{"id":4300,"date":"2014-04-05T07:36:35","date_gmt":"2014-04-05T06:36:35","guid":{"rendered":"http:\/\/blogs.scienceforums.net\/ajb\/?p=4300"},"modified":"2014-04-05T07:36:35","modified_gmt":"2014-04-05T06:36:35","slug":"what-is-a-connection-on-a-vector-bundle","status":"publish","type":"post","link":"http:\/\/blogs.scienceforums.net\/ajb\/2014\/04\/05\/what-is-a-connection-on-a-vector-bundle\/","title":{"rendered":"What is a connection on a vector bundle?"},"content":{"rendered":"<p>I will assume the reader has some basic knowledge of differential geometry including the standard notions of a connection.<\/p>\n<p>Connections on (or in) fibre bundles are confusing things when you first meet them.  They maybe introduced as a way of defining parallel transport of objects along curves, for the tangent bundle you may have been introduced to them as directional derivatives and so on. A  rather general notion is that put forward by Ehresmann as the compliment to the vertical bundle which we call the horizontal bundle. Note that Ehresmann&#8217;s defintion works for any fibre bundle and not just vector bundles. Even more abstractly we can define a connection on a fibred manifold as a section of its first jet bundle.<\/p>\n<p>Okay, so we all agree that are confusing things. However, for vector bundles we can describe them completely geometrically as certain maps between double vector bundle. For double vector bundles see for example [1]<\/p>\n<p><strong>Double vector bundles<\/strong><br \/>\nI won&#8217;t describe general double vector bundles here, all we really need is the tangent bundle of a vector bundle \\(TE \\), which we equip with natural local coordinates \\((x^{A}, y^{a}, \\dot{x}^{A}, \\dot{y}^{a}) \\). The coordinate transformations here are given by<\/p>\n<p>\\(x^{A&#8217;} = x^{A&#8217;}(x)\\)<br \/>\n\\(y^{a&#8217;} = y^{b}T_{b}^{\\:\\: a&#8217;}(x)\\)<br \/>\n\\(\\dot{x}^{A&#8217;} = \\dot{x}^{B}\\frac{\\partial x^{A&#8217;}}{\\partial x^{B}}(x)\\)<br \/>\n\\(\\dot{y}^{a&#8217;} = \\dot{y}^{b}T_{b}^{\\:\\: a&#8217;}(x) + y^{b}\\dot{x}^{B}\\frac{\\partial T_{b}^{\\:\\:a&#8217;}}{\\partial x^{B}}(x)\\)<\/p>\n<p>It is now convenient to view this double vector bundle as a bi-graded manifold by assigning a bi-weight as \\(w(x) = (0,0)\\), \\(w(y) = (1,0)\\), \\(w(\\dot{x}) = (0,1)\\) and \\(w(\\dot{y}) = (1,1)\\).<\/p>\n<p>Note that as a graded manifold the coordinate transformations must respect this bi-grading. A quick look at the transformation laws show that they do preserve this bi-grading. In fact, under some mild conditions, we can always view a double vector bundle as such a bi-graded manifold. This is far more convenient than the original definitions as a vector bundle in the category of vector bundles, where one has to check all the compatibility conditions.<\/p>\n<p>We have the following vector bundle structures<\/p>\n<p><img decoding=\"async\" src=\"http:\/\/blogs-new.scienceforums.net\/ajb\/wp-content\/uploads\/sites\/10\/2014\/04\/DVB-1-300x181.png\" alt=\"DVD\" \/><\/p>\n<p>The core of this double vector bundle is identified with \\(ker(T\\pi)\\) which is the vertical bundle \\(VE\\), which you can easily see is isomorphic to \\(E\\).<\/p>\n<p>As we have a bi-graded structure here we can also pass to a graded structure by considering the total weight. That is just sum the components of the bi-weight. In doing so we get what is know as a graded bundle of degree two [2]. In simpler language we have the series of fibrations<\/p>\n<p>\\(TE \\rightarrow E \\times_{M} TM \\rightarrow M \\),<\/p>\n<p>defined by first projecting out \\(\\dot{y}\\), which has weight 2 and the then projecting out \\(y\\) and \\(\\dot{x}\\) which have weight 1. Again you can see that this is all consistent with the coordinate transformations.<\/p>\n<p>Also note that \\(E \\times TM\\) also carries the structure of a double vector bundle simply by projection onto the two factors and then the obvious projections to \\(M\\). This can be viewed keeping the bi-weight as inherited by that on \\(TE \\).<\/p>\n<p><strong>Connection in a vector bundle<\/strong><br \/>\nNow let me define in a non-standard way a linear connection.<\/p>\n<p><strong>Def<\/strong> A (linear) connection on a vector bundle \\(E\\) is a morphisms of double vector bundles  (bi-graded manifolds)<\/p>\n<p>\\(\\Gamma: E \\times_{M}TM \\rightarrow TE\\),<\/p>\n<p>that acts as the identity on the vector bundles \\(E\\) and \\(TM\\).<\/p>\n<p>So, does this reproduce the more standard definitions? Let us look at this in local coordinates;<\/p>\n<p>\\((x^{A}, y^{a}, \\dot{x}^{A}, \\dot{y}^{a})\\circ \\Gamma  = (x^{A}, y^{a}, \\dot{x}^{A}, \\dot{x}^{A} y^{b} (\\Gamma_{b}^{\\:\\: a})_{A}(x))\\).<\/p>\n<p>The question now is how do the components of the morphisms \\((\\Gamma_{b}^{\\:\\: a})_{A}\\) transform under changes of local coordinates?<\/p>\n<p>It is not  hard to see that we have<\/p>\n<p>\\((\\Gamma_{b&#8217;}^{\\:\\: a&#8217;})_{A&#8217;} = \\frac{\\partial x^{B}}{\\partial x^{A&#8217;}}\\left(T_{b&#8217;}^{\\:\\: c} (\\Gamma_{c}^{\\:\\: d})_{B}T_{d}^{\\:\\: a&#8217;} + T_{b&#8217;}^{\\:\\: c} \\frac{\\partial T_{c}^{\\:\\: a&#8217;}}{\\partial x^{B}} \\right)\\),<\/p>\n<p>thus the local data of a linear connection as I defined it coincides exactly with the local data of the connection coefficients of connection on a vector  as defined in any textbook on differential geometry.<\/p>\n<p><strong>Acknowledgements <\/strong><br \/>\nI won&#8217;t claim originality here, this understanding was largely inspired by a lecture given by Prof. Urbanski.<\/p>\n<p><strong>References<\/strong><br \/>\n[1] Kirill C. H. Mackenzie, &#8220;General Theory of Lie Groupoids and Lie Algebroids&#8221;, Cambridge University Press, 2005.<\/p>\n<p>[2] J. Grabowski and M. Rotkiewicz, Graded bundles and homogeneity structures, <em>J. Geom. Phys.<\/em> <strong>62<\/strong> (2012), no. 1, 21-36.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>I will assume the reader has some basic knowledge of differential geometry including the standard notions of a connection. Connections on (or in) fibre bundles are confusing things when you first meet them. They maybe introduced as a way of defining parallel transport of objects along curves, for the tangent bundle you may have been &hellip; <a href=\"http:\/\/blogs.scienceforums.net\/ajb\/2014\/04\/05\/what-is-a-connection-on-a-vector-bundle\/\" class=\"more-link\">Continue reading <span class=\"screen-reader-text\">What is a connection on a vector bundle?<\/span> <span class=\"meta-nav\">&rarr;<\/span><\/a><\/p>\n","protected":false},"author":7,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[],"class_list":["post-4300","post","type-post","status-publish","format-standard","hentry","category-general-mathematics"],"_links":{"self":[{"href":"http:\/\/blogs.scienceforums.net\/ajb\/wp-json\/wp\/v2\/posts\/4300","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/blogs.scienceforums.net\/ajb\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/blogs.scienceforums.net\/ajb\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/blogs.scienceforums.net\/ajb\/wp-json\/wp\/v2\/users\/7"}],"replies":[{"embeddable":true,"href":"http:\/\/blogs.scienceforums.net\/ajb\/wp-json\/wp\/v2\/comments?post=4300"}],"version-history":[{"count":0,"href":"http:\/\/blogs.scienceforums.net\/ajb\/wp-json\/wp\/v2\/posts\/4300\/revisions"}],"wp:attachment":[{"href":"http:\/\/blogs.scienceforums.net\/ajb\/wp-json\/wp\/v2\/media?parent=4300"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/blogs.scienceforums.net\/ajb\/wp-json\/wp\/v2\/categories?post=4300"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/blogs.scienceforums.net\/ajb\/wp-json\/wp\/v2\/tags?post=4300"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}