. Modelling the mathematical graph of the mergers of these dark matter haloes and in turn the corresponding star formation was initially treated either by analysing purely gravitational N-body simulations. or by using numerical realisations of statistical ('semi-analytical') formulae. In a 1992 observational cosmology conference in Milan, Roukema, Quinn and Peterson showed the first merger history trees of dark matter haloes extracted from cosmological N-body simulations. These merger history trees were combined with formulae for star formation rates and evolutionary population synthesis, yielding synthetic luminosity functions of galaxies (statistics of how many galaxies are intrinsically bright or faint) at different cosmological epochs. Given the complex dynamics of dark matter halo mergers, a fundamental problem in modelling merger history tree is to define when a halo at one time step is a descendant of a halo at the previous time step. Roukema's group chose to define this relation by requiring the halo at the later time step to contain strictly more than 50 percent of the particles in the halo at the earlier time step; this guaranteed that between two time steps, any halo could have at most a single descendant. This galaxy formation modelling method yields rapidly calculated models of galaxy populations with synthetic spectra and corresponding statistical properties comparable with observations.