<?xml version="1.0"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en">
	<id>https://biomolmd.org/mw/index.php?action=history&amp;feed=atom&amp;title=Ammm%3AMm_ewald</id>
	<title>Ammm:Mm ewald - Revision history</title>
	<link rel="self" type="application/atom+xml" href="https://biomolmd.org/mw/index.php?action=history&amp;feed=atom&amp;title=Ammm%3AMm_ewald"/>
	<link rel="alternate" type="text/html" href="https://biomolmd.org/mw/index.php?title=Ammm:Mm_ewald&amp;action=history"/>
	<updated>2026-11-16T11:15:55Z</updated>
	<subtitle>Revision history for this page on the wiki</subtitle>
	<generator>MediaWiki 1.43.1</generator>
	<entry>
		<id>https://biomolmd.org/mw/index.php?title=Ammm:Mm_ewald&amp;diff=1566&amp;oldid=prev</id>
		<title>Admin: /* Particle Mesh Ewald */</title>
		<link rel="alternate" type="text/html" href="https://biomolmd.org/mw/index.php?title=Ammm:Mm_ewald&amp;diff=1566&amp;oldid=prev"/>
		<updated>2022-05-10T01:50:03Z</updated>

		<summary type="html">&lt;p&gt;&lt;span class=&quot;autocomment&quot;&gt;Particle Mesh Ewald&lt;/span&gt;&lt;/p&gt;
&lt;table style=&quot;background-color: #fff; color: #202122;&quot; data-mw=&quot;interface&quot;&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;col class=&quot;diff-marker&quot; /&gt;
				&lt;col class=&quot;diff-content&quot; /&gt;
				&lt;tr class=&quot;diff-title&quot; lang=&quot;en&quot;&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;← Older revision&lt;/td&gt;
				&lt;td colspan=&quot;2&quot; style=&quot;background-color: #fff; color: #202122; text-align: center;&quot;&gt;Revision as of 01:50, 10 May 2022&lt;/td&gt;
				&lt;/tr&gt;&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l59&quot;&gt;Line 59:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 59:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&amp;lt;/span&amp;gt;&amp;lt;/span&amp;gt;&lt;/del&gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt; &lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:Ewald-real.png|580px]]&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:Ewald-real.png|580px]]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l128&quot;&gt;Line 128:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 128:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The &amp;lt;math&amp;gt;\alpha \,or\, \xi &amp;lt;/math&amp;gt; is the Ewald coefficient controls how much work is done in the real or reciprocal space. There should be a range of &amp;lt;math&amp;gt;\alpha&amp;lt;/math&amp;gt; gives the same total energy but with different efficiency.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The &amp;lt;math&amp;gt;\alpha \,or\, \xi &amp;lt;/math&amp;gt; is the Ewald coefficient controls how much work is done in the real or reciprocal space. There should be a range of &amp;lt;math&amp;gt;\alpha&amp;lt;/math&amp;gt; gives the same total energy but with different efficiency.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The CPU time required for a fully optimized Ewald summation scales with the number of particles as &amp;lt;math&amp;gt;O(N^{3/2})&amp;lt;/math&amp;gt;. the real-space can use a small cutofff (~9 Ang). The calculation of the Fourier part of the Ewald summation scales as N^2 , which makes the Ewald summation inefficient for large system.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;The CPU time required for a fully optimized Ewald summation scales with the number of particles as &amp;lt;math&amp;gt;O(N^{3/2})&amp;lt;/math&amp;gt;. the real-space can use a small cutofff (~9 Ang). The calculation of the Fourier part of the Ewald summation scales as N^2 , which makes the Ewald summation inefficient for large system.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt; &lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:Pme.pdf|thumb]]&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Particle mesh method takes advantage of Fast Fourier Transform to solve the reciprocal contribution. The potential and charge are pre-stored on a grid, which are used later in the interpolation to compute the actual energy and force.&amp;lt;ref name=&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&#039;&lt;/del&gt;pme&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&#039;&lt;/del&gt;&amp;gt;Darden, York and Pedersen, J. Chem. Phys., Vol. 98, No. 12, 15 June 1993&amp;lt;/ref&amp;gt; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:pme.pdf]]&lt;/del&gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Particle mesh method takes advantage of Fast Fourier Transform to solve the reciprocal contribution. The potential and charge are pre-stored on a grid, which are used later in the interpolation to compute the actual energy and force.&amp;lt;ref name=&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&quot;&lt;/ins&gt;pme&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&quot;&lt;/ins&gt;&amp;gt;Darden, York and Pedersen, J. Chem. Phys., Vol. 98, No. 12, 15 June 1993&amp;lt;/ref&amp;gt;  &lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:pme0.png|390px]]&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:pme0.png|390px]]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&amp;#039;&amp;#039;&amp;#039;Approximate the potential using interpolation (e.g. order p spline) of precomputed phi on the integer grid.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&amp;#039;&amp;#039;&amp;#039;Approximate the potential using interpolation (e.g. order p spline) of precomputed phi on the integer grid.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&#039;&#039;&#039;&lt;/del&gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt; &lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:pme.png|590px]]&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:pme.png|590px]]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt; &lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:Spme.pdf|thumb]]&lt;/ins&gt;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&#039;&#039;&#039;Early interpolation uses Lagrangian &amp;lt;ref name=&quot;pme&quot; /&amp;gt;  but later B spline, which is smoother (continuous derivative)&amp;lt;ref name+&quot;spme&quot;&amp;gt;Essman, JCP 1995, v103 p8577&amp;lt;/ref&amp;gt; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:spme.pdf]]&lt;/del&gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&#039;&#039;&#039;Early interpolation uses Lagrangian &amp;lt;ref name=&quot;pme&quot; /&amp;gt;  but later B spline, which is smoother (continuous derivative)&amp;lt;ref name+&quot;spme&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;&quot;=&quot;&lt;/ins&gt;&quot;&amp;gt;Essman, JCP 1995, v103 p8577&amp;lt;/ref&amp;gt; &#039;&#039;&#039;&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;&#039;&#039;&#039;&lt;/div&gt;&lt;/td&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-side-added&quot;&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:pme1.png|390px]]&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;[[File:pme1.png|390px]]&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot; id=&quot;mw-diff-left-l167&quot;&gt;Line 167:&lt;/td&gt;
&lt;td colspan=&quot;2&quot; class=&quot;diff-lineno&quot;&gt;Line 166:&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Comparison of Ewald and FMM:&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;Comparison of Ewald and FMM:&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;−&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;*Sagui and Darden &amp;lt;ref&amp;gt;Annu. Rev. Biophys. Biomol. Struct. 1999. 28:155–79&amp;lt;/ref&amp;gt; &lt;del style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:Sagui-annurev.biophys.28.1.pdf‎ ]]&lt;/del&gt;&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot; data-marker=&quot;+&quot;&gt;&lt;/td&gt;&lt;td style=&quot;color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;*&lt;ins style=&quot;font-weight: bold; text-decoration: none;&quot;&gt;[[File:Sagui-annurev.biophys.28.1.pdf|thumb]]&lt;/ins&gt;Sagui and Darden &amp;lt;ref&amp;gt;Annu. Rev. Biophys. Biomol. Struct. 1999. 28:155–79&amp;lt;/ref&amp;gt;  &lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;*For conceivable system size, PME is more efficient.&lt;/div&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;div&gt;*For conceivable system size, PME is more efficient.&lt;/div&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;td class=&quot;diff-marker&quot;&gt;&lt;/td&gt;&lt;td style=&quot;background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;&quot;&gt;&lt;br&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/table&gt;</summary>
		<author><name>Admin</name></author>
	</entry>
	<entry>
		<id>https://biomolmd.org/mw/index.php?title=Ammm:Mm_ewald&amp;diff=1565&amp;oldid=prev</id>
		<title>Admin: Created page with &quot;==Truncation of long-range interaction== For condense-phase simualtion, Periodic Boundary Conditon is used to avoid the boundary effect.  480px  It is expensive to compute all pairs of interactions.  Solution: cutoff?  Longer cutoff is (may be) more accurate but more expensive.  480px  ==Electrostatics== Electrostatic interactions die off very slowly. In crystal the interaction can be rather long-ranged. &#039;&#039;&#039;Cutoff is a bad idea.&#039;...&quot;</title>
		<link rel="alternate" type="text/html" href="https://biomolmd.org/mw/index.php?title=Ammm:Mm_ewald&amp;diff=1565&amp;oldid=prev"/>
		<updated>2022-05-10T01:48:07Z</updated>

		<summary type="html">&lt;p&gt;Created page with &amp;quot;==Truncation of long-range interaction== For condense-phase simualtion, Periodic Boundary Conditon is used to avoid the boundary effect.  &lt;a href=&quot;/mw/index.php/File:Pbc.png&quot; title=&quot;File:Pbc.png&quot;&gt;480px&lt;/a&gt;  It is expensive to compute all pairs of interactions.  Solution: cutoff?  Longer cutoff is (may be) more accurate but more expensive.  &lt;a href=&quot;/mw/index.php/File:Cutoff-pair.png&quot; title=&quot;File:Cutoff-pair.png&quot;&gt;480px&lt;/a&gt;  ==Electrostatics== Electrostatic interactions die off very slowly. In crystal the interaction can be rather long-ranged. &amp;#039;&amp;#039;&amp;#039;Cutoff is a bad idea.&amp;#039;...&amp;quot;&lt;/p&gt;
&lt;p&gt;&lt;b&gt;New page&lt;/b&gt;&lt;/p&gt;&lt;div&gt;==Truncation of long-range interaction==&lt;br /&gt;
For condense-phase simualtion, Periodic Boundary Conditon is used to avoid the boundary effect.&lt;br /&gt;
&lt;br /&gt;
[[File:pbc.png|480px]]&lt;br /&gt;
&lt;br /&gt;
It is expensive to compute all pairs of interactions.&lt;br /&gt;
&lt;br /&gt;
Solution: cutoff?&lt;br /&gt;
&lt;br /&gt;
Longer cutoff is (may be) more accurate but more expensive.&lt;br /&gt;
&lt;br /&gt;
[[File:Cutoff-pair.png|480px]]&lt;br /&gt;
&lt;br /&gt;
==Electrostatics==&lt;br /&gt;
Electrostatic interactions die off very slowly. In crystal the interaction can be rather long-ranged. &amp;#039;&amp;#039;&amp;#039;Cutoff is a bad idea.&amp;#039;&amp;#039;&amp;#039;&lt;br /&gt;
&lt;br /&gt;
[[File:ele-cut-quartz.png|480px]]&lt;br /&gt;
&lt;br /&gt;
We typically use Ewald with dipole correction (tin-foil boundary condition)&lt;br /&gt;
&lt;br /&gt;
== Ewald ==&lt;br /&gt;
&lt;br /&gt;
&amp;#039;&amp;#039;Detailed description can be found in many books and papers. Here we present the discussion found in book by Frenkel and Smit.&amp;lt;ref&amp;gt;Understanding Molecular Simulation From Algorithms to Applications, Frenkel and Smit, acdemic press 2002&amp;lt;/ref&amp;gt;&amp;#039;&amp;#039;&lt;br /&gt;
&lt;br /&gt;
Ewald is a rigorous method to treat long range electrostatic interaction.&amp;lt;ref&amp;gt; EE Ewald. Die Berechnung optischer und elektrostatischer Gitterpotentiale. Ann. Phys., 64:253-287, 1921.&amp;lt;/ref&amp;gt; It is as accurate as infinite cutoff, and the computaitonal cost is rughly similiar to ~ 10 Ang cutoff (when Particle mesh Ewals is used). &lt;br /&gt;
&lt;br /&gt;
&amp;lt;math&amp;gt;U_{ele}=1/2\sum_{i=1}^{N}q_{i}\phi \left ( r_{i} \right )&amp;lt;/math&amp;gt; &lt;br /&gt;
&lt;br /&gt;
&amp;lt;math&amp;gt;\phi\left ( r_{i} \right )=\sum_{j,n}^{&amp;#039;}\frac{q_{j}}{\left |  r_{ij}+nL\right |}&amp;lt;/math&amp;gt; &lt;br /&gt;
&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;Problem: &amp;lt;math&amp;gt;\frac{1}{r}&amp;lt;/math&amp;gt; dies off slowly&amp;#039;&amp;#039;&amp;#039;&lt;br /&gt;
&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;Solution: Rewrite the potential in two parts&amp;#039;&amp;#039;&amp;#039;&lt;br /&gt;
&amp;lt;math&amp;gt; \frac{1}{r}=\frac{erfc(r)}{r}+\frac{erf(r)}{r}&amp;lt;/math&amp;gt;&lt;br /&gt;
&lt;br /&gt;
&amp;lt;br&amp;gt; &amp;lt;math&amp;gt;erf(r) = \frac{2}{\sqrt{\pi}}\int_0^r e^{-t^2} dt.&amp;lt;/math&amp;gt; &lt;br /&gt;
&lt;br /&gt;
The &amp;#039;&amp;#039;&amp;#039;complementary error function&amp;#039;&amp;#039;&amp;#039;, denoted &amp;#039;&amp;#039;erfc&amp;#039;&amp;#039;, is defined in terms of the error function: &lt;br /&gt;
&lt;br /&gt;
&amp;lt;math&amp;gt;erfc(r) = 1-erf(r)= \frac{2}{\sqrt{\pi}} \int_r^{\infty} e^{-t^2}\,dt &amp;lt;/math&amp;gt; &lt;br /&gt;
&lt;br /&gt;
[[Image:Erfcr.png|480px]] &lt;br /&gt;
&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;Note: erfc(r)/r dies off much faster, i.e. interaction can be ignored within machine precision beyond a short cutoff.&amp;#039;&amp;#039;&amp;#039; &lt;br /&gt;
&lt;br /&gt;
&amp;lt;span style=&amp;quot;color: rgb(255, 102, 0); &amp;quot;&amp;gt;&amp;lt;span style=&amp;quot;font-size: larger; &amp;quot;&amp;gt;Physically, this effect can be achieved by adding (two equal and opposite) Gaussian charge distributions to screen the point charges:&amp;lt;/span&amp;gt;&amp;lt;/span&amp;gt;&lt;br /&gt;
&lt;br /&gt;
*&amp;#039;&amp;#039;&amp;#039;The Gaussian charge distribution leads to a potential associated with erf(r).&amp;#039;&amp;#039;&amp;#039;&lt;br /&gt;
&lt;br /&gt;
*&amp;#039;&amp;#039;&amp;#039;The screen potential (point charge plus the negative Gaussian) is associated with erfc(r).&amp;#039;&amp;#039;&amp;#039; &lt;br /&gt;
&lt;br /&gt;
[[Image:Ewald-illus.png|580px]]&lt;br /&gt;
&lt;br /&gt;
==Ewald real space contribution==&lt;br /&gt;
&lt;br /&gt;
&amp;lt;span style=&amp;quot;color: rgb(255, 102, 0); &amp;quot;&amp;gt;&amp;lt;span style=&amp;quot;font-size: larger; &amp;quot;&amp;gt;The screen potential, erfc(r) can now be computed using a short cutoff:&lt;br /&gt;
&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;we will show later why the potential of a Gaussian charge distribution is erf via solving Poisson equation.&amp;#039;&amp;#039;&amp;#039;&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&amp;lt;/span&amp;gt;&amp;lt;/span&amp;gt;&lt;br /&gt;
[[File:Ewald-real.png|580px]]&lt;br /&gt;
&lt;br /&gt;
==Ewald reciprocal space contribution==&lt;br /&gt;
&lt;br /&gt;
&amp;lt;span style=&amp;quot;color: rgb(255, 102, 0); &amp;quot;&amp;gt;&amp;lt;span style=&amp;quot;font-size: larger; &amp;quot;&amp;gt;What about the Gaussian charge distribution, the potential, erf(r)/r, diverges in the real-space now --- but dies off quickly in the reciprocal (Fourier) space.&amp;lt;/span&amp;gt;&amp;lt;/span&amp;gt;&lt;br /&gt;
&lt;br /&gt;
[[File:Ewald-grho.png|580 px]]&lt;br /&gt;
&lt;br /&gt;
[[File:Ewald-recip1.png|580 px]]&lt;br /&gt;
&lt;br /&gt;
[[File:Ewald-recip2.png|580px]]&lt;br /&gt;
&lt;br /&gt;
== Ewald self correction  ==&lt;br /&gt;
&lt;br /&gt;
Note the equation 12.1.17 include a single summation so there is an interaction between the point charge qi and the (negative) Gaussian charge distribution also located at &amp;#039;&amp;#039;i&amp;#039;&amp;#039; (r=0). This contribution can be removed. &lt;br /&gt;
&lt;br /&gt;
First we need to solve the potential due to a (positive) Gaussian. &lt;br /&gt;
&lt;br /&gt;
&amp;lt;math&amp;gt;\rho _{Gauss}(r)=q_{i}(\alpha/\pi)^{\frac{3}{2}}\exp(-\alpha r^{2})&amp;lt;/math&amp;gt; &lt;br /&gt;
&lt;br /&gt;
  &amp;lt;span style=&amp;quot;font-size: larger; &amp;quot;&amp;gt;The Dirac delta function is the limit of Gaussians &amp;lt;math&amp;gt;\delta_a(x) = \frac{1}{\sigma \sqrt{\pi}} \mathrm{e}^{-x^2/\sigma^2}&amp;lt;/math&amp;gt; as &amp;lt;math&amp;gt;\sigma &amp;lt;/math&amp;gt;, the width of Gaussian,&amp;lt;math&amp;gt; \to 0.&amp;lt;/math&amp;gt; &lt;br /&gt;
 or &amp;lt;math&amp;gt;\delta_a(r) = \frac{1}{\sigma^{3} {\pi}^{3/2}} \mathrm{e}^{-r^2/\sigma^2}&amp;lt;/math&amp;gt;&amp;lt;/span&amp;gt;&lt;br /&gt;
&lt;br /&gt;
[[Image:Poisson-gau.png|580px]] &lt;br /&gt;
&lt;br /&gt;
[[Image:Poisson-gau2.png|580px]] &lt;br /&gt;
&lt;br /&gt;
 &amp;lt;span style=&amp;quot;font-size: larger; &amp;quot;&amp;gt;Here we proved that the potential due to Gaussian charge distribution is erf. &lt;br /&gt;
 When alpha -&amp;gt; infinity, Gaussian charge distribution approaches a point charge and the erf approaches 1. &lt;br /&gt;
 Therefore, the potential φ (r) approaches Coulomb&amp;#039;s law (1/r). Also notice, to turn off &amp;quot;Ewald&amp;quot;, we need alpha=0, &lt;br /&gt;
 so we are not adding any (Gaussian} charges to screen the point charges&amp;lt;/span&amp;gt;&lt;br /&gt;
[[Image:Self_smit.png|580px]]&lt;br /&gt;
&lt;br /&gt;
==Ewald energy due to point charges==&lt;br /&gt;
[[File:Ewald-total.png|580px]]&lt;br /&gt;
&lt;br /&gt;
==boundary correction==&lt;br /&gt;
&lt;br /&gt;
[[File:Ewald-bc.png|580px]]&lt;br /&gt;
&lt;br /&gt;
This contribution to the potential energy corresponds to&lt;br /&gt;
the k = 0 term that we have neglected thus far. It is permissible to ignore this&lt;br /&gt;
term if the depolarizing field vanishes. This is the case if our periodic system&lt;br /&gt;
is embedded in a medium with infinite dielectric constant (a conductor), which is what we have assumed throughout.&lt;br /&gt;
For simulations of ionic systems, it is essential to use such &amp;quot;conducting&amp;quot;&lt;br /&gt;
boundary conditions; for polar systems, it is merely advantageous. For a&lt;br /&gt;
discussion of these subtle points, see &amp;lt;ref&amp;gt;M. Neumann. Dipole moment fluctuation formulas in computer simulations&lt;br /&gt;
of polar systems. Mol. Phys., 50:841-858, 1983.&amp;lt;/ref&amp;gt;.&lt;br /&gt;
&lt;br /&gt;
== Ewald for multipoles  ==&lt;br /&gt;
&lt;br /&gt;
Detailed derivation is in Smith article in CCP5 newsletter &amp;lt;ref&amp;gt; W. Smith CCP5, No 46. Point multipoles in the Ewald Summation revisited 1998&amp;lt;/ref&amp;gt; &lt;br /&gt;
&lt;br /&gt;
[[Image:Ewald-mpole-energy.png|680px]]&amp;lt;br&amp;gt;&lt;br /&gt;
[[Image:Ewald-mpole-g.png|680px]]&amp;lt;br&amp;gt;&lt;br /&gt;
&lt;br /&gt;
----&lt;br /&gt;
&lt;br /&gt;
[[Image:Ewald-bl.png|480px]]&amp;lt;br&amp;gt;&lt;br /&gt;
[[Image:Ewald-bl0.png|380px]]&amp;lt;br&amp;gt;&lt;br /&gt;
&lt;br /&gt;
----&lt;br /&gt;
&lt;br /&gt;
[[Image:Ewald-mpole-self.png|680px]]&amp;lt;br&amp;gt;&lt;br /&gt;
&lt;br /&gt;
==Particle Mesh Ewald==&lt;br /&gt;
The &amp;lt;math&amp;gt;\alpha \,or\, \xi &amp;lt;/math&amp;gt; is the Ewald coefficient controls how much work is done in the real or reciprocal space. There should be a range of &amp;lt;math&amp;gt;\alpha&amp;lt;/math&amp;gt; gives the same total energy but with different efficiency.&lt;br /&gt;
The CPU time required for a fully optimized Ewald summation scales with the number of particles as &amp;lt;math&amp;gt;O(N^{3/2})&amp;lt;/math&amp;gt;. the real-space can use a small cutofff (~9 Ang). The calculation of the Fourier part of the Ewald summation scales as N^2 , which makes the Ewald summation inefficient for large system.&lt;br /&gt;
&lt;br /&gt;
Particle mesh method takes advantage of Fast Fourier Transform to solve the reciprocal contribution. The potential and charge are pre-stored on a grid, which are used later in the interpolation to compute the actual energy and force.&amp;lt;ref name=&amp;#039;pme&amp;#039;&amp;gt;Darden, York and Pedersen, J. Chem. Phys., Vol. 98, No. 12, 15 June 1993&amp;lt;/ref&amp;gt; [[File:pme.pdf]]&lt;br /&gt;
&lt;br /&gt;
[[File:pme0.png|390px]]&lt;br /&gt;
&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;Approximate the potential using interpolation (e.g. order p spline) of precomputed phi on the integer grid.&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;&lt;br /&gt;
[[File:pme.png|590px]]&lt;br /&gt;
&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;Early interpolation uses Lagrangian &amp;lt;ref name=&amp;quot;pme&amp;quot; /&amp;gt;  but later B spline, which is smoother (continuous derivative)&amp;lt;ref name+&amp;quot;spme&amp;quot;&amp;gt;Essman, JCP 1995, v103 p8577&amp;lt;/ref&amp;gt; [[File:spme.pdf]]&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;&lt;br /&gt;
&lt;br /&gt;
[[File:pme1.png|390px]]&lt;br /&gt;
&lt;br /&gt;
[[File:pme2.png|390px]]&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
[[File:pme3.png|390px]]&lt;br /&gt;
&lt;br /&gt;
&amp;#039;&amp;#039;&amp;#039;&amp;lt;span style=&amp;quot;font-size: larger&amp;quot;&amp;gt;The PME scales as N log(N).&amp;lt;/span&amp;gt;&amp;#039;&amp;#039;&amp;#039;&lt;br /&gt;
&lt;br /&gt;
Typical use:&lt;br /&gt;
*Real space cutoff 9 Ang for point charge&lt;br /&gt;
**For point multipole, we can afford 6.8 Ang&lt;br /&gt;
**longer cutoff smaller Ewald coefficient (less work in the reciprocal space) will lead to slower speed&lt;br /&gt;
*Reciprocal grid size is &amp;lt;1 Ang (smaller more accurate but slower). X Ang cube box, the no. of grid should be at least 1.2X. &lt;br /&gt;
**For example, 30 A cube box should have 36 grid point.&lt;br /&gt;
*Reciprocal vector K should be big enough (depending on the Ewald coefficient)&lt;br /&gt;
**typical Ewald coefficient is 0.45 when 9 Ang real-space cutoff is used. Larger value needed for shorter real-space cutoff.&lt;br /&gt;
*Spline order: 5 minimum&lt;br /&gt;
&lt;br /&gt;
See http://www.ncbi.nlm.nih.gov/pubmed/15267263 for PME for multipole (the one implemented in TINKER).&lt;br /&gt;
&lt;br /&gt;
==Fast Multipole Method==&lt;br /&gt;
FMM is an alternative O(N) method. It is more efficient than PME for an extreme large system.&lt;br /&gt;
An algorithm that is of order O(N) is the fast multipole method. The multipole method is based on the idea that a group of particles at a large distance can be considered one big cluster, for which it is not necessary to calculate&lt;br /&gt;
all particle-particle interactions individually. By clustering the system into bigger and bigger groups, the interactions can be approximated. This approach of Appel &amp;lt;ref&amp;gt;A.W. Appel. An efficient program for many-body simulation. SIAM J. Sci. Stat. Comput., 6:85-103, 1985.&amp;lt;/ref&amp;gt; leads to an order O(N) algorithm &amp;lt;ref&amp;gt;K. Esselink. The order of Appel&amp;#039;s algorithm. Inf. Proc. Lett., 41:141-147, 1992.&amp;lt;/ref&amp;gt;. This algorithm was further refined by Barnes and Hut &amp;lt;ref&amp;gt;J. Barnes and P. Hut. A hierarchical O(N log N) force-calculation algorithm.Nature, 324:446-449, 1986.&amp;lt;/ref&amp;gt;. In the original algorithm of Appel, the clusters were approximated as a single charge. Greengard and Rokhlin &amp;lt;ref&amp;gt;L. Greengard and V. Rokhlin. A fast algorithm for particle simulations. J. Comp. Phys., 73:325-348, 1987.&amp;lt;/ref&amp;gt; developed an algorithm in which the charge distribution in a cluster is approximated by a multipole expansion. Schmidt and Lee extended this method to systems with periodic boundary conditions &amp;lt;ref&amp;gt;K.E. Schmidt and M.A. Lee. Implementing the fast multipole method in three dimensions. J. Stat. Phys., 63:1223-1235, 1991.&amp;lt;/ref&amp;gt;.&lt;br /&gt;
&lt;br /&gt;
Comparison of Ewald and FMM:&lt;br /&gt;
*Sagui and Darden &amp;lt;ref&amp;gt;Annu. Rev. Biophys. Biomol. Struct. 1999. 28:155–79&amp;lt;/ref&amp;gt; [[File:Sagui-annurev.biophys.28.1.pdf‎ ]]&lt;br /&gt;
*For conceivable system size, PME is more efficient.&lt;br /&gt;
&lt;br /&gt;
Goddard implementation for million atoms &amp;lt;ref&amp;gt;J. Chem. Phys., Vol. 97, No. 6, 15 September 1992&amp;lt;/ref&amp;gt;:&lt;br /&gt;
&lt;br /&gt;
[[File:fmm.png|400px]]&lt;br /&gt;
&lt;br /&gt;
==vdw interactions==&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
If we truncate the potential at a distance rc, the contribution of the tail of the potential u(r) can be estimated (in three dimensions) using&lt;br /&gt;
&lt;br /&gt;
&amp;lt;math&amp;gt;E_{tail}=\frac{N\rho }{2}\int_{rc}^{\infty }dru(r)4\pi r^{2}&amp;lt;/math&amp;gt;&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Vdw interactions dies off faster, especially the repulsion component &amp;lt;math&amp;gt;1/r^{12}&amp;lt;/math&amp;gt;. The pairwise interactions are cutoff at certain distance (e.g. 12 Ang). It is possible to correct the ignored long-range interactions for isotropic liquids and additive potentials (RDF approaches unit at the cutoff distance).&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Cutoff is less &amp;quot;dangers&amp;quot;; in addition, correction can be made for &amp;quot;homogeneous&amp;quot; system.&lt;br /&gt;
 &lt;br /&gt;
Recall the RDF (radial distribution function), or g(r), not only characterizes the microscopic structure but also relates to macroscopic ensemble averages of thermodynamics properties via the following relationship: &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
 &amp;lt;math&amp;gt;E=(3/2)N K_{b}T + 2\pi N\rho \int_{0}^{\infty }r^{2}v\left ( r \right)g\left(r \right) dr&amp;lt;/math&amp;gt; &lt;br /&gt;
&lt;br /&gt;
Beyond the cutoff, the g(r)=1, and the &amp;quot;long tail&amp;quot; correction &amp;lt;ref&amp;gt; Allen &amp;amp; Tildesley Computer Simulation of liquids &amp;lt;/ref&amp;gt;&lt;br /&gt;
&lt;br /&gt;
[[Image:Allen Tildesley LRC.png]]&lt;br /&gt;
&lt;br /&gt;
&amp;lt;references/&amp;gt;&lt;/div&gt;</summary>
		<author><name>Admin</name></author>
	</entry>
</feed>