Yandex in the new CERN experiment: how to find dark matter in just 13 years
Despite the fact that physicists are sometimes tried to be conservative, in reality they are just waiting to find something that goes beyond the current understanding of nature. But they have long failed this.
Once again, hopes for updating the Standard Model collapsed after the Higgs boson was found at CERN. And despite the fact that, according to Stephen Hawking , this discovery made physics more boring, the problems that the Standard Model cannot explain, still remain. One of them - which particle can become a candidate for dark matter ? As you know, it is contained in the Universe, but we cannot see it.
And now scientists at CERN are starting a new experiment - Search for Hidden Particles. If such particles are found, the Standard Model can be expanded. This will mean that our understanding of the structure and evolution of the Universe may change. But scientists may well claim the Nobel Prize. Astrophysical research for the SHiP will be carried out by the Astro-H space telescope. Yandex for this experiment will not only provide CERN with its machine learning technologies: students and researchers at the Yandex Data Analysis School will work together with its scientists.
')
The cooperation between Yandex and CERN began in 2011, when we gave it our servers. In 2012, we developed a search service for the organization, which was used in one of the four main experiments of CERN at the Large Hadron Collider - Large Hadron Collider beauty experiment ( LHCb ). In 2013, physicists had the opportunity to use our own machine learning technology, Matrixnet. At the same time, Yandex became an associate member of the European Center for Nuclear Research in the framework of the CERN openlab project .
Two years ago, Andrei Golutvin , scientific advisor to the director of CERN, spoke at Yandex. It was exactly one day before the discovery of the Higgs boson was officially announced. And last week, Andrew, at a special seminar, talked about the new SHiP experiment, which already at the planning stage assumes the use of technology and knowledge of Yandex. The lecture consists of five parts:
Detailed decoding - under the cut.
Any experimental physicist tried to discover something in his life. And the idea of ​​finding particles of this sort is not new. You see on this distribution the results of a large number of experiments that have already tried to detect them. Typically, these results are represented as a two-dimensional distribution. Along one axis is a certain value that characterizes the strength with which these new particles interact with known particles, and on the other scale the mass of these particles is laid. And here we see the result of a very large number of experiments that have set quite impressive limits in the absence of such particles.
It turned out that the Standard Model works very well and describes virtually all phenomena in the microworld, particle physics. And it turns out that this microcosm is not very complicated, despite the huge number of different experimental particles. Of course, you all heard about protons , neutrons , electrons , muons . In fact, there are very few fundamental particles. These are the so-called fermions (a total of six fundamental fermions of leptons and six quarks are known).
From these quarks you can build all the other mesons that we know today. The interaction between these fundamental fermions is described using three forces:
So, there are six fundamental leptons - an electron, a muon, a tau lepton and three corresponding neutrinos. Neutrinos are particles with a very small mass. They participate only in weak interactions, because neutrinos are electrically neutral.
And there are six quarks: up , down , charm , strange , top and bottom . It turns out that there are many similarities between the properties of leptons and quarks, and it is accepted to say that there are three generations of leptons and quarks: the first, second and third. In fact, the properties of the electron and the neutrino, the properties of the muon and the neutrino, the properties of the tau lepton and the tau neutrino are absolutely the same. How do they differ? Only by its mass. The mass of the muon is approximately 200 times greater than the electron, and the mass of the tau lepton is 20 times greater than the mass of the muon. Why this is so, we do not know. The same thing happens with quarks. Quarks of the first, second and third generation have exactly the same properties - only different masses. And why their mass is also completely different, we also do not know.
The interactions that these fundamental fermions experience in particle physics are commonly associated with carriers of interactions. Electromagnetic interactions are transferred by the photon, weak interactions - by these three bosons. Photon is massless, the mass of these bosons is about 100g GeV. Massless gluons tolerate strong interactions.
In fact, all physical processes predicted by the standard model were confirmed experimentally. The only particle that was absent, but was predicted by it, is the Higgs boson. The fact is that the equations of the standard model were written for massless particles, and then Higgs and several other theorists came up with a very elegant mechanism for making the particles massive using the Higgs boson.
This is a very interesting distribution, which shows the mass distribution of the Higgs boson, as it was collected during data processing. And at first, no peak is visible. Today we know that the Higgs boson exists, and its mass corresponds to this peak and is about 125 GeV. And today we can definitely say that these initial reports on the Higgs mass were correct.
It turns out that this mass is very amazing. If the Higgs is in a very narrow mass range, then it may turn out that apart from the standard model, nothing more is needed, that it can be rightfully classified as a complete quantum field theory that works up to the Planck scale, i.e. she describes all the processes in the microworld. Moreover, having Higgs with this mass, one can even describe the expansion of the Universe without resorting to any other new particles.
Nevertheless, physicists continue to search for what is called beyond standard model physics or new physics. Why do they do it? There are three major issues that the Standard Model does not explain.
The first problem is connected with experiments in the microworld. In the beginning, it was believed that the three fundamental fermions are massless. The standard model with the help of the Higgs mechanism can give mass to all other fermions and bosons, but not neutrinos. Today we know for sure that neutrinos have mass. We do not know what it is equal to, but we know for sure that there is a difference in mass between three generations of neutrinos, and we have accurately measured the difference of two masses between three generations of neutrinos. We need to come up with a mechanism to add the standard model so that the neutrino gains mass. Given the mass with which the Higgs boson was found, the Standard Model claims to be the title of a complete quantum field theory. I do not want to change its equation, inventing an additional mechanism for how to make neutrinos massive.
The second two big problems arise mainly from very detailed studies in cosmology and astronomy, which have successfully passed the last 10-20 years. Today we know for sure about the presence of dark matter. The fact is that the overwhelming gravitational mass of our Universe does not absorb and does not emit light — it is invisible. At the same time, we do not have a candidate for a particle of dark matter among known particles. Since dark matter cannot be detected, its particle must be electrically neutral, otherwise we could see it through electromagnetic interactions, and it must interact very, very weakly. In principle, an ordinary neutrino could claim to be a candidate for dark matter, but we know that it is distributed very unevenly in the Universe. Those inhomogeneities that we see in the Universe can be attributed to very small inhomogeneities in the first moments after the Big Bang. If dark matter consisted of neutrinos, then due to the fact that neutrinos are very, very light, the particles would fly apart at the same speed of light and there would be no irregularities.
There may be a more complicated explanation. We approximately know in what places dark matter is concentrated. Using the language of classical physics, you can calculate the mass of a fermion. Remember that these are particles that must satisfy the Pauli principle — you cannot have two identical fermions in one place of space. And you can calculate the mass of this fermion so that the gravitational force, which seeks to compress the clusters of dark matter, is somehow compensated. And it is compensated by the pressure of Fermi. And from here you can calculate the lower limit on the mass of such a fermion. It should be of the order of 1 keV, which is many times the mass of the neutrino. That is, the neutrino mass cannot be dark matter, which means there must be something else that would explain this huge cosmological effect.
And finally, the third major problem of the standard model. We know that when a collision occurs on accelerators, approximately the same number of particles and antiparticles are born. This happened during the Big Bang. But there is a slight violation of symmetry between the development in time of matter and antimatter. We measured it very well in quarks. This is a well-known phenomenon, the study of which was awarded two Nobel Prizes. The only problem is that the effect we observe with neutral mesons in the quark sector is 10 10 times less than the mechanism needed to explain the absence of antimatter in the Universe. And today we know that the lower limit on the admixture of antimatter in the universe is very, very small. There were even special experiments on spacecraft that tried to find traces of antimatter in the universe. In general, you need to do something with these three problems.
The fundamental fermions that are known in the standard model are three generations of quarks, three generations of leptons, interaction vectors, and the Higgs boson.
One of the very interesting things that can be done in the framework of SHiP is to try to find three more fundamental fermions, the so-called massive Majorana neutrinos . They are designated on this scheme as N 1 , N 2 , N 3 . And it turns out that using these fundamental Majorana neutrinos, dark matter can also be explained. The role of a candidate for dark matter would be played by N 1 , whose mass should be of the order of several kiloelectronvolts. For heavier N 2 and N 3 it is in the range from 0.5 GeV to 40-50 GeV. And they can explain how to make an ordinary neutrino massive, without breaking the equation of the standard model, and greatly strengthen the mechanism of symmetry breaking between matter and antimatter. Quite a large number of both Russian physicists and foreign scientists took part in creating such a model, which is called the Neutrino Minimum Standard Model.
How to find them? In order to find the N1, a program of experiments in space is planned. If you have N 1 - a candidate for dark matter, and its mass is quite large, a few kiloelectronvolts , it can decay into ordinary neutrinos, it can - into neutrinos and photons. It is important that this decay take place very, very slowly - so that the lifetime is much longer than the lifetime of the Universe. Otherwise, everything will disintegrate, and this will not explain the dark matter. The decay into three neutrinos is the main decay, but experimentally it is very difficult to detect. But the decay of such a sterile neutrino into a photon and the usual electron and neutrino can be seen in space. Since dark matter is cold, you just have two-particle decay and you need to look for a monoline in the galactic photon spectrum. In this case, the energy resolution should be very good. The fact is that kiloelectrovolts are the region of the x-ray spectrum. It is very filled with lines from known elements, therefore we know it very well. Therefore, to say that the new line does not coincide with any of the known lines, the resolution of your device must be very good.
Four astro-missions are currently being discussed, with the launch of Astro-H scheduled for 2015. The photon detector, which will be located on Astro – H, has the required resolution. Since dark matter surrounds us from all sides, it was ideal that it was located on a space satellite and could scan 4π - all areas of the Universe. But, as a rule, astrophysical experiments are aimed at studying a certain interval of angles and their aperture is very narrow. But the three astromissies, whose launch is being discussed for the years 2019-2020, should have a much wider viewing angle.
To find particles N 2 and N 3 , we need experiments on accelerators. It is expected that the mass of these particles is in the region from 0.5 GeV to 30-40 GeV. Their search is one of the tasks of the SHiP experiment. Since the probability of the appearance of such particles is a very, very rare process, it is necessary to ensure that the data array where such particles can appear is also very, very large. Where is this available at CERN accelerators? This is a chain of all accelerators that exists. The SHiP experiment is proposed to be carried out on the SPS accelerator.
The fact is that the lifetime of a proton beam in a large hadron collider is approximately 20 hours. That is, after the proton injection from the SPS to the Large Hadron Collider has passed, you, roughly speaking, have 20 hours to use the SPS proton accelerators for something else. In fact, until recently, these protons were used for the production of ordinary neutrinos, and these neutrinos were taken along this line and then sent to a special underground laboratory, which is located in Italy at a distance of about 750 km from CERN. These experiments have been completed, and today 70% of the protons that can be used and which are produced on the SPS accelerator are actually free. And one of the motivations of the SHiP experiment is to understand how to use these protons.
This is the place where the SHiP experiment is planned. It will require a rather expensive proton extraction line from the SPS and a rather large hall (approximately 20x20 meters and a length of about 100 meters). Creating such additional infrastructure at CERN costs about 100-120 million Swiss francs. The first engineering study of such a special bundle line will be published now.
What should the SHiP experiment look like? For every second you can produce a clot of 5x10 13 protons. And such a conclusion can be done every seven seconds. If the experiment in this mode will work for 3-4 years, then you can type in approximately 2x10 20 proton interactions. They all need to be analyzed, and to find among them several signaling events. These events will be searched for in this experimental setup, but between the site of the impact of the protons on the target and the installation, special protection should be located that will allow to absorb unwanted particles in order to leave only those that interact very weakly and which pass through this protection.
This is a very schematic view of this setup. Now it is planned that the length of this cylinder will be approximately 50 meters, diameter - 5 meters. Thus, a heavy Majorana neutrino, which is born as a result of interactions of 2x10 20 protons, will fly through this protection with a target, will fall into a vacuum volume and will decay into two other particles, for example, pion and muon. It is necessary in some way to measure the mass of this decay peak on a pion and a muon and identify the particles into which this heavy neutrino has decayed. Installation is not very complicated, because all the technologies that are needed to create it exist. The most difficult thing in planning this experiment is to understand very well how we can remove the backgrounds from this huge number of initial 2x10 20 interactions so that the final background is zero. But zero does not happen, say, 0.1 with an accuracy of 1%. If we know this for sure, then each signal event registered in such an experiment will indicate that this is a signal, and not a fluctuation of the background.
To compare what the SHiP experiment can do compared to previous experiments, again let's look at this picture: "The strength of the interaction of these heavy neutral neutrinos with conventional neutrinos and mass."
It can be seen that the region that the SHiP experiment can cover, it covers the results achieved by previous experiments by several orders of magnitude. That, in general, is encouraging.
In October 2013, we published a proposal for this experiment. There were 12 or 14 authors. It immediately attracted attention at CERN, and we were invited to speak at the seminar before the relevant committee, which is responsible for evaluating our idea: how much it costs, whether it is possible to implement it. From his recommendations all our next steps will depend. Immediately after this seminar, which took place on October 23, this SPS committee (this is an abbreviation of the accelerator, on which this experiment will be done) appointed a team of referees who asked us quite a few questions. We answered these questions on January 3. And on January 15 we were recommended what is written on this slide. Please note: the essential part is framed. We were required to write a technical proposal and show how much interest in this project exists in the world.
We created a website with experiment details. Then we held the first two-day rally of the SHiP collaboration. It was June 10-12 in Zurich. On the first day we discussed the physical program (what other experiments, in addition to the search for the Majorana neutrino, can still be carried out). And the second day was devoted to various technologies and issues of computing, which need to pay attention to write a technical proposal.
It was a fairly successful rally, because after it and until today we received offers from 41st groups from 15 countries. From Russia, Switzerland, Great Britain, Italy, France and from the Yandex Data Analysis School.
These are approximate stages of this experiment. I must say at once that the next stage is possible only with the approval of the previous stage by the relevant CERN committee. Now we are busy to consider all these proposals from 41 groups, and to answer who we need in order to write a technical proposal.
Technical proposal must be written quickly. We were asked to do it by next year. And our plan is to publish it in March 2015. Then we will need money to start production of the detector. We are planning to collect it and start data collection in 2023. And then, having worked for four years, you can dial 2 to 10 20 interactions, and when I turn 70 in 2027, publish the results.
What should be reflected in the technical proposal? Of course, we need to show what other physics can be done in addition to the search for such Majorana neutrinos. We need to show that this detector can be made on technologies that are known today. Or offer some new technology.
And we need to offer some kind of data processing model. And actually, why am I here today. Now we are at the initial stage of project creation, and therefore, if there are any ideas about new technologies, then, of course, it's time to start discussing them. Because when a group joins an experiment that is already in adulthood, it is virtually impossible to convince people that their concepts are right or wrong.
What exactly is the complexity of the SHiP experiment data processing model? I do not think that when everything is done and the data set starts, this project will be linked to the requirements of analyzing a very large amount of data. Compared to the large Hadron Collider, this will be a very light project. But in order to design this experiment, we need a very competent infrastructure. Because we want to start with 10 20 interactions, among which, if we are lucky, there will be 5-10 events from the signal. We need to turn these 2x10 20 interactions to zero. Using hardware and software. And for this, a very friendly infrastructure is very important, which allows you to keep track of how much background was suppressed. And since this work will be carried out by a large number of researchers from different countries, the main request is to develop such a model. And we very much hope that we will have every reason to accept the Yandex Data Analysis School in our collaboration.
Once again, hopes for updating the Standard Model collapsed after the Higgs boson was found at CERN. And despite the fact that, according to Stephen Hawking , this discovery made physics more boring, the problems that the Standard Model cannot explain, still remain. One of them - which particle can become a candidate for dark matter ? As you know, it is contained in the Universe, but we cannot see it.
And now scientists at CERN are starting a new experiment - Search for Hidden Particles. If such particles are found, the Standard Model can be expanded. This will mean that our understanding of the structure and evolution of the Universe may change. But scientists may well claim the Nobel Prize. Astrophysical research for the SHiP will be carried out by the Astro-H space telescope. Yandex for this experiment will not only provide CERN with its machine learning technologies: students and researchers at the Yandex Data Analysis School will work together with its scientists.
')
The cooperation between Yandex and CERN began in 2011, when we gave it our servers. In 2012, we developed a search service for the organization, which was used in one of the four main experiments of CERN at the Large Hadron Collider - Large Hadron Collider beauty experiment ( LHCb ). In 2013, physicists had the opportunity to use our own machine learning technology, Matrixnet. At the same time, Yandex became an associate member of the European Center for Nuclear Research in the framework of the CERN openlab project .
Two years ago, Andrei Golutvin , scientific advisor to the director of CERN, spoke at Yandex. It was exactly one day before the discovery of the Higgs boson was officially announced. And last week, Andrew, at a special seminar, talked about the new SHiP experiment, which already at the planning stage assumes the use of technology and knowledge of Yandex. The lecture consists of five parts:
- why do we need the SHiP experiment,
- Standard Model issues
- how the detector is arranged and what it should measure
- how an international collaboration is created to create and conduct a large experiment,
- main stages of the experiment,
- what the SHiP collaboration expects from Yandex.
Detailed decoding - under the cut.
Why do we need SHiP experiment
Any experimental physicist tried to discover something in his life. And the idea of ​​finding particles of this sort is not new. You see on this distribution the results of a large number of experiments that have already tried to detect them. Typically, these results are represented as a two-dimensional distribution. Along one axis is a certain value that characterizes the strength with which these new particles interact with known particles, and on the other scale the mass of these particles is laid. And here we see the result of a very large number of experiments that have set quite impressive limits in the absence of such particles.
It turned out that the Standard Model works very well and describes virtually all phenomena in the microworld, particle physics. And it turns out that this microcosm is not very complicated, despite the huge number of different experimental particles. Of course, you all heard about protons , neutrons , electrons , muons . In fact, there are very few fundamental particles. These are the so-called fermions (a total of six fundamental fermions of leptons and six quarks are known).
From these quarks you can build all the other mesons that we know today. The interaction between these fundamental fermions is described using three forces:
- Electromagnetic Interactions. Describe the interaction of charged particles.
- Weak interactions. They involve leptons (for example, neutrinos , electrons, muons . But electrons and muons also participate in electromagnetic interactions, because they are charged), quarks.
- Strong interactions. They participate in quarks. They are needed so that you can create a proton and a neutron.
So, there are six fundamental leptons - an electron, a muon, a tau lepton and three corresponding neutrinos. Neutrinos are particles with a very small mass. They participate only in weak interactions, because neutrinos are electrically neutral.
And there are six quarks: up , down , charm , strange , top and bottom . It turns out that there are many similarities between the properties of leptons and quarks, and it is accepted to say that there are three generations of leptons and quarks: the first, second and third. In fact, the properties of the electron and the neutrino, the properties of the muon and the neutrino, the properties of the tau lepton and the tau neutrino are absolutely the same. How do they differ? Only by its mass. The mass of the muon is approximately 200 times greater than the electron, and the mass of the tau lepton is 20 times greater than the mass of the muon. Why this is so, we do not know. The same thing happens with quarks. Quarks of the first, second and third generation have exactly the same properties - only different masses. And why their mass is also completely different, we also do not know.
The interactions that these fundamental fermions experience in particle physics are commonly associated with carriers of interactions. Electromagnetic interactions are transferred by the photon, weak interactions - by these three bosons. Photon is massless, the mass of these bosons is about 100g GeV. Massless gluons tolerate strong interactions.
In fact, all physical processes predicted by the standard model were confirmed experimentally. The only particle that was absent, but was predicted by it, is the Higgs boson. The fact is that the equations of the standard model were written for massless particles, and then Higgs and several other theorists came up with a very elegant mechanism for making the particles massive using the Higgs boson.
This is a very interesting distribution, which shows the mass distribution of the Higgs boson, as it was collected during data processing. And at first, no peak is visible. Today we know that the Higgs boson exists, and its mass corresponds to this peak and is about 125 GeV. And today we can definitely say that these initial reports on the Higgs mass were correct.
It turns out that this mass is very amazing. If the Higgs is in a very narrow mass range, then it may turn out that apart from the standard model, nothing more is needed, that it can be rightfully classified as a complete quantum field theory that works up to the Planck scale, i.e. she describes all the processes in the microworld. Moreover, having Higgs with this mass, one can even describe the expansion of the Universe without resorting to any other new particles.
Problems of the Standard Model
Nevertheless, physicists continue to search for what is called beyond standard model physics or new physics. Why do they do it? There are three major issues that the Standard Model does not explain.
The first problem is connected with experiments in the microworld. In the beginning, it was believed that the three fundamental fermions are massless. The standard model with the help of the Higgs mechanism can give mass to all other fermions and bosons, but not neutrinos. Today we know for sure that neutrinos have mass. We do not know what it is equal to, but we know for sure that there is a difference in mass between three generations of neutrinos, and we have accurately measured the difference of two masses between three generations of neutrinos. We need to come up with a mechanism to add the standard model so that the neutrino gains mass. Given the mass with which the Higgs boson was found, the Standard Model claims to be the title of a complete quantum field theory. I do not want to change its equation, inventing an additional mechanism for how to make neutrinos massive.
The second two big problems arise mainly from very detailed studies in cosmology and astronomy, which have successfully passed the last 10-20 years. Today we know for sure about the presence of dark matter. The fact is that the overwhelming gravitational mass of our Universe does not absorb and does not emit light — it is invisible. At the same time, we do not have a candidate for a particle of dark matter among known particles. Since dark matter cannot be detected, its particle must be electrically neutral, otherwise we could see it through electromagnetic interactions, and it must interact very, very weakly. In principle, an ordinary neutrino could claim to be a candidate for dark matter, but we know that it is distributed very unevenly in the Universe. Those inhomogeneities that we see in the Universe can be attributed to very small inhomogeneities in the first moments after the Big Bang. If dark matter consisted of neutrinos, then due to the fact that neutrinos are very, very light, the particles would fly apart at the same speed of light and there would be no irregularities.
There may be a more complicated explanation. We approximately know in what places dark matter is concentrated. Using the language of classical physics, you can calculate the mass of a fermion. Remember that these are particles that must satisfy the Pauli principle — you cannot have two identical fermions in one place of space. And you can calculate the mass of this fermion so that the gravitational force, which seeks to compress the clusters of dark matter, is somehow compensated. And it is compensated by the pressure of Fermi. And from here you can calculate the lower limit on the mass of such a fermion. It should be of the order of 1 keV, which is many times the mass of the neutrino. That is, the neutrino mass cannot be dark matter, which means there must be something else that would explain this huge cosmological effect.
And finally, the third major problem of the standard model. We know that when a collision occurs on accelerators, approximately the same number of particles and antiparticles are born. This happened during the Big Bang. But there is a slight violation of symmetry between the development in time of matter and antimatter. We measured it very well in quarks. This is a well-known phenomenon, the study of which was awarded two Nobel Prizes. The only problem is that the effect we observe with neutral mesons in the quark sector is 10 10 times less than the mechanism needed to explain the absence of antimatter in the Universe. And today we know that the lower limit on the admixture of antimatter in the universe is very, very small. There were even special experiments on spacecraft that tried to find traces of antimatter in the universe. In general, you need to do something with these three problems.
The fundamental fermions that are known in the standard model are three generations of quarks, three generations of leptons, interaction vectors, and the Higgs boson.
One of the very interesting things that can be done in the framework of SHiP is to try to find three more fundamental fermions, the so-called massive Majorana neutrinos . They are designated on this scheme as N 1 , N 2 , N 3 . And it turns out that using these fundamental Majorana neutrinos, dark matter can also be explained. The role of a candidate for dark matter would be played by N 1 , whose mass should be of the order of several kiloelectronvolts. For heavier N 2 and N 3 it is in the range from 0.5 GeV to 40-50 GeV. And they can explain how to make an ordinary neutrino massive, without breaking the equation of the standard model, and greatly strengthen the mechanism of symmetry breaking between matter and antimatter. Quite a large number of both Russian physicists and foreign scientists took part in creating such a model, which is called the Neutrino Minimum Standard Model.
How will the detector be arranged and what should it measure?
How to find them? In order to find the N1, a program of experiments in space is planned. If you have N 1 - a candidate for dark matter, and its mass is quite large, a few kiloelectronvolts , it can decay into ordinary neutrinos, it can - into neutrinos and photons. It is important that this decay take place very, very slowly - so that the lifetime is much longer than the lifetime of the Universe. Otherwise, everything will disintegrate, and this will not explain the dark matter. The decay into three neutrinos is the main decay, but experimentally it is very difficult to detect. But the decay of such a sterile neutrino into a photon and the usual electron and neutrino can be seen in space. Since dark matter is cold, you just have two-particle decay and you need to look for a monoline in the galactic photon spectrum. In this case, the energy resolution should be very good. The fact is that kiloelectrovolts are the region of the x-ray spectrum. It is very filled with lines from known elements, therefore we know it very well. Therefore, to say that the new line does not coincide with any of the known lines, the resolution of your device must be very good.
Four astro-missions are currently being discussed, with the launch of Astro-H scheduled for 2015. The photon detector, which will be located on Astro – H, has the required resolution. Since dark matter surrounds us from all sides, it was ideal that it was located on a space satellite and could scan 4π - all areas of the Universe. But, as a rule, astrophysical experiments are aimed at studying a certain interval of angles and their aperture is very narrow. But the three astromissies, whose launch is being discussed for the years 2019-2020, should have a much wider viewing angle.
To find particles N 2 and N 3 , we need experiments on accelerators. It is expected that the mass of these particles is in the region from 0.5 GeV to 30-40 GeV. Their search is one of the tasks of the SHiP experiment. Since the probability of the appearance of such particles is a very, very rare process, it is necessary to ensure that the data array where such particles can appear is also very, very large. Where is this available at CERN accelerators? This is a chain of all accelerators that exists. The SHiP experiment is proposed to be carried out on the SPS accelerator.
The fact is that the lifetime of a proton beam in a large hadron collider is approximately 20 hours. That is, after the proton injection from the SPS to the Large Hadron Collider has passed, you, roughly speaking, have 20 hours to use the SPS proton accelerators for something else. In fact, until recently, these protons were used for the production of ordinary neutrinos, and these neutrinos were taken along this line and then sent to a special underground laboratory, which is located in Italy at a distance of about 750 km from CERN. These experiments have been completed, and today 70% of the protons that can be used and which are produced on the SPS accelerator are actually free. And one of the motivations of the SHiP experiment is to understand how to use these protons.
This is the place where the SHiP experiment is planned. It will require a rather expensive proton extraction line from the SPS and a rather large hall (approximately 20x20 meters and a length of about 100 meters). Creating such additional infrastructure at CERN costs about 100-120 million Swiss francs. The first engineering study of such a special bundle line will be published now.
What should the SHiP experiment look like? For every second you can produce a clot of 5x10 13 protons. And such a conclusion can be done every seven seconds. If the experiment in this mode will work for 3-4 years, then you can type in approximately 2x10 20 proton interactions. They all need to be analyzed, and to find among them several signaling events. These events will be searched for in this experimental setup, but between the site of the impact of the protons on the target and the installation, special protection should be located that will allow to absorb unwanted particles in order to leave only those that interact very weakly and which pass through this protection.
This is a very schematic view of this setup. Now it is planned that the length of this cylinder will be approximately 50 meters, diameter - 5 meters. Thus, a heavy Majorana neutrino, which is born as a result of interactions of 2x10 20 protons, will fly through this protection with a target, will fall into a vacuum volume and will decay into two other particles, for example, pion and muon. It is necessary in some way to measure the mass of this decay peak on a pion and a muon and identify the particles into which this heavy neutrino has decayed. Installation is not very complicated, because all the technologies that are needed to create it exist. The most difficult thing in planning this experiment is to understand very well how we can remove the backgrounds from this huge number of initial 2x10 20 interactions so that the final background is zero. But zero does not happen, say, 0.1 with an accuracy of 1%. If we know this for sure, then each signal event registered in such an experiment will indicate that this is a signal, and not a fluctuation of the background.
To compare what the SHiP experiment can do compared to previous experiments, again let's look at this picture: "The strength of the interaction of these heavy neutral neutrinos with conventional neutrinos and mass."
It can be seen that the region that the SHiP experiment can cover, it covers the results achieved by previous experiments by several orders of magnitude. That, in general, is encouraging.
How an international collaboration is created
In October 2013, we published a proposal for this experiment. There were 12 or 14 authors. It immediately attracted attention at CERN, and we were invited to speak at the seminar before the relevant committee, which is responsible for evaluating our idea: how much it costs, whether it is possible to implement it. From his recommendations all our next steps will depend. Immediately after this seminar, which took place on October 23, this SPS committee (this is an abbreviation of the accelerator, on which this experiment will be done) appointed a team of referees who asked us quite a few questions. We answered these questions on January 3. And on January 15 we were recommended what is written on this slide. Please note: the essential part is framed. We were required to write a technical proposal and show how much interest in this project exists in the world.
We created a website with experiment details. Then we held the first two-day rally of the SHiP collaboration. It was June 10-12 in Zurich. On the first day we discussed the physical program (what other experiments, in addition to the search for the Majorana neutrino, can still be carried out). And the second day was devoted to various technologies and issues of computing, which need to pay attention to write a technical proposal.
It was a fairly successful rally, because after it and until today we received offers from 41st groups from 15 countries. From Russia, Switzerland, Great Britain, Italy, France and from the Yandex Data Analysis School.
These are approximate stages of this experiment. I must say at once that the next stage is possible only with the approval of the previous stage by the relevant CERN committee. Now we are busy to consider all these proposals from 41 groups, and to answer who we need in order to write a technical proposal.
Technical proposal must be written quickly. We were asked to do it by next year. And our plan is to publish it in March 2015. Then we will need money to start production of the detector. We are planning to collect it and start data collection in 2023. And then, having worked for four years, you can dial 2 to 10 20 interactions, and when I turn 70 in 2027, publish the results.
What should be reflected in the technical proposal? Of course, we need to show what other physics can be done in addition to the search for such Majorana neutrinos. We need to show that this detector can be made on technologies that are known today. Or offer some new technology.
And we need to offer some kind of data processing model. And actually, why am I here today. Now we are at the initial stage of project creation, and therefore, if there are any ideas about new technologies, then, of course, it's time to start discussing them. Because when a group joins an experiment that is already in adulthood, it is virtually impossible to convince people that their concepts are right or wrong.
What SHiP collaboration expects from Yandex
What exactly is the complexity of the SHiP experiment data processing model? I do not think that when everything is done and the data set starts, this project will be linked to the requirements of analyzing a very large amount of data. Compared to the large Hadron Collider, this will be a very light project. But in order to design this experiment, we need a very competent infrastructure. Because we want to start with 10 20 interactions, among which, if we are lucky, there will be 5-10 events from the signal. We need to turn these 2x10 20 interactions to zero. Using hardware and software. And for this, a very friendly infrastructure is very important, which allows you to keep track of how much background was suppressed. And since this work will be carried out by a large number of researchers from different countries, the main request is to develop such a model. And we very much hope that we will have every reason to accept the Yandex Data Analysis School in our collaboration.
Source: https://habr.com/ru/post/229821/
All Articles