CINXE.COM

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">     <html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en" xmlns:og="http://ogp.me/ns#" xmlns:fb="http://ogp.me/ns/fb#"> <head> <title>Machine learning based crystal collimator alignment optimization - CERN Document Server</title> <link href='https://framework.web.cern.ch/framework/2.0/fonts/PTSansWeb/PTSansWeb.css' rel='stylesheet' type='text/css' /> <link rel="stylesheet" href="https://cds.cern.ch/img/invenio.css?v=20141127" type="text/css" /> <link rel="stylesheet" href="https://cds.cern.ch/img/cern_theme/css/cern_theme.css?v=20141127" type="text/css" /> <link rel="stylesheet"href="/css/font-awesome.min.css"> <meta http-equiv="X-UA-Compatible" content="IE=Edge"/> <link rel="stylesheet" href="https://cds.cern.ch/img/cern_toolbar/css/toolbar.css" type="text/css" />    <link rel="canonical" href="https://cds.cern.ch/record/2912720" /> <link rel="alternate" hreflang="el" href="https://cds.cern.ch/record/2912720?ln=el" /> <link rel="alternate" hreflang="fr" href="https://cds.cern.ch/record/2912720?ln=fr" /> <link rel="alternate" hreflang="bg" href="https://cds.cern.ch/record/2912720?ln=bg" /> <link rel="alternate" hreflang="zh-TW" href="https://cds.cern.ch/record/2912720?ln=zh_TW" /> <link rel="alternate" hreflang="pt" href="https://cds.cern.ch/record/2912720?ln=pt" /> <link rel="alternate" hreflang="no" href="https://cds.cern.ch/record/2912720?ln=no" /> <link rel="alternate" hreflang="hr" href="https://cds.cern.ch/record/2912720?ln=hr" /> <link rel="alternate" hreflang="ca" href="https://cds.cern.ch/record/2912720?ln=ca" /> <link rel="alternate" hreflang="de" href="https://cds.cern.ch/record/2912720?ln=de" /> <link rel="alternate" hreflang="it" href="https://cds.cern.ch/record/2912720?ln=it" /> <link rel="alternate" hreflang="zh-CN" href="https://cds.cern.ch/record/2912720?ln=zh_CN" /> <link rel="alternate" hreflang="sv" href="https://cds.cern.ch/record/2912720?ln=sv" /> <link rel="alternate" hreflang="sk" href="https://cds.cern.ch/record/2912720?ln=sk" /> <link rel="alternate" hreflang="en" href="https://cds.cern.ch/record/2912720?ln=en" /> <link rel="alternate" hreflang="pl" href="https://cds.cern.ch/record/2912720?ln=pl" /> <link rel="alternate" hreflang="ru" href="https://cds.cern.ch/record/2912720?ln=ru" /> <link rel="alternate" hreflang="ka" href="https://cds.cern.ch/record/2912720?ln=ka" /> <link rel="alternate" hreflang="ja" href="https://cds.cern.ch/record/2912720?ln=ja" /> <link rel="alternate" hreflang="es" href="https://cds.cern.ch/record/2912720?ln=es" /> <link rel="alternate" type="application/rss+xml" title="CERN Document Server RSS" href="/rss?ln=en" /> <link rel="search" type="application/opensearchdescription+xml" href="https://cds.cern.ch/opensearchdescription" title="CERN Document Server" /> <link rel="unapi-server" type="application/xml" title="unAPI" href="https://cds.cern.ch/unapi" /> <link rel="apple-touch-icon" href="/apple-touch-icon.png"/> <link rel="apple-touch-icon-precomposed" href="/apple-touch-icon-precomposed.png"/> <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> <meta http-equiv="Content-Language" content="en" /> <meta name="description" content="In the CERN Large Hadron Collider (LHC), bent crystals are used to efficiently deflect beam halo particles toward secondary collimators used as absorbers. In this crystal collimation scheme, a crystal with a length of a few millimeters can produce a deflection equivalent to a magnetic field of hundreds of Tesla at LHC top energies, improving the cleaning performance of the machine. However, crystals must be in optimal alignment with respect to the circulating beam to maximize the efficiency of the channeling process. A newly developed machine learning model automatically classifies the channeling condition of crystals using beam loss monitor signals during slow rotation of the crystal. This advancement represents a crucial step toward refining the process of identifying the optimal channeling orientation. The algorithm has been tested for the fist time in operation with Pb ion beams at the record energy of 6.8 Z TeV demonstrating its reliability. Ricci, Gianmarco; D'Andrea, Marco; Di Castro, Mario; Matheson, Eloise; Mirarchi, Daniele; Mostacci, Andrea; Redaelli, Stefano" /> <meta name="keywords" content="CERN Document Server, WebSearch, CERN Document Server" /> <script type="text/javascript" src="https://cds.cern.ch/js/jquery.min.js"></script>  <link rel="stylesheet" href="https://cds.cern.ch/img/webnews.css" type="text/css" />  <script type="text/javascript" src="https://cds.cern.ch/js/webnews.js?v=20131009"></script> <meta property="fb:app_id" content="137353533001720"/> <script type="text/x-mathjax-config"> MathJax.Hub.Config({ tex2jax: {inlineMath: [['$','$']], processEscapes: true}, showProcessingMessages: false, messageStyle: "none" }); </script> <script src="/MathJax/MathJax.js?config=TeX-AMS_CHTML" type="text/javascript"> </script>  <meta content="APS : Machine learning based crystal collimator alignment optimization" name="citation_title" /> <meta content="Ricci, Gianmarco" name="citation_author" /> <meta content="Di Castro, Mario" name="citation_author" /> <meta content="Mostacci, Andrea" name="citation_author" /> <meta content="Mirarchi, Daniele" name="citation_author" /> <meta content="Redaelli, Stefano" name="citation_author" /> <meta content="D'Andrea, Marco" name="citation_author" /> <meta content="Matheson, Eloise" name="citation_author" /> <meta content="10.1103/PhysRevAccelBeams.27.093001" name="citation_doi" /> <meta content="Phys. Rev. Accel. Beams" name="citation_journal_title" /> <meta content="27" name="citation_volume" /> <meta content="093001" name="citation_firstpage" /> <meta content="2024" name="citation_publication_date" /> <meta name="citation_online_date" content="2024/10/08"> <meta content="10.1103/PhysRevAccelBeams.27.093001" name="citation_doi" /> <meta name="citation_pdf_url" content="https://cds.cern.ch/record/2912720/files/Publication.pdf" />  <meta content="APS" property="og:title" /> <meta content="Machine learning based crystal collimator alignment optimization" property="og:title" /> <meta content="website" property="og:type" /> <meta content="https://cds.cern.ch/record/2912720" property="og:url" /> <meta content="CERN Document Server" property="og:site_name" /> <meta content="APS" property="og:description" /> <meta content="In the CERN Large Hadron Collider (LHC), bent crystals are used to efficiently deflect beam halo particles toward secondary collimators used as absorbers. In this crystal collimation scheme, a crystal with a length of a few millimeters can produce a deflection equivalent to a magnetic field of hundreds of Tesla at LHC top energies, improving the cleaning performance of the machine. However, crystals must be in optimal alignment with respect to the circulating beam to maximize the efficiency of the channeling process. A newly developed machine learning model automatically classifies the channeling condition of crystals using beam loss monitor signals during slow rotation of the crystal. This advancement represents a crucial step toward refining the process of identifying the optimal channeling orientation. The algorithm has been tested for the fist time in operation with Pb ion beams at the record energy of 6.8 Z TeV demonstrating its reliability." property="og:description" />  <meta content="summary" name="twitter:card" /> <style></style> </head> <body class="CERN32Document32Server search" lang="en">  <div id="cern-toolbar"> <h1><a href="http://cern.ch" title="CERN">CERN <span>Accelerating science</span></a></h1> <ul> <li class="cern-accountlinks"><a class="cern-account" href="https://cds.cern.ch/youraccount/login?ln=en&referer=https%3A//cds.cern.ch/record/2912720" title="Sign in to your CERN account">Sign in</a></li> <li><a class="cern-directory" href="http://cern.ch/directory" title="Search CERN resources and browse the directory">Directory</a></li> </ul> </div>   <div role="banner" class="clearfix" id="header"> <div class="header-inner inner"> <hgroup class="clearfix"> <h2 id="site-name"> <a rel="home" title="Home" href="/"><span>CERN Document Server</span></a> </h2> <h3 id="site-slogan">Access articles, reports and multimedia content in HEP</h3> </hgroup> <div role="navigation" id="main-navigation" class="cdsmenu"> <h2 class="element-invisible">Main menu</h2><ul class="links inline clearfix"> <li class="menu-386 first active-trail"><a class="active-trail" href="https://cds.cern.ch/?ln=en">Search</a></li> <li class="menu-444 "><a class="" title="" href="https://cds.cern.ch/submit?ln=en">Submit</a></li> <li class="menu-426 "><a class="" href="https://cds.cern.ch/help/?ln=en">Help</a></li> <li class="leaf hassubcdsmenu"> <a hreflang="en" class="header" href="https://cds.cern.ch/youraccount/display?ln=en">Personalize</a> <ul class="subsubcdsmenu"><li><a href="https://cds.cern.ch/youralerts/list?ln=en">Your alerts</a></li><li><a href="https://cds.cern.ch/yourbaskets/display?ln=en">Your baskets</a></li><li><a href="https://cds.cern.ch/yourcomments?ln=en">Your comments</a></li><li><a href="https://cds.cern.ch/youralerts/display?ln=en">Your searches</a></li></ul></li> </ul> </div> </div> </div>  <table class="navtrailbox"> <tr> <td class="navtrailboxbody"> <a href="/?ln=en" class="navtrail">Home</a> > Machine learning based crystal collimator alignment optimization </td> </tr> </table> </div> <div class="pagebody"><div class="pagebodystripemiddle"> <div class="detailedrecordbox"> <div class="detailedrecordtabs"> <div> <ul class="detailedrecordtabs"><li class="on first"><a href="/record/2912720/?ln=en">Information </a></li><li class=""><a href="/record/2912720/files?ln=en">Files </a></li></ul> <div id="tabsSpacer" style="clear:both;height:0px"> </div></div> </div> <div class="detailedrecordboxcontent"> <div class="top-left-folded"></div> <div class="top-right-folded"></div> <div class="inside">  <abbr class="unapi-id" title="2912720"></abbr> <style type="text/css">  </style> <table class="formatRecordTableFullWidth" > <tr> <td class="formatRecordHeader" style="background-image: url('https://cds.cern.ch/img/journals.jpg');" colspan="2">  Article </td> </tr> <script type="text/javascript"> $( document ).ready(function() { $('.showAuthor').on('click', function() { var author = '<p>' + $(this).data('name') + '</p>'; var affiliation = $(this).data('affiliation') + '</br>'; var contribution = $(this).data('contribution') + '</br>'; $.magnificPopup.open({ items: { src: '<div id="ovelary-mathjax" class="overlay-white oc-content overlay-white-500">' + author + affiliation + contribution + '</div>', type: 'inline' }, callbacks: { open: function() { var div = document.getElementById("overlay-mathjax") MathJax.Hub.Queue(["Typeset", MathJax.Hub, div]); }, } }) }) }); </script> <tr><td class="formatRecordLabel"> Title </td><td style="padding-left:5px;"><b>Machine learning based crystal collimator alignment optimization</b></td></tr> <tr><td class="formatRecordLabel"><span style="white-space:nowrap;"> Author(s) </span> </td><td style="padding-left:5px;"><a href="https://cds.cern.ch/search?f=author&p=Ricci%2C%20Gianmarco&ln=en">Ricci, Gianmarco</a> (CERN) ; <a href="https://cds.cern.ch/search?f=author&p=D%27Andrea%2C%20Marco&ln=en">D'Andrea, Marco</a> (CERN) ; <a href="https://cds.cern.ch/search?f=author&p=Di%20Castro%2C%20Mario&ln=en">Di Castro, Mario</a> (CERN) ; <a href="https://cds.cern.ch/search?f=author&p=Matheson%2C%20Eloise&ln=en">Matheson, Eloise</a> (CERN) ; <a href="https://cds.cern.ch/search?f=author&p=Mirarchi%2C%20Daniele&ln=en">Mirarchi, Daniele</a> (CERN) ; <a href="https://cds.cern.ch/search?f=author&p=Mostacci%2C%20Andrea&ln=en">Mostacci, Andrea</a> (CERN) ; <a href="https://cds.cern.ch/search?f=author&p=Redaelli%2C%20Stefano&ln=en">Redaelli, Stefano</a> (CERN)</td></tr> <tr><td class="formatRecordLabel"> Publication </td><td style="padding-left:5px;">2024</td></tr> <tr><td class="formatRecordLabel"> Number of pages </td><td style="padding-left:5px;">8</td></tr> <tr><td class="formatRecordLabel"> In: </td><td style="padding-left:5px;"><a href="http://dx.doi.org/10.1103/PhysRevAccelBeams.27.093001"><i>Phys. Rev. Accel. Beams</i> 27 (2024) 093001</a> </a></td></tr> <tr><td class="formatRecordLabel"> DOI </td><td style="padding-left:5px;"><a href="http://dx.doi.org/10.1103/PhysRevAccelBeams.27.093001" title="DOI" target="_blank">10.1103/PhysRevAccelBeams.27.093001</a> (publication) <tr><td class="formatRecordLabel"> Subject category </td><td style="padding-left:5px;">Accelerators and Storage Rings</td></tr> <tr><td class="formatRecordLabel"> Abstract </td><td style="padding-left:5px;">In the CERN Large Hadron Collider (LHC), bent crystals are used to efficiently deflect beam halo particles toward secondary collimators used as absorbers. In this crystal collimation scheme, a crystal with a length of a few millimeters can produce a deflection equivalent to a magnetic field of hundreds of Tesla at LHC top energies, improving the cleaning performance of the machine. However, crystals must be in optimal alignment with respect to the circulating beam to maximize the efficiency of the channeling process. A newly developed machine learning model automatically classifies the channeling condition of crystals using beam loss monitor signals during slow rotation of the crystal. This advancement represents a crucial step toward refining the process of identifying the optimal channeling orientation. The algorithm has been tested for the fist time in operation with Pb ion beams at the record energy of 6.8 Z TeV demonstrating its reliability.</td></tr> <tr><td class="formatRecordLabel"> Copyright/License </td><td style="padding-left:5px;">publication: © 2024 authors (License: <a href="https://creativecommons.org/licenses/by/4.0/">CC BY 4.0</a>)</td></tr> </table> <br/>Corresponding record in: <a href="http://inspirehep.net/record/2834232">Inspire</a> <small> </small> <br/> <br/><br/><div align="right"><div style="padding-bottom:2px;padding-top:30px;"><span class="moreinfo" style="margin-right:10px;"> <a href="" class="moreinfo">Back to search</a> </span></div></div> <div class="bottom-left-folded"><div class="recordlastmodifiedbox" style="position:relative;margin-left:1px"> Record created 2024-10-08, last modified 2024-10-08</div></div> <div class="bottom-right-folded" style="text-align:right;padding-bottom:2px;"> <span class="moreinfo" style="margin-right:10px;"><a href="/search?ln=en&p=recid%3A2912720&rm=wrd" class="moreinfo">Similar records</a></span></div> </div> </div> </div> <br/> <br /> <div class="detailedrecordminipanel"> <div class="top-left"></div><div class="top-right"></div> <div class="inside"> <div id="detailedrecordminipanelfile" style="width:33%;float:left;text-align:center;margin-top:0"> <div><small class="detailedRecordActions">Fulltext:</small> <br /><a href="/record/2912720/files/Publication.pdf"><img style="border:none" src="/img/file-icon-text-34x48.gif" alt="Download fulltext" /><br />PDF</a><br /></div> </div> <div id="detailedrecordminipanelreview" style="width:30%;float:left;text-align:center"> </div> <div id="detailedrecordminipanelactions" style="width:36%;float:right;text-align:right;"> <ul class="detailedrecordactions"> <li><a href="/yourbaskets/add?ln=en&recid=2912720">Add to personal basket</a></li> <li>Export as <a style="text-decoration:underline;font-weight:normal" href="/record/2912720/export/hx?ln=en">BibTeX</a>, <a style="text-decoration:underline;font-weight:normal" href="/record/2912720/export/hm?ln=en">MARC</a>, <a style="text-decoration:underline;font-weight:normal" href="/record/2912720/export/xm?ln=en">MARCXML</a>, <a style="text-decoration:underline;font-weight:normal" href="/record/2912720/export/xd?ln=en">DC</a>, <a style="text-decoration:underline;font-weight:normal" href="/record/2912720/export/xe?ln=en">EndNote</a>,  <a style="text-decoration:underline;font-weight:normal" href="/record/2912720/export/xn?ln=en">NLM</a>, <a style="text-decoration:underline;font-weight:normal" href="/record/2912720/export/xw?ln=en">RefWorks</a> </li> </ul> <div style='padding-left: 13px;'>  <div id="bookmark"></div> <div id="bookmark_sciencewise"></div> <style type="text/css"> #bookmark_sciencewise, #bookmark {float: left;} #bookmark_sciencewise li {padding: 2px; width: 25px;} #bookmark_sciencewise ul, #bookmark ul {list-style-image: none;} </style> <script type="text/javascript" src="/js/jquery.bookmark.min.js"></script> <style type="text/css">@import "/css/jquery.bookmark.css";</style> <script type="text/javascript">// <![CDATA[ $.bookmark.addSite('sciencewise', 'ScienceWise.info', 'https://cds.cern.ch/img/sciencewise.png', 'en', 'bookmark', 'http://sciencewise.info/bookmarks/cds:2912720/add'); $('#bookmark_sciencewise').bookmark({sites: ['sciencewise']}); $('#bookmark').bookmark({ sites: ['facebook', 'twitter', 'linkedin', 'google_plusone'], icons: '/img/bookmarks.png', url: 'https://cds.cern.ch/record/2912720', addEmail: true, title: "Machine learning based crystal collimator alignment optimization", description: "In the CERN Large Hadron Collider (LHC), bent crystals are used to efficiently deflect beam halo particles toward secondary collimators used as absorbers. In this crystal collimation scheme, a crystal with a length of a few millimeters can produce a deflection equivalent to a magnetic field of hundreds of Tesla at LHC top energies, improving the cleaning performance of the machine. However, crystals must be in optimal alignment with respect to the circulating beam to maximize the efficiency of the channeling process. A newly developed machine learning model automatically classifies the channeling condition of crystals using beam loss monitor signals during slow rotation of the crystal. This advancement represents a crucial step toward refining the process of identifying the optimal channeling orientation. The algorithm has been tested for the fist time in operation with Pb ion beams at the record energy of 6.8 Z TeV demonstrating its reliability." }); // ]]> </script>  </div> </div> <div style="clear:both;margin-bottom: 0;"></div> </div> <div class="bottom-left"></div><div class="bottom-right"></div> </div> </div></div> <footer id="footer" class="pagefooter clearfix">  <div class="pagefooterstripeleft"> CERN Document Server :: <a class="footer" href="https://cds.cern.ch/?ln=en">Search</a> :: <a class="footer" href="https://cds.cern.ch/submit?ln=en">Submit</a> :: <a class="footer" href="https://cds.cern.ch/youraccount/display?ln=en">Personalize</a> :: <a class="footer" href="https://cds.cern.ch/help/?ln=en">Help</a> :: <a class="footer" href="https://cern.service-now.com/service-portal?id=privacy_policy&se=CDS-Service" target="_blank">Privacy Notice</a> <br /> Powered by <a class="footer" href="http://invenio-software.org/">Invenio</a> <br /> Maintained by <a class="footer" href="https://cern.service-now.com/service-portal?id=service_element&name=CDS-Service">CDS Service</a> - Need help? Contact <a href="https://cern.service-now.com/service-portal?id=service_element&name=CDS-Service">CDS Support</a>. <br /> </div> <div class="pagefooterstriperight"> <div class="cern-logo"> <a id="logo" href="http://cern.ch" title="CERN" rel="CERN" ><img src="https://cds.cern.ch/img/cern_theme/img/cern-logo-large.png" alt="CERN" /></a> </div> <div class="cern-languagebox"> This site is also available in the following languages:<br /><a href="/record/2912720?ln=bg" class="langinfo">Български</a>  <a href="/record/2912720?ln=ca" class="langinfo">Català</a>  <a href="/record/2912720?ln=de" class="langinfo">Deutsch</a>  <a href="/record/2912720?ln=el" class="langinfo">Ελληνικά</a>  <span class="langinfo">English</span>  <a href="/record/2912720?ln=es" class="langinfo">Español</a>  <a href="/record/2912720?ln=fr" class="langinfo">Français</a>  <a href="/record/2912720?ln=hr" class="langinfo">Hrvatski</a>  <a href="/record/2912720?ln=it" class="langinfo">Italiano</a>  <a href="/record/2912720?ln=ja" class="langinfo">日本語</a>  <a href="/record/2912720?ln=ka" class="langinfo">ქართული</a>  <a href="/record/2912720?ln=no" class="langinfo">Norsk/Bokmål</a>  <a href="/record/2912720?ln=pl" class="langinfo">Polski</a>  <a href="/record/2912720?ln=pt" class="langinfo">Português</a>  <a href="/record/2912720?ln=ru" class="langinfo">Русский</a>  <a href="/record/2912720?ln=sk" class="langinfo">Slovensky</a>  <a href="/record/2912720?ln=sv" class="langinfo">Svenska</a>  <a href="/record/2912720?ln=zh_CN" class="langinfo">中文(简)</a>  <a href="/record/2912720?ln=zh_TW" class="langinfo">中文(繁)</a> </div> </div>  </footer> <script type="text/javascript"> var SyndeticsBookCovers = (function() { var SMALL_SIZE = "sc.gif", MEDIUM_SIZE = "mc.gif", RAW_URL = "https://secure.syndetics.com/index.aspx?isbn=THEISBN/THESIZE&client=cernlibrary"; replaceCover = function(imgElement, isbns, hdFormat) { var img = new Image(), size = hdFormat ? MEDIUM_SIZE : SMALL_SIZE; var _isbns = isbns.sort(function(a, b) { // sort from shortest to longest ISBN (more modern) return a.length > b.length ? 1 : -1; }); function next() { var isbn = _isbns.pop(); if (isbn) { var url = RAW_URL.replace("THEISBN", isbn).replace("THESIZE", size); img.src = url; } } function done() { imgElement.src = img.src; } img.onload = function() { if (this.width > 1) { done(); } else { next(); } }; next(); }; return { replaceCover: replaceCover }; })(); $(document).ready(function() { // get book covers $("img.book-cover").each(function() { var $this = $(this), strIsbns = $this.data("isbns") || "", isbnsArray = String(strIsbns).split(","), hdFormat = $this.hasClass("hd"); SyndeticsBookCovers.replaceCover(this, isbnsArray, hdFormat); }); // WebNews tooltips $.ajax({ url: "/news/tooltips", success: function(data) { create_tooltips(data); }, dataType: "json", cache: false }); }); </script>  <script src="//cds.cern.ch/js/feedback.js"></script>   <script> var _paq = window._paq = window._paq || []; /* tracker methods like "setCustomDimension" should be called before "trackPageView" */ _paq.push(['trackPageView']); _paq.push(['enableLinkTracking']); (function() { var u="https://webanalytics.web.cern.ch/"; _paq.push(['setTrackerUrl', u+'matomo.php']); _paq.push(['setSiteId', '756']); var d=document, g=d.createElement('script'), s=d.getElementsByTagName('script')[0]; g.async=true; g.src=u+'matomo.js'; s.parentNode.insertBefore(g,s); })(); </script>  </body> </html>