<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Γ(z) on Medium]]></title>
        <description><![CDATA[Stories by Γ(z) on Medium]]></description>
        <link>https://medium.com/@anacabreu?source=rss-be6f053229d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*tRpmT5ntXQjX2EO94gaRZw@2x.jpeg</url>
            <title>Stories by Γ(z) on Medium</title>
            <link>https://medium.com/@anacabreu?source=rss-be6f053229d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 09:28:25 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@anacabreu/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Regular Sturm-Liouville problems IV- Green’s function, existence, completeness and the spectral…]]></title>
            <link>https://anacabreu.medium.com/regular-sturm-liouville-problems-iv-greens-function-existence-completeness-and-the-spectral-3e96d8c22802?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/3e96d8c22802</guid>
            <category><![CDATA[sturm-liouville]]></category>
            <category><![CDATA[differential-equations]]></category>
            <category><![CDATA[functional-analysis]]></category>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Mon, 10 Aug 2026 00:29:26 GMT</pubDate>
            <atom:updated>2026-08-10T00:57:51.494Z</atom:updated>
            <content:encoded><![CDATA[<h2>Regular Sturm-Liouville problems IV- Green’s function, existence, completeness and the spectral theorem</h2><p>This final article of the series on regular Sturm-Liouville problems brings together everything developed so far into a complete spectral theory for a regular Sturm-Liouville problems brings together on a closed interval .</p><p>The central construction is the Green’s function. Once it exists, the differential eigenvalue problem (for reference, see <a href="https://anacabreu.medium.com/regular-sturm-liouville-problems-iii-self-adjoint-operators-and-the-formal-adjoint-of-second-order-d46a5530dcbb">the third part of this series</a>) is rewritten as a compact self-adjoint integral equation.</p><p>The spectral theorem for such operators then results in: (i) the existence of infinitely many eigenvalues, (ii) their tendency to infinity, and (iii) the completeness of the eigenfunctions in <em>ℒ²</em> (the fact that they span the space). With an independent quadratic-form argument, it is easy to show that the eigenvalues cannot escape to minus infinity, i.e.: they are bounded from below though unbounded from above.</p><p>Hence, we end up with a discrete spectrum and a complete orthogonal system of eigenfunctions.</p><p>From the last piece, recall that the Sturm-Liouville operator written in self-adjoint form, looks like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/986/1*Mbwp2pBrWq_X4XdXTjSPBA@2x.jpeg" /></figure><p>We are interested in the following eigenvalue problem (Sturm-Liouville eigenvalue problem):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/931/1*Ck-TTN8s1hjW9PlfkE8ZHA@2x.jpeg" /></figure><p>To study this eigenvalue problem, we first consider<em> Lu = f</em>. We want to find a function <em>G(x,ξ)</em> such that, for each point ξ, it satisfies the homogeneous equation <em>L[G] = 0</em> everywhere except for at <em>x = ξ </em>diagonal and whose first derivative jumps by <em>1/p(ξ) </em>when <em>x</em> crosses <em>ξ</em>.</p><p>Once such a function is known, the unique solution is recovered by the integral fromula for <em>f = -λu</em> such that the original eigenvalue equation is converted into a homogeneous integral.</p><h3>Green’s function</h3><p>A second-order linear differential equation is a local relation connecting the value of an unknown function at a point to the values of its first and second derivatives at the same point . On the other hand, boundary conditions are the global constraints.</p><p>Green’s function is like a translator that converts the differential problem, boundary conditions included, into a single integral equation.</p><p>Intuitively, think of <em>L</em> as describing a linear system, <em>Lu = f </em>asks for the response of the system to an external distributed force f. <em>G</em> answers what is the response of the system when driven by a unit point source concentrated at a single locus <em>ξ</em>, i.e.: the response observed at <em>x</em> is <em>G(x, ξ)</em>.</p><p>Since the equation is linear, the response to f is the continuous superposition of the response to all point sources that constitute <em>f:</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/987/1*Bj5Tm3HoZBctghcUvenCMg@2x.jpeg" /></figure><p>And <em>Lu = λu</em>. becomes <em>u = λKu</em>, where <em>K</em> is the integral operator with <em>G</em> as kernel. Integral operators with continuous kernels on compact intervals are compact. Thus, since <em>L</em> is self-adjoint, <em>K</em> is also self-adjoint.</p><p>Let’s construct Green’s function:</p><p>Let <em>G: [a,b]x[a,b] → ℝ</em>. be continuous. <em>G</em> is a Green’s function for <em>L</em>, with the given boundary conditions if, for all fixed <em>ξ</em> in<em> (a,b)</em>:</p><p>(i) <em>Lₓ[G(., ξ)] = 0</em> on <em>(a, ξ)</em> and <em>(ξ,b).</em></p><p>(ii) <em>G(.,ξ)</em> satisfies the left boundary condition at <em>x = a</em> and the right one at <em>x = b</em>.</p><p>(iii) <em>G</em> is continuous across the diagonal <em>x = ξ</em>.</p><p>(iv)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/984/1*72QEiAm0hgzK5sR-dgearw@2x.jpeg" /></figure><p><strong>Theorem 1.- Existence and Uniqueness of <em>G</em></strong></p><p>Assume <em>λ = 0</em> is not an eigenvalue of <em>L</em>. Then <em>G</em> exists an it is unique. In addition, it is symmetric: <em>G(x,ξ) ‎ = G(ξ,x)</em>.</p><p><strong>Existence proof</strong></p><p>Since zero is not an eigenvalue, <em>Lu = 0</em> admits one dimensional space of solutions satisfying the boundary conditions. Choose a non-trivial solution from that space: <em>uₗ</em> and <em>uᵣ</em>. of <em>Lu = 0</em>. Both linearly independent.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/972/1*i4fhrbhn_3_0wK3tU3jvlw@2x.jpeg" /></figure><p>Verifying properties:</p><p>(i) Off the diagonal<em> x = ξ,</em> <em>G</em> is a scalar multiple of either <em>uₗ</em> or <em>uᵣ</em>, satisfying the homogeneous equation.</p><p>(ii) At <em>x = a</em>, it is a multiple of <em>uₗ</em> satisfying the left boundary condition, similarly, at <em>x = b</em>, it is a multiple of <em>uᵣ</em>, doing the same with the right one.</p><p>(iii) Continuity on <em>[a,b]x[a,b]</em> is immediate from two expressions agreeing when <em>x = ξ</em>.</p><p>(iv) Differentiating,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/987/1*3wt0rz_bHcWmPQ3dt6ndnA@2x.jpeg" /></figure><p>Therefore, <em>G</em> is a Green’s function. Q.E.D</p><p><strong>Uniqueness proof</strong></p><p>Let <em>G₁</em> and <em>G₂</em>. be two Green’s functions such that <em>H = G₁ — G₂</em> satisfies the homogeneous equation off the diagonal, boundary conditions, is continuous and the jump in the first derivative is zero. Therefore, <em>H(., ξ)</em> is a classical solution of the homogeneous boundary value problem for all ξ. But we have assumed that zero is not an eigenvalue, therefore <em>H ≡ 0</em>.</p><p>Hence, <em>G</em> exists and it is unique. Q.E.D</p><p>It is your turn to prove symmetry <em>G(x,ξ) = G(ξ,x).</em></p><h3>Boundedness from below</h3><p><strong>Theorem 2.-</strong> There exists <em>l</em> real which depends on <em>p</em> and <em>q</em>, such that every eigenvalue <em>λ</em> of the Sturm-Liouville satisfies that <em>λ ≥ l.</em></p><p><strong>Proof</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/916/1*2iSw_Mf1-K9WIMGyrY9t0Q@2x.jpeg" /></figure><p>Integration by parts converts <em>&lt;Lu,u&gt;</em> into an expression that contains the square of the derivative weighted by <em>p &gt; 0</em>. The only term that can be negative is <em>q</em>, but <em>q</em> is continuous on the compact interval, thus, it is bounded from below and the whole quadratic is bounded from below and so the eigenvalues.</p><h3>Associated integral operator</h3><p>Here, we want to translate the original differential problem into the eigenvalue problem for a compact, self-adjoint integral operator.</p><p>Assume serio is not an eigenvalue, let <em>G</em> be the Green’s function of the <em>L</em> operator. By definition, <em>G</em> conversación every non-homogeneous boundary value problem <em>Lu = f</em> into the integral:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/854/1*O9FmmVGLhS421kxGxQ5A8Q@2x.jpeg" /></figure><p>Which means that Green’s function is an integral inverse of <em>L</em>. Now, for the eigenvalue equation <em>Lu ‎ = λu</em>, <em>f = λu</em> and</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/856/1*_NAClxyDbyQYo5gJd5mo5A@2x.jpeg" /></figure><p>Conversely, if a continuous function <em>u</em> satisfies that integral equation, then:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ec9iULm_nF-Rcxy7DWQP8Q@2x.jpeg" /></figure><p>And we recover <em>Lu = λu</em>.</p><p>Besides, the integral satisfies boundary conditions since <em>G</em> does it. Thus, the equivalence is true.</p><p>In operator language let <em>K</em> on <em>ℒ²[a,b</em>]:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/995/1*nNLjNMHjjwr-HnRLx2hQIw@2x.jpeg" /></figure><p>First, <em>K</em> is compact. The kernel of <em>G</em> is continuous on <em>[a,b]x[a,b]</em> which is also compact. Thus, the image under <em>K</em> of the <em>ℒ²[a,b]</em> unit ball is made of uniformly bounded and equicontinuous function. By Arzelà-Ascoli, this image is precompact in <em>ℒ²[a,b]</em>.</p><p>Second, <em>K</em> is self-adjoint. From Green’s function symmetry, the kernel of <em>K</em> is hermitian, thus, for all <em>u</em>, <em>v</em> in <em>ℒ²[a,b]: &lt;Ku,v&gt; = &lt;u,Kv&gt;.</em></p><p>Compact self-adjoint operators possess an extremely rigid and complete spectral structure. All of conclusions about eigenvalues and eigenfunctions of <em>L</em> flow from that structure, i.e.: the spectrum of <em>K</em> consists of discrete eigenvalues of finite multiplicity.</p><h3>The spectral theorem</h3><p>The spectral theorem is the engine that turns compactness and self-adjointness into a complete orthonormal basis of eigenfunctions.</p><p><strong>Theorem 3.- The spectral theorem for compact self-adjoint operators.</strong></p><p>Let <em>H</em> be a Hilbert space and <em>K: H → H</em> is compact and self-adjoint, then:</p><p>(i) All the eigenvalues of <em>K</em> are real.</p><p>(ii) Eigenvectors for distinct eigenvalues are orthogonal.</p><p>(iii) The set of eigenvalues is, at most, countable and can accumulate only a the origin.</p><p>(iv) every non-zero eigenvalue has finite multiplicity.</p><p>(v) There exists an orthonormal basis of <em>H</em> consisting, absolutely, of eigenvectors of <em>K</em>.</p><p>(vi) Every vector f in <em>H</em> admits an expansion:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1007/1*cXkc7vtfM5Y9FUHekC-qTg@2x.jpeg" /></figure><p>Which converges in the norm of <em>H</em>, where the sum runs over an enumeration of the eigenvalues according to multiplicity. If there are infinitely many eigenvalues, one has <em>μₙ → 0</em>.</p><p>The proof of this theorem is pretty standard, you can found it in Kreyszig’s chapter 8 (8.3 spectral properties of compact linear operators on normed spaes), but here is a brief summary:</p><p>Let <em>K</em> be a compact self-adjoint operator on <em>H</em>. If <em>K = 0</em>, then every vector is an eigenvector and any orthonormal basis work. Otherwise, if. <em>K ≠ 0</em>, compactness and self-adjointness guarantee that <em>K</em> has, at least, one eigenvalue <em>μ</em> such that the absolute value is equal to the operator norm <em>||K||</em>. The corresponding eigenspace is finite dimensional. Now, the orthogonal complement of that eigenspace is invariant under <em>K</em>. Restrict <em>K</em> to the complement and repeat the argument: extract another eigenvalue of maximal removing modulus and so on.</p><p>Continuing this process produces a sequence of pairwise orthogonal finite-dimensional eigenspaces. Their closed span is the whole space. Choosing an orthonormal basis inside each eigenspace results in an orthonormal basis inside each eigenspace results in an orthonormal basis of <em>H</em> consisting of eigenvectors. The eigenvalues are real, can accumulate only at zero and every non-zero eigenvalue has finite multiplicity.</p><p>In short, compactness supplies the existence of eigenvalues and prevents them from accumulating (except at zero) and self-adjointness supplies orthogonality, forcing the eigenvectors to span the whole space.</p><h3>Existence and completeness for the Sturm-Liouville Problem</h3><p><strong>Theorem 4.- Existence and completeness</strong></p><p>The regular Sturm-Liouville problem possesses a countably infinite sequence of of real eigenvalues: <em>λ₁ ≤ λ₂ ≤ λ₃</em>… satisfying <em>λₙ → ∞</em> and a corresponding sequence eigenfunctions {uₙ} that forms an orthonormal basis of <em>ℒ²[a,b]</em>. In particular the system is complete, i.e.: every f in <em>ℒ²[a,b]</em> admits an expansion:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/815/1*ucLRWGG2LL0bjCHeJfXFmw@2x.jpeg" /></figure><p>Convergent in the <em>ℒ²</em> norm</p><p><strong>Proof</strong></p><p>With the spectral theorem on the compact self-adjoint <em>K</em> we obtain an orthonormal basis of eigenvectors of <em>K</em> with eigenvalues <em>μₙ → 0</em>. Non-zero eigenvalues of K are related to those of L by. <em>λ = 1/ μ</em>.</p><p>By theorem 2, the eigenvalues is bounded from below, thus, the only possible accumulation point of <em>λₙ</em> is infinity. Particularly, there must be infinitely many eigenvalues. Hence <em>λₙ → ∞.</em></p><p>The eigenfunctions of <em>L</em> are, exactly, the eigenvectors of <em>K</em> which by the spectral theorem, they form an orthonormal basis of <em>ℒ²[a,b]</em>, i.e.: this basis expands every square-integrable function (completeness).</p><h3>Conclusion</h3><p>Now we know that even beginning with a modest question, the answer, developed across this series is far richer and deeper than a mere existence statement. We started with a second-order linear differential expression and a pair of boundary conditions, then we asked if the associated eigenvalue problem has solution.</p><p>We then constructed a function, Green’s function, who converts the differential operator into an integral operator. Since this integral operator is compact and self-adjoint, the spectral theorem gives us a countable infinite sequence of real eigenvalues that can accumulate only at infinity, together with a corresponding or thorn normal basis of eeigenfunctions. Then a on a the eigenvalues arThe bounded from below, thus completing the picture, i.e.: a discrete spectrum bounded from below and unbounded from above, with a complete orthogonal system in <em>ℒ²</em>.</p><p>Intuitively, the Green’s function is the system’s response to , point source. Once every possible forcing term can be reconstructed by superposing these responses , the original differential eigenvalue problem shows to be equivalent to a compact integral eigenvalue problem.</p><p>Compactness (the fact that a continuous kernel on a finite interval cannot see infinitely many directions at once) and self-adjointness (symmetry or energy conservation), together, force the spectrum to be discrete and the eigenfunctions to span the entire space.</p><p>s one turns to linear PDEs with self-adjoint partial operators. The method of variables separation reduces many of the well-known equations of mathematical physics to, precisely, Sturm-Liouville problems. Just three examples are: heat, wave and Laplace’s equations.</p><p>For the heat equation, on a finite interval (or a rectangle, cylinder or disk in appropriate coordinate), the temperature <em>u(x,t)</em> is expanded in the eigenfunctions <em>{uₙ}</em> of the spatial operator:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/973/1*P_SbxeHJDsUJfJk4QEmqSg@2x.jpeg" /></figure><p>Completeness guarantees that arbitrary initial data can be represented, the fact that <em>λₙ → ∞</em> guarantees that high modes decay rapidly, which is the mathematical expression of smoothing.</p><p>In the case of the wave equation, the same eigenfunctions result in normal modes of vibration and the general solution is the superposition:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/899/1*Q99_8WflZmrzKzBmQs94vQ@2x.jpeg" /></figure><p>Here, orthogonality makes coefficients easy to compute from initial displacement and velocity.</p><p>Lastly, Laplace’s: in a rectangle, a disk, or an annular region, separation of variables again produces Sturm-Liouville problems in one or more of the coordinates. The eigenfunction expansion converts the PDE into a sequence of elementary ordinary differential equations (or algebraic equations) for the coefficients. The same pattern appears in polar, cylindrical, and spherical coordinates, where the radial or angular equations are singular Sturm-Liouville problems whose regular counterparts we have studied.</p><p>In all of these, the results of the present piece become concrete tools: the eigenvalues determine the decay rates or the frequencies, the eigenfunctions form the natural basis in which the solution is written, and completeness guarantees that the series can represent the prescribed boundary or initial data. The Green’s function itself reappears as the spatial kernel of the solution operators. Thus the spectral theory of a single ordinary differential operator on an interval is the backbone of the classical method of eigenfunction expansions for partial differential equations.</p><h4>Recommended books</h4><p>Al-Gwaiz, M. A. (2008). Sturm — Liouville theory and its applications. Springer.</p><p>Kreyszig, E. (1978). Introductory functional analysis with applications. John Wiley &amp; Sons.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3e96d8c22802" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Regular Sturm-Liouville problems III- Self-Adjoint Operators and the Formal Adjoint of Second-Order…]]></title>
            <link>https://anacabreu.medium.com/regular-sturm-liouville-problems-iii-self-adjoint-operators-and-the-formal-adjoint-of-second-order-d46a5530dcbb?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/d46a5530dcbb</guid>
            <category><![CDATA[partial-differential-equa]]></category>
            <category><![CDATA[linear-operators]]></category>
            <category><![CDATA[sturm-liouville]]></category>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Sun, 02 Aug 2026 06:39:51 GMT</pubDate>
            <atom:updated>2026-08-02T07:33:29.060Z</atom:updated>
            <content:encoded><![CDATA[<h2>Regular Sturm-Liouville problems III- Self-Adjoint Operators and the Formal Adjoint of Second-Order Linear Differential Operators</h2><p>From the <a href="https://anacabreu.medium.com/regular-sturm-liouville-problems-ii-second-order-linear-ordinary-differential-equations-and-zeros-daf9487d80ca">second article</a> of this series, recall that a homogeneous second-order linear ordinary differential equation may be written as follows: <em>p(x)y’’ + q(x)y’+ r(x)y = 0, </em>or, more compactly: <em>L[y] = 0</em>, where the linear differential operator <em>L = p(x) d^2/x^2 + q(x)d/x + r(x)</em> acts on a suitable space of twice-differentiable functions.</p><h3><strong>Linear operators</strong></h3><p>A linear operator is a function between two vector spaces, particularly normed spaces. Let <em>T</em> be an operator, we say that <em>T:D(T)⊂ X→ Y</em> between vector spaces (over the same field) is called linear if:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/947/1*QkV-ZurL1fXgHuuUX6T_3g@2x.jpeg" /></figure><p>In other words, <em>T</em> is a homomorphism of vector spaces. When <em>T</em> is bijective (a monomorphism and an epimorphism) it is an isomorphism, and the domain and the image are algebraically indistinguishable.</p><p>Differentiation, integration and matrix multiplication are the most familiar examples of linear operators:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/922/1*O2WJFb1pjYUaqBk6X9HL-w@2x.jpeg" /></figure><p>Both preserve linear dependence and linear independence. If <em>T</em> is injective, then an inverse operator <em>T⁻¹:R(T)→ D(T)</em> exists.</p><h3><strong>The adjoint of an operator</strong></h3><p>Let <em>X</em> be an inner-product space. An operator <em>\widehat{T}</em> is called an adjoint of <em>T</em> if:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/905/1*-Y5CPL0q5UREWHk5ke1C8w@2x.jpeg" /></figure><p>Thus, <em>\widehat{T}</em> is the unique partner that mirrors the action of <em>T</em> through the inner product, i.e.: whatever <em>T</em> does to the first argument is exactly compensated, in the inner product sense, by what <em>\widehat{T}</em> does to the second argument.</p><p>Let us work on two examples. The first is from Axler’s Algebra Done Right (p. 204), the second is a pretty common one.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/973/1*cLd82YdlChXY0QubPMsLMQ@2x.jpeg" /></figure><p>Now, consider the operator <em>T = d/dx</em> acting on smooth functions on <em>[0,1]</em> that vanish at the left endpoint (<em>f(0) = 0</em>). Integration by parts gives:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/935/1*gAH6HreUp6l3CPf-eKh3AQ@2x.jpeg" /></figure><p>The boundary term disappears whenever <em>g(1) = 0</em>. Consequently the adjoint is<em> \widehat{T} = -d/dx</em> with domain consisting of functions that vanish at the right endpoint.</p><h3><strong>Matrix representation</strong></h3><p>When <em>X</em> is finite-dimensional and <em>{eᵢ}</em> is an orthonormal basis, every linear operator <em>T</em> is represented by the matrix:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/912/1*KakPqQ_78qVT64Dti7d1JQ@2x.jpeg" /></figure><p>The same definition applied to the adjoint immediately yields:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/957/1*8YZZmAqEhrOz3n4inY1Qyg@2x.jpeg" /></figure><p>So,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/910/1*uJpkm29OFmErs9R2ZWi5iQ@2x.jpeg" /></figure><p>This matrix is the conjugate transpose. The abstract adjoint is therefore independent of the choice of orthonormal basis.</p><h3><strong>Self-adjointness</strong></h3><p>A bounded linear operator <em>T:X → X</em> on a Hilbert space is called self-adjoint (or Hermitian) if<em> T = \widehat{T}</em>. In matrix language, this means <em>M = M*</em> (Hermitian matrix) in the complex case, or <em>M = Mᵀ</em> (symmetric matrix) in the real case.</p><p>Intuitively, think of the inner product as a way of measuring angles and lengths. The bilinear form defined by <em>T</em> is completely symmetric:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/909/1*hdHUxVL3eY9pW0f1SnPixA@2x.jpeg" /></figure><p>The work done by <em>T</em> on <em>x</em> when tested against <em>y</em> is exactly the same as the work done by <em>T</em> on <em>y</em> when tested against <em>x</em>. There is no preferred direction and no hidden asymmetry; the operator is indistinguishable from its own mirror image under the inner product.</p><h3><strong>Consequences of self-adjointness</strong></h3><p>Let <em>T</em> be self-adjoint and suppose <em>Tu = λu</em> with u≠0. Then:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/935/1*N5NuM4BgPdC_CW_euPxf3g@2x.jpeg" /></figure><p>If <em>Tu = λu</em> and <em>Tv = μv</em> with <em>λ ≠ μ</em>, the same argument yields:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/961/1*JfWaY16gnTGAnXsmL6ZJcg@2x.jpeg" /></figure><p>Thus eigenvectors belonging to distinct eigenvalues are orthogonal.</p><p>To sum up, in finite dimensions, the set of all eigenvectors forms an orthonormal basis of the whole space and all eigenvalues are real numbers. In the infinite-dimensional setting of Sturm - Liouville theory the same geometric picture survives once a suitable domain has been chosen so that the operator becomes truly self-adjoint.</p><h3><strong>The formal adjoint of a second-order differential operator</strong></h3><p>In the theory of Sturm-Liouville operators, or generally, in the spectral theory of ordinary differential operators, the notion of self-adjointness becomes relevant. In this section, I will introduce the formal adjoint of <em>L</em> directly from the inner product definition on a Hilbert space.</p><p>We now work in the Hilbert space <em>ℒ ²[a,b]</em> equipped with the inner product:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/937/1*l0dJuVilqcQzIsaN7KEMPg@2x.jpeg" /></figure><p>Consider, again, the differential expression: <em>L = p(x) d^2/x^2 + q(x)d/x + r(x)</em> where <em>p, q, r</em> are sufficiently smooth, i.e.: p∈ C², q∈ C¹, r∈ C⁰ and real-valued. The operator <em>L</em> is unbounded, so for the adjoint to be defined we should use more than just algebraic formulation, its domain must therefore be specified carefully. First we look for a differential expression <em>L* </em>(the formal adjoint) such that:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/927/1*IrEaCJWplq70bVbkhKczqw@2x.jpeg" /></figure><p>Boundary terms appear since each integration by parts produces an evaluation at the end points. If the functions under consideration belong to a domain on which those boundary terms vanish, then:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/918/1*0K1Nvb5qtqmkTMXR58-y8Q@2x.jpeg" /></figure><p>such that <em>L</em> is Hermitian with respect to that specific domain. Moreover, if the domain of <em>L</em> matches with the domain of its true Hilbert-space adjoint, then <em>L</em> is self-adjoint. Hence, the formal adjoint is the differential expression that would make the operator symmetric if boundary terms can disappear by a suitable choice of domain.</p><p>The boundary terms tell us which precise boundary conditions are candidates for producing a self-adjoint. Let us find the formal adjoint step by step.</p><p>Integrating by parts twice:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/934/1*FaTMK6Zi6Mdh7kTmzBzExg@2x.jpeg" /></figure><p>The integral term shows that the formal adjoint is given by:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/929/1*vk6iT4e325lAAoV7d5-aAw@2x.jpeg" /></figure><p>-assuming real coefficients; the complex case requires the obvious conjugations-</p><p><em>L</em> is said to be formally self-adjoint when <em>L = L*</em>. This occurs precisely when <em>q = p’</em>. In that case <em>L</em> can be rewritten in the divergence (Sturm - Liouville) form:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/836/1*weasnNmCnedLWNc3eR6Xpg@2x.jpeg" /></figure><p>When <em>q = p’</em> the boundary form reduces to the classical Wronskian expression:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/820/1*PA-zkbpiuuTz9rAh-JImlQ@2x.jpeg" /></figure><p>Consequently Lagrange’s identity (Green’s formula) becomes:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0jug6BBHV_rXG3vNotFF6A@2x.jpeg" /></figure><p>If a domain 𝒟(L) is chosen so that Φ(u,v)=0 for all u,v∈𝒟(L), then:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1005/1*d9XvBTHi202-a9XaJmkUIQ@2x.jpeg" /></figure><p>and <em>L</em> is symmetric. When, in addition, the domain of the true Hilbert-space adjoint coincides with 𝒟(L), the operator is self-adjoint.</p><h3><strong>Spectral consequences</strong></h3><p>Once a differential expression has been turned into a genuinely self-adjoint operator on a Hilbert space, the spectral theorem becomes available. It converts the abstract self-adjointness into concrete statements about eigenvalues, eigenfunctions and expansions.</p><p>We are interested in the eigenvalue problem for <em>-L</em>, <em>Lu + λu = 0 </em>because the sign convention yields positive eigenvalues whenever <em>p&gt;0.</em></p><p><strong>Theorem</strong></p><p>Let <em>L = p d^2/dx^2 + q d/dx + r</em> act on a dense subspace <em>ℒ ²[a,b]</em> . Then:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1009/1*uiq79HPnix75-YcNQHQPCg@2x.jpeg" /></figure><p><strong>Proof of (iii)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/975/1*LRVgDC3vUV0_dsl1FT9OVA@2x.jpeg" /></figure><p>As a corollary: If <em>ρ</em> is a positive continuous function on <em>[a,b]</em>, then all eigenvalues of <em>Lu + λρu = 0</em> are real and the corresponding eigenfunctions are orthogonal with respect to the weight ρ.</p><p>In this article we clarified what it means for a linear operator to have an adjoint and, more importantly, when an operator is equal to its own adjoint, that is: when it is self-adjoint.</p><p>We then moved from the abstract setting into differential operators. By carefully integrating by parts we found the formal adjoint of a general second-order expression and saw the precise condition that makes the operator formally self-adjoint. That condition is exactly what turns the operator into the classical Sturm - Liouville form we all recognize.</p><p>Once suitable boundary conditions are chosen so that the leftover boundary terms disappear, the operator becomes truly self-adjoint. At that moment the geometry becomes clean and inevitable: every eigenvalue must be real, and eigenfunctions belonging to different eigenvalues are automatically orthogonal.</p><p>With these foundations in place, we are ready for the next step: the actual Sturm - Liouville eigenvalue problem and the spectral theorem that guarantees a complete set of eigenfunctions.</p><h3>Recommended books</h3><p>Al-Gwaiz, M. A. (2008). Sturm — Liouville theory and its applications. Springer.</p><p>Axler, S. (2024). Linear algebra done right (4th ed.). Springer. https://doi.org/10.1007/978-3-031-41026-0</p><p>Friedberg, S. H., Insel, A. J., &amp; Spence, L. E. (2019). Linear algebra (5th ed.). Pearson.</p><p>Kreyszig, E. (1978). Introductory functional analysis with applications. John Wiley &amp; Sons.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d46a5530dcbb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Regular Sturm-Liouville problems II- Second order linear ordinary differential equations and zeros…]]></title>
            <link>https://anacabreu.medium.com/regular-sturm-liouville-problems-ii-second-order-linear-ordinary-differential-equations-and-zeros-daf9487d80ca?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/daf9487d80ca</guid>
            <category><![CDATA[sturm-liouville]]></category>
            <category><![CDATA[differential-equations]]></category>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Wed, 22 Jul 2026 06:42:57 GMT</pubDate>
            <atom:updated>2026-07-22T07:21:24.528Z</atom:updated>
            <content:encoded><![CDATA[<h2>Regular Sturm-Liouville problems II- Second order linear ordinary differential equations and zeros of solutions</h2><p>Second-order linear ordinary differential equations (ODE) stand at the heart of countless models in mathematics and physics, from the simplest harmonic oscillator to far more intricate systems with variable coefficients. While explicit closed-form solutions are sometimes available, much of their essential character can be read from a purely qualitative feature: the zeros of the solutions (the points at which they cross or touch the equilibrium line). The number, spacing, and interlacing of these zeros reveal how rapidly a solution oscillates, how its behavior changes with the coefficients, and how different solutions of the same equation relate to one another.</p><p>In the f<a href="https://anacabreu.medium.com/regular-sturm-liouville-problems-i-functional-analysis-elements-the-%E2%84%92%C2%B2-space-904e36e3c192">irst article</a>, I talked about the functional analysis fundamentals. In this article, I explore the basic tools for constructing and understanding such solutions, beginning with the auxiliary equation for constant-coefficient problems, the Euler-Cauchy and the power-series method for equations with variable coefficients. I then introduce Sturm’s classical comparison and separation theorems, which provide powerful, explicit-free ways to locate and compare zeros across solutions and across different equations. These ideas form the foundation for the deeper oscillation theory that will later illuminate the structure of regular Sturm-Liouville problems.</p><h3><strong>Second Order Linear ODEs</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8x1vSODCWLbm8FPGhCpCOg@2x.jpeg" /></figure><p>Then <em>L[y]=f </em>is a second-order ODE on a real interval <em>I</em>. We say <em>L</em> is linear since, for any two class <em>C²</em> functions \<em>varphi</em>, ψ, two complex numbers <em>c₁</em> and <em>c₂</em>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/799/1*KWR2CxEPkgnL589l9HvONg@2x.jpeg" /></figure><p>Now, let <em>c</em> by a point belonging to the real interval <em>I</em> such that <em>a₀(c)=0</em>, then we call <em>c</em> a singularity and the equation is irregular. On the other hand, if for all points <em>x</em> of <em>I a₀(x)</em> is not zero, then the equation is regular and we can let:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/796/1*iuZ859rUz_JjCKfprmOtAw@2x.jpeg" /></figure><p>By the existence and uniqueness theorems, if <em>q(x)</em>,<em>r(x)</em>, and <em>g(x)</em> are continuous on <em>I </em>and <em>x₀</em> belongs to <em>I</em>, then:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/828/1*xMhEkGLBOfIBs136nQxFOQ@2x.jpeg" /></figure><p>Solving these kind of equations requires us to identify the structure: if <em>g(x)=0</em>, the equation is homogeneous and there exists two linearly independent solutions <em>y₁(x), y₂(x)</em> on <em>I </em>such that the general solution is a linear combination:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/819/1*ZRxP-1Gh4032cB7DXDtugA@2x.jpeg" /></figure><p>If <em>g(x) </em>is not zero, then the equation is inhomogeneous and a particular integral<em> yₚ</em> shall be found (many methods) and added to the general solution:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/819/1*eAaIvp1q4veC57GkzJIvug@2x.jpeg" /></figure><p><em>Constant coefficient linear ODES</em></p><p>Consider the equation:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/823/1*TuXivZSsbY0dH0aCLa_xcQ@2x.jpeg" /></figure><p>And let <em>f(x)=0</em>, when the coefficients are constants, the linear homogeneous equation admits an elegant exact solution via the auxiliary (or characteristic) equation. We simply try an exponential trial solution:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*mQQwOSAnqDW8A4blDuPM0w@2x.jpeg" /></figure><p>Differentiating twice and substituting immediately yields this quadratic:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/815/1*wvk2H3mUXcuA4WEs3bUVaA@2x.jpeg" /></figure><p>The two roots <em>λ₁, λ₂</em> completely determine the shape of the general solution (a linear combination of two independent basis functions). Distinct real roots give decaying or growing exponentials; a repeated root brings a factor of <em>xe^{\lambda x}</em>; complex conjugate roots produce the familiar damped or undamped oscillations <em>e^{\alpha x} (\cos \beta x + \sin \beta x)</em>.</p><p>Intuitively, think of the auxiliary equation as the frequency selector for the linear operator. Because constant coefficients make the equation translation invariant (due to linearity), pure exponentials are eigenfunctions of differentiation, and the algebraic equation simply asks which growth or decay rates (frequencies) are compatible with the differential equation. This is why we obtain closed-form solutions for free particles, simple harmonic motion, uniform vibrating strings, etc.</p><p><em>Euler-Cauchy</em></p><p>For the Euler-Cauchy equations, where:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/796/1*UxKdsbD2y3IeP7eVsdt2-g@2x.jpeg" /></figure><p><em>Power series</em></p><p>As soon as the coefficients become functions of <em>x</em> (the generic situation in Sturm-Louville problems coming from separation of variables in cylindrical or spherical coordinates) the exponential ansatz (solution guess) fails. There is no single <em>λ</em> that works for every <em>x</em>. The power series method (and its Frobenius extension for regular singular points) restores constructivity; we assume a solution of the form:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/792/1*IsdokVPJ_oJq1gyFAGDXHw@2x.jpeg" /></figure><p>around an ordinary or regular singular point <em>x₀</em>. Substituting into the differential equation produces a recurrence relation that determines every coefficient <em>aₙ</em> for <em>n</em> greater or equal than 2 in terms of the two arbitrary constants <em>a₀, a₁</em>. The resulting series converges in a disk (or up to the nearest singularity), giving an analytic solution that can be continued or matched to boundary conditions.</p><p>We can found solutions to certain second order ODEs through alternative methods. If one solution is known or guessed, the second is found by a first-order equation in the unknown multipliers. For non-homogeneous problems, variation of parameters can be used. Sometimes, numerical analysis methods can also work.</p><h3><strong>Boundary conditions</strong></h3><p>A second-order linear differential equation by itself describes an entire two-parameter family of possible solutions -think of it as a broad canvas of curves that all satisfy the same local relationship between a function, its slope, and its curvature. To single out one specific solution (or a discrete set of them) that matches a concrete physical or mathematical situation, we must supply two additional pieces of information. When these constraints are imposed at the ends of the interval on which we are working, they are called boundary conditions. They tell us, for example, that the solution must take a prescribed value at each endpoint (Dirichlet conditions, such as <em>y(a)=0 </em>and <em>y(b)=0</em>), that its slope must vanish there (Neumann conditions, such as <em>y’(a)= 0</em>), or that a linear combination of the value and the slope must satisfy a certain relation (Robin or mixed conditions).</p><p>Intuitively, boundary conditions act like the anchors that hold the solution in place: fixed ends of a vibrating string, insulated tips of a heat-conducting rod, or an elastically supported beam. Once these anchors are fixed, the differential equation and the boundary conditions together select solutions with very particular zero distributions (certain numbers of crossings inside the interval become allowed or forbidden). In this way, boundary conditions transform the general theory of second-order equations and their zeros into precise, well-posed problems whose solutions exhibit the structured oscillatory behavior we explore in the rest of this article.</p><p>For (1) on the real closed interval <em>I</em> that goes from <em>a</em> to <em>b</em>, the solution may be subjected to boundary conditions at interval’s end points, taking the form of:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/790/1*Ii4sos_0a7G4q1LVyZnWVQ@2x.jpeg" /></figure><p>Generalizing:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/813/1*7LLv8Pz-Ftrk8PF1guZQpQ@2x.jpeg" /></figure><p>Where:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/820/1*5OrFjnQ1ui9c-1-r2FMZJQ@2x.jpeg" /></figure><p>Then, the boundary value problem (BVP) is given by equations (1), (2) and (3). If<em> ξ = η =0</em> then the boundary conditions are homogeneous. If <em>α₃=α₄=β₃=β₄=0</em>, we say that the boundary conditions are separated. We call periodic boundary conditions to: <em>y(a)=y(b)</em>,<em> y’(a)=y’(b)</em>.</p><h3>The Wronskian</h3><p>Let<em> f </em>and <em>g</em> be two differentiable functions, the Wronskian of <em>f</em> and <em>g</em> is defined as:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/795/1*xydVz216KTeHD_WnNWQeVg@2x.jpeg" /></figure><p>The Wronskian <em>W</em> helps us detect independence. If you imagine the two functions as curves, <em>W</em> measures whether their value-and-slope pairs are pointing in linearly independent directions at each point (remember that the determinant of a 2x2 matrix tells you whether two vectors are parallel -linearly independent-).</p><p>When <em>W</em> is nowhere zero on an interval, the two functions are linearly independent: neither is a constant multiple of the other, and together they span the entire solution space of a second-order linear equation.</p><p>For solutions of the same homogeneous equation, Abel’s identity shows that <em>W</em> is either zero or never zero; it can change sign only if the solutions are dependent. This fact is the hidden engine behind Sturm’s separation theorem (which I will mention and prove later): because a non-vanishing Wronskian forces the ratio y₂/y₁ to be strictly monotonic between consecutive zeros of y₁ , the second solution is compelled to cross zero exactly once in every such interval. In short, the Wronskian converts the local geometric idea of two curves not being multiples of each other into a global statement about how their zeros must alternate.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/849/1*nVPHdEFyItNjyBldIGaXhg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/781/1*OaFi0D3ZSPcCYKjyyDfprQ@2x.jpeg" /></figure><p>Now, regarding linear independence:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/831/1*7UyizyZim7mfU1-yfoTthQ@2x.jpeg" /></figure><p>Try to prove yourself this little lemma with the definition of linear dependence.</p><h3>Zeros distribution</h3><p>A zero of a solution <em>y(x)</em> is simply a point where the graph crosses or touches the equilibrium line <em>y=0</em> . For a second-order linear equation the solution is uniquely determined by its value and derivative at any single point; consequently, a non-trivial solution can have at most a simple zero (if <em>y(x₀)=y’(x₀)=0 </em>, uniqueness forces y is equivalent to 0).</p><p>Intuitively, picture <em>y(x)</em> as the transverse displacement of a vibrating string or the wave function of a quantum particle. Each zero is a node, i.e.: a quiet point that never moves. The local spacing between consecutive zeros tells us the local wavelength and therefore the local frequency or energy cost. Where the restoring coefficient is large, zeros cluster more tightly; where it weakens, the solution stretches out.</p><p>It is important to say that the number and distribution of zeros along with other qualitative features of a solution (if any) to:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*vTpz4MqPrcTJ_NapIGTa-A@2x.jpeg" /></figure><p>are governed by the coefficients <em>q(x) </em>and <em>r(x)</em>. For example, consider the BVP:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/798/1*tSZRBU8oowY7VM6x6dhCuw@2x.jpeg" /></figure><p>We say that a function f has an isolated zero <em>x₀</em> in <em>I</em> if:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/814/1*Mz2KM9HGlC4llrK0u8hYmw@2x.jpeg" /></figure><p>Using that definition, we prove the following lemma:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/871/1*-Zn8N7zHJsbAYWnRPt5vRw@2x.jpeg" /></figure><p>It is time to talk about (and prove it) the Sturm’s separation Theorem which is the precise statement that independent solutions of the same equation take turns crossing zero.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/851/1*TyvE2_jrczqzou-X1n9HPg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yuhOvO8bpMGWcXSSf1gT5Q@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/856/1*PL1FtydOzqDhh70v0TC0OQ@2x.jpeg" /></figure><p>Now, to state and prove the Sturm’s comparison theorem we need the self-adjoint form of the equation (it is in the third article of this series that I’ll talk fully about the self-adjoint, do not worry about the fundamentals for now).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/840/1*ZLV2y9Vlm27SoJDqkZkbaw@2x.jpeg" /></figure><p>Since <em>v</em> cannot be zero, the zeros of <em>u</em> coincide with those of <em>y</em>.</p><p>While separation compares two solutions of one equation, comparison Theorem compares the oscillation rates of two different equations.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/876/1*CZbu_6IpESjLfInwuK3RPg@2x.jpeg" /></figure><p>The statement of Sturm’s comparison Theorem can be reduced to: the solution of <em>y</em> oscillates at least as fast as that of <em>z</em>. Larger <em>r</em> means a stronger restoring force. Think of two strings with the same length and boundary conditions but different tensions or densities. The higher-tension string has nodes closer together.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dt-yuzEohAFfU_fmDz1ouQ@2x.jpeg" /></figure><p>Lastly, we say that a non-trivial solution for <em>y’’+r(x)y=0</em>. is oscillatory if it has infinite number of zeros. If <em>r</em> is less or equal to zero, then the solution does not oscillate. But, if:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/828/1*Cz3zaggfSirvPHDhQPwWAg@2x.jpeg" /></figure><p>then any solution of the equation has infinite number of zeros distributed between the zeros of any solution of y’’+k2y=0.</p><p>The tools developed in this article (the Wronskian and Abel’s identity, the precise control of zeros via Sturm’s separation and comparison theorems, the transformation that puts any second-order equation into self-adjoint form, and the role of boundary conditions in selecting solutions with definite nodal patterns) are the concrete, geometric skeleton on which the entire abstract theory of regular Sturm-Liouville operators rests.</p><p>In short, everything that feels abstract in the next articles (domains of self-adjoint operators, essential self-adjointness, discrete spectrum, eigenfunction expansions, etc.) acquires clear intuitive meaning once we have seen how zeros behave, how independent solutions take turns, and how boundary conditions turn a differential expression into a well-posed, self-adjoint boundary-value problem. The tools developed in this article (the Wronskian and Abel’s identity, the precise control of zeros via Sturm’s separation and comparison theorems, the transformation that puts any second-order equation into self-adjoint form, and the role of boundary conditions in selecting solutions with definite nodal patterns) are the concrete, geometric skeleton on which the entire abstract theory of regular Sturm-Liouville operators rests.</p><p>In short, everything that feels abstract in the next articles (domains of self-adjoint operators, essential self-adjointness, discrete spectrum, eigenfunction expansions, etc.) acquires clear intuitive meaning once we have seen how zeros behave, how independent solutions take turns, and how boundary conditions turn a differential expression into a well-posed, self-adjoint boundary-value problem.</p><h4>Recommended books</h4><p>Al-Gwaiz, M. A. (2008). Sturm — Liouville theory and its applications. Springer.</p><p>Bear, H. S. (1999). Differential equations: A concise course. Dover Publications.</p><p>Graef, J. R., Henderson, J., Kong, L., &amp; Liu, S. X. (2018). Ordinary differential equations and boundary value problems: Volume I. Advanced ordinary differential equations. World Scientific.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=daf9487d80ca" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Draft: Ordinary Differential Equations done (rigorously) right]]></title>
            <link>https://anacabreu.medium.com/draft-ordinary-differential-equations-done-rigorously-right-16702f6f76c2?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/16702f6f76c2</guid>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Wed, 15 Jul 2026 18:20:55 GMT</pubDate>
            <atom:updated>2026-07-15T18:20:55.710Z</atom:updated>
            <content:encoded><![CDATA[<p>Besides my Regular S-L problems series, I’ve been seriously thinking on writing a series of pieces on ODEs far beyond solution techniques (a rigorous treatment).</p><p>Most ODEs undergrad courses and textbooks teach you how to solve specific classes of equations (separable, linear, exact, etc.) and that’s useful for computation but it leaves you with no deep mathematical understanding needed for research, modeling complex systems, or advancing to PDEs and dynamical systems.</p><p>For a good series on ODEs I am thinking on: Existence, uniqueness and data dependence (Picard — Lindelöf theorem via Banach fixed-point theorem on the integral operator, Gronwall’s inequality, Peano existence, etc.); Linear systems (some espectral theory, Floquet’s, stability, etc.); Qualitative aspects for nonlinear (Lyapunov’s, Hartman-Grobman, invariant sets, omega-limit sets, attractors); some perturbation theory, and a long etcetera.</p><p>I want to make this series concise but comprehensive and, to me, the best books for this are Arnold’s and Hale’s. They represent complementary pillars of a serious mathematical treatment: geometric intuition (Arnold) and analytic/qualitative depth (Hale). While Arnold’s treats ODEs as vector fields on manifolds from the beginning, revealing the geometry behind every phenomenon, Hale’s is the definitive analytic treatment of nonlinear qualitative theory. It is more comprehensive on several key topics that Arnold treats lightly or omits.</p><p>I mean, this is just an idea, it’s messy right now but, let me know (comments or X) if you would read the series or if there’s a particular topic or treatment you’d expect. This will take some time since I’m working on proving my first conjecture (each day it gets harder and harder, folks! I don’t know how to tame it already and it’s starting to embody the ‘I know I should sleep but the voices in my head go…’ song), so be patient.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/169/1*eYK3UOGJ-WxfZprYe57S8w@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/329/1*XlFwbMROT1_AIG2C6X6DKw@2x.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=16702f6f76c2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Regular Sturm-Liouville problems I-Functional analysis elements: the ℒ² space]]></title>
            <link>https://anacabreu.medium.com/regular-sturm-liouville-problems-i-functional-analysis-elements-the-%E2%84%92%C2%B2-space-904e36e3c192?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/904e36e3c192</guid>
            <category><![CDATA[fourier-series]]></category>
            <category><![CDATA[partial-differential-equa]]></category>
            <category><![CDATA[sturm-liouville]]></category>
            <category><![CDATA[functional-analysis]]></category>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Sat, 11 Jul 2026 05:12:18 GMT</pubDate>
            <atom:updated>2026-07-11T05:42:28.915Z</atom:updated>
            <content:encoded><![CDATA[<p>Regular Sturm-Liouville problems constitute the cornerstone of the spectral theory for second-order linear ordinary differential operators on a finite interval. Their primary use is the method of separation of variables for linear partial differential equations such as the heat, wave, or Laplace equation on bounded domains where the spatial operator reduces precisely to a Sturm-Liouville problem whose eigenfunction expansion yields the general solution as an infinite series that satisfies the boundary conditions.</p><p>This is a first article from a series on Regular Sturm-Liouville problems. In this opening article I therefore develop the indispensable functional-analytic tools in a self-contained, sequential manner.</p><h3><strong>Mathematical spaces: intuition</strong></h3><p>A mathematical space is a collection of objects (points) together with additional structure (given by axioms). Different axioms give rise to different spaces. The relations or rules between objects allow us to meaningfully talk about closeness, shape, distance (if the space is metrizable), direction, continuity or transformations.</p><p>Think of it as a flexible stage or arena where mathematical objects live and interact according to chosen rules. The Euclidean space is one simple example (points with the euclidean distance and angles), but spaces can be far more abstract: curved surfaces, manifolds, infinite dimensional function rules, networks of connection, or even sets of probability distributions.</p><p>In the late 19th-early 20th, Cantor set theory, Frèchet metric spaces, Hausdorff topological spaces and Hilbert/Banach spaces generalized the idea dramatically. The notion of space became any set with structure useful for analysis, physics or abstractions <em>per se</em>.</p><p>Here, we are interested in inner product spaces and complete inner product spaces called Hilbert spaces. An inner product space is a vector space equipped with an inner product function that allows us to measure lengths via the norm, angles and orthogonality (a central concept here) between vectors in a way that feels geometrically natural.</p><p>A Hilbert space is just a complete inner product space , i.e.: every Cauchy sequence of elements converges to a limit contained in the space: this ensures the arena is nicely closed. These spaces are the universe in which objects like Fourier series, quantum wavefunctions, PDEs solutions, etc., live with well defined angles and distances.</p><h3><strong>Inner product spaces</strong></h3><p>As mentioned before, an inner product space is a metric space equipped with an inner product.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/860/1*4koxQbB239YFMuXho-1EIg@2x.jpeg" /></figure><p>Now, from the inner product, a norm ||.|| which measures the length of an object in the space and induces a metric (distance), arises naturally:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/822/1*xb54dHs6f9MKyG1A_CwreQ@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/866/1*uwm2y2HuIMVi40ZFpdzVOw@2x.jpeg" /></figure><p>This axiomatizes the intuitive notion of distance and induces a topology via open balls.</p><p>In an inner product space:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/852/1*hqvRaSRblQZuaxYQU0Nwuw@2x.jpeg" /></figure><p>This is called the Cauchy-Schwartz inequality and though it is quite easy to prove (give it a try) it is tremendously powerful. As a corollary:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/859/1*uf4d5hPCPKhTCWdqiBo7Ow@2x.jpeg" /></figure><p>Let’s prove this corollary:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/854/1*mQ6XubGvNs4TSnKQtRKYJg@2x.jpeg" /></figure><p>From that, the triangle inequality follows:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/862/1*WVBNuG8fT2WTe3Ntz1kjSw@2x.jpeg" /></figure><p>Now we have the generalization of our arena: a metric space equipped with an inner product, from which a measure of length called norm naturally arises. From the norm, a distance between elements is obtained. From this, we can define orthogonality:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/857/1*xItJrDb3VXaxzduJe_3NKg@2x.jpeg" /></figure><h3><strong>The ℒ² space</strong></h3><p>It is time to conceptualize the space in which the objects we are interested in (Fourier series, Legendre polynomials, PDEs solutions) live: the ℒ² space.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/856/1*v0j01_C27Frc5IMoEpqg2w@2x.jpeg" /></figure><p>This space is equipped with the inner product:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/852/1*crAJ3rM8YoIV96q5d2tohA@2x.jpeg" /></figure><p>This induces the norm:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/862/1*LmI9Ge_5g0CdOPLBVjaqbg@2x.jpeg" /></figure><p>And, consequently, the induced metric (distance) between two functions is:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/855/1*x9uUUUG7yiBv49rn1yzrLQ@2x.jpeg" /></figure><p>Let us now prove that the inner product in our space satisfies the required axioms:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/860/1*Z1TRUaQBrb565LMFxcfc_A@2x.jpeg" /></figure><p>Try to prove that the norm and the metric satisfy the corresponding axioms.</p><p><strong><em>Completeness</em></strong></p><p>ℒ² is not just an inner product space, but also a complete inner product space or a Hilbert space.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/857/1*6Mh-TcQJkd4cErdG_eywjQ@2x.jpeg" /></figure><h3><strong>Orthogonality in ℒ²</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/860/1*WQQrorGzLdTypWuwV2cF3w@2x.jpeg" /></figure><p>We now enter the interesting part: we can write a function in ℒ² as a linear combination of orthogonal functions. This is the functional analysis foundation for objects like Fourier series.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/853/1*btvE1dBZt5J_wcxD-BbPsg@2x.jpeg" /></figure><p>Which means that the <em>kth</em> coefficient of the series is determined by the projection of the function on element of the orthogonal functions set. Hence:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/853/1*g-w7ugYP-jcpDQSAWyUeow@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/822/1*zWjOcGgqbwaXGmAlNYWeAw@2x.jpeg" /></figure><p>Which means that <em>f</em> can be represented by the infinite series on the right hand side.</p><p>Lastly, an orthogonal set of functions in ℒ² is complete if:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/823/1*wJqthSK2MIDtzr9QirKoEg@2x.jpeg" /></figure><p>And, the Bessel’s inequality becomes an equality (Perseval’s relation):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/819/1*fef0BZbSr59c4Xshx0F8rw@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/823/1*CRwOFLz85ONSeQiveeZiuw@2x.jpeg" /></figure><p>We say that a complete set of orthogonal functions in ℒ² is a basis of the entire space (it spans the space) and every function in ℒ² can be then expressed as a linear combination of that set.</p><h3><strong>Trigonometric Fourier series as a prototypical example</strong></h3><p>A concrete and historically central illustration of these functional analysis tools is the classical trigonometric Fourier series. Consider the real trigonometric system on the interval [π,-π] in the space ℒ² restricted to the interval and with the standard inner product. The functions:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/816/1*RzwT1YfQ4g9dsOcgavF5rA@2x.jpeg" /></figure><p>form a complete orthogonal set. In addition, their squared lengths (norms) are:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/817/1*j1sZiiUnT11UwYedO0i-Dg@2x.jpeg" /></figure><p>Hence, any function in ℒ²([π,-π]) admits the Fourier expansion:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/818/1*_Va_XOXc7ifxyQUXH4ctxA@2x.jpeg" /></figure><p>Which converges to <em>f</em> in the ℒ² norm. We can obtain the coefficients via the general formula for an orthogonal set:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/818/1*9EALBvwOUj9FYe1NIJUcNA@2x.jpeg" /></figure><p>The factor of 1/2 in the constant term arises from the conventional normalization that makes the formula for <em>a_n</em> uniform. This explicit dependence on the squared norms of the basis functions carries over directly to the eigenfunction expansions of regular Sturm-Liouville problems, where the weight function modifies the inner product while the coefficient formula retains the same structure:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/814/1*hQ_QiKnvISs_ih_WE8yDMw@2x.jpeg" /></figure><p>With this example we finish this first article from the series. Armed with this foundation, the next article in the series will introduce the regular Sturm-Liouville eigenvalue problem in its self-adjoint form, establish the reality and discreteness of its eigenvalues, prove the orthogonality of its eigenfunctions with respect to the weight function, and begin exploring the oscillation properties that make these eigenfunctions a natural basis for boundary-value problems.</p><h3><em>Recommended functional analysis books</em></h3><p>Kreyszig, E. (1978). Introductory functional analysis with applications. John Wiley &amp; Sons.</p><p>Brezis, H. (2011). Functional analysis, Sobolev spaces and partial differential equations. Springer.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=904e36e3c192" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The two-body problem (Lagrangian approach) to derive the orbital period of Earth: Happy New…]]></title>
            <link>https://anacabreu.medium.com/the-two-body-problem-lagrangian-approach-to-derive-the-orbital-period-of-earth-happy-new-2d379f8ac0b7?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/2d379f8ac0b7</guid>
            <category><![CDATA[the-two-body-problem]]></category>
            <category><![CDATA[lagrangian]]></category>
            <category><![CDATA[celestial-mechanics]]></category>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Thu, 01 Jan 2026 01:00:58 GMT</pubDate>
            <atom:updated>2026-01-01T01:00:58.601Z</atom:updated>
            <content:encoded><![CDATA[<h3>The two-body problem (Lagrangian approach) to derive the orbital period of Earth: Happy New sidereal Year!</h3><p>It is nearly 2026 and I cannot think about anything except celestial mechanics. The idea that a full cycle -which datum has been established by our species- has been completed: a new sidereal year, the completion of a full orbit.</p><p>So, to celebrate with you, here, I will use the principle of least action (the Lagrangian approach) to show a beautiful model that, at our scale, approximates nicely the sidereal year (without perturbations), i.e.: <em>T</em> -&gt; 365.2564 days. I will avoid cumbersome algebraic calculations since I am aiming for a brief text.</p><p>Let’s then use the elegant Lagrangian approach to solve the two-body problem where the system is made of two entities: our Sun and our Earth. With the equations of motion, Kepler’s laws can be deduced along with the Earth’s orbital period.</p><p>I will use planar polar coordinates since conservation of angular momentum allows it. The coordinates (<em>r</em>,<em>θ</em>) will describe, respectively, radial position and angular position.</p><h3>The equations of motion</h3><p>Let <em>M</em> be the mass of the Sun and let <em>m</em> be the mass of Earth such that <em>M</em>&gt;&gt;<em>m. </em>This will allow us to treat the Sun as a fixed body though I will keep the general reduced mass (rigor), i.e.: μ= (<em>Mm</em>)/(<em>M+m</em>)→<em>m</em>.</p><p>Consider this two-body system in an inertial frame of reference such that<strong> r </strong>is the relative position vector whose magnitude<em> r</em> is the distance bewteen the bodies. Now we look for the Lagrangian ℒ = <em>T−V, </em>where <em>T</em> is the kinetic energy of the system and<em> V</em> is the potential (gravitational) energy. Taking <em>G</em> as the gravitational constant:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bUTmpbNgqhEK34_XcFVmiA.jpeg" /></figure><p>By Euler-Lagrange, for angular motion:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*apicUfqYI6ZacKLZOyGMbA.jpeg" /></figure><p>It is worth mentioning that this confirms Kepler’s second law about constant areal velocities:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fRCGNDoRRRv0F3TzV7axKw.jpeg" /></figure><p>For radial motion</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XPXJN9q_8v9e-bhZCX9m4Q.jpeg" /></figure><p>Therefore,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SoAm_PaSuNqx84IxhP1LZQ.jpeg" /></figure><p>Is the radial equation of motion.</p><h3><strong>The elliptical shape of the orbit</strong></h3><p>Now, let’s solve for the orbit’s shape (elliptical). For that, we shall use Binet transformation. Let<em> </em>ϕ=1/<em>r :</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3CdVvgCYwCTyq-JfeWVB8Q.jpeg" /></figure><p>After some cumbersome algebra, we get this linear, inhomogeneous differential equation:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QBuGIl9sWCTkYqdQsUrKpA.jpeg" /></figure><p>Solving it as simple harmonic motion with constant shift <em>ψ</em>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Slx7IcHU4X21ws8fn15UyA.jpeg" /></figure><p>Thus, taking the perihelion at <em>θ = 0, ψ</em> = 0 and defining eccentricity:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Dczgsaazf7lMQKRWzCqVmg.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZTL2N4ADvMxOhyBZh0HqPw.jpeg" /></figure><p>Letting</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1msvJmhTI5is6yp759X7pQ.jpeg" /></figure><p>And, finally, substituting ϕ, we obtain the elliptical orbit (first Kepler’s law):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QPhPxPixnjck3BKeSD15DA.jpeg" /></figure><h3>The orbital period</h3><p>Finally, we derive the orbital period (one Earth year cycle). Let T be the orbital period, i.e.: the time for <em>θ </em>to achieve 2π radians.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Wfe8gyzDoVmD19c4cdW_og.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9ZimFAhz_WmVGPq-MJq0eg.jpeg" /></figure><p>Takin a as one AU, we get, as expected: <em>T</em> -&gt; 365.2564 days.</p><p>Happy New sidereal Year!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2d379f8ac0b7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The breeding ground for the Special Relativity Theory]]></title>
            <link>https://anacabreu.medium.com/the-breeding-ground-for-the-the-special-relativity-theory-lets-make-the-quabla-operator-invariant-e1082cf54ae6?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/e1082cf54ae6</guid>
            <category><![CDATA[special-relativity]]></category>
            <category><![CDATA[wave-equation]]></category>
            <category><![CDATA[physics]]></category>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Wed, 27 Aug 2025 02:49:39 GMT</pubDate>
            <atom:updated>2025-08-27T18:51:13.423Z</atom:updated>
            <content:encoded><![CDATA[<h3>The breeding ground for the Special Relativity Theory: let’s make the quabla operator invariant again</h3><p>In 1632, in his “<em>Dialogo sopra i due massimi sistemi del mondo</em>”, Galileo proposed the Galilean relativity principle, also known as the Galilean invariance, which states that the fundamental laws of physics are invariant in all inertial frames of reference (IFR).</p><p>Later, Newton would formalize this principle through the ideas of classical mechanics, which embodied the concepts of absolute, independent time and relative space for inertial observers, where time is universal and spatial coordinates change in accordance to the Galilean addition of velocities.</p><p>Although Newton rejected the Cartesian mechanical <em>Aether</em> (more like a <em>plenum</em> in Aristotelian terms), he did believe in absolute space. In Book 1 of the <em>Principia</em>, he states:</p><blockquote>Absolute space, in its own nature, without regard to anything external, remains always similar and immovable…</blockquote><p>In group theory terms, the Galilean invariance is represented by a Lie Group called Galilean group <em>SGa</em>l(3), in which four types of transformations (symmetries) live and combine through composition: spatial translations, rotations, time translations and Galilean Boost. These transformations and their compositions preserve isochronicity (invariance of time intervales), isotropy (invariance under rotation), homogeneity and -for isolated systems- energy conservation.</p><p>Let us think of ℝ⁴, with the Euclidean metric as the space in which Newtonian mechanics comes into play along with <em>SGal</em>(3). The Galilean transformations map (<em>x y z t</em>)ᵀ into (<em>x’ y’ z’ t’</em>)ᵀ and viceversa. Let (<em>x y z t</em>)ᵀ be the position vector of an arbitrary event in the IFR of the observer <em>O</em>: <em>S</em> and let (<em>x’ y’ z’ t’</em>)ᵀ be the position vector of the same event but through the perspective of observer <em>O’</em> in <em>S’</em>. Now suppose that <em>S’</em> is moving at a constant velocity <em>v</em>, in the <em>x</em>-axis direction from, starting both at <em>t=t’=</em>0 (time in which the origins of both IFR coincide), then:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0vr9UUWCbb1JMcmmyvwN_A.jpeg" /></figure><p>To go from<em> S</em> to <em>S’</em>, we just need to apply the inverse.</p><p>In fact these transformations satisfy the Galilean principle of relativity when it comes to the invariance of Newton’s laws of motion. What about light?</p><p>In 1678, in his “<em>Traité de la lumière</em>”, Huygens proposed the undulatory theory of light by modeling light as an object which propagates as spherical wavefronts through a continuous medium: the <em>Luminiferous Aether</em>. This hypothesis will compete with Newton’s corpuscular theory of light (<em>Opticks,</em> 1717).</p><p>During the dawn of the electromagnetic theory, in the 19th C., both the undulatory theory of light and the Cartesian <em>Aether </em>(with properties necessary to be the medium of propagation of light waves) would reappear.</p><p>In 1801, Thomas Young performed the double-slit experiment, showing interference fringes (a wave pattern) produced by light, thus confirming the undulatory theory of light, perfected a couple of decades later by Fresnel, who incorporated transverse waves to explain polarization (after his diffraction experiments). With this, it was empirically (experimentally) proven that light, indeed, behaves like waves: it diffracts, it reflects and shows interference patterns.</p><p>On July 21, 1820, the results of Ørsted observations were published. He noticed that, in presence of a wire in which electric current flows, a magnetic needle deflects, i.e.: an electric current induces a magnetic field. A decade later, Faraday proved the converse: a magnetic field induces an electric current. In addition, Faraday gave birth to the theory of fields by proposing the concept of lines of force, keeping the need for a continuous action between two particles of matter.</p><p>Finally, in 1860s, the 19th C. Euclid of electromagnetism: Maxwell, unified the works of Gauss, Ørsted and Faraday, deriving the electromagnetic theory through which the constancy of the speed of light (an electromagnetic wave) was predicted. According to Maxwell’s equations, electromagnetic waves travel at a constant speed of:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lIR8Rhr-NQ1Iq2XCViWPIg.jpeg" /></figure><p>Where μ₀ is the permeability of free space and ε₀ is the permittivity of free space (both constants).</p><p>Now, since the lines of force of electricity and magnetism are orthogonal to the propagation of the wave, light waves shall be transversal waves, which, in accordance with the undulatory theory, a continuous medium of propagation was needed: the <em>Luminiferous Aether, </em>again.</p><blockquote>The undulatory theory of light requires us to admit this kind of elasticity in the luminiferous medium, in order to account for the transverse vibrations. We need not then be surprised if the magneto-electric medium possesses the same property.<em> (Maxwell, 1861)</em></blockquote><p>Now, if light behaves like a wave, then it certainly satisfies the wave equation:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/893/1*wOaXFMh7iQRUsjfKKPolPA.png" /></figure><p>Where ∇² is the Laplacian, ψ is the wave function and <em>c </em>is the speed of propagation of the wave -in this case, it is the speed of light-.</p><p>As a law of nature, according to the principle of Galilean relativity, the wave operator (D’Alembertian or quabla operator):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AdGA3Tabh1KJ1HboLI8xyw@2x.jpeg" /></figure><p>Should be invariant under any Galilean transformation. Let us take the Galilean boost in order to transform the wave operator, noticing that both<em> x </em>and <em>t </em>are both functions of <em>x’, t’</em>.</p><p>By the chain rule:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/664/1*27oOmzpozx_HSZrAR-j2Hg.png" /></figure><p>Since<em> x’=x-vt</em>, <em>v</em> constant:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/590/1*AgD6Ui3Pr3p4ZL7gCQkIfA.png" /></figure><p>Thus:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*W2o1WCntqEZu-0rUwVOjow.png" /></figure><p>Now, we need the second partial derivatives, but first observe that ∇²=∇ ’², so we just have to find the second partial derivative with respect to time:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/637/1*o-rRtGRRfcVkaZ1kLEYqiQ.png" /></figure><p>Since:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/674/1*VMeF1zgO7MpE545bz1lDDg.png" /></figure><p>then:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/625/1*IaUYydov2Vn0b-n1G3CLKw.png" /></figure><p>And, after applying the Galilean boost, we have obtained:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/922/1*B68prSP2By8MmuorD6XEMw@2x.jpeg" /></figure><p>which means that the quabla operator is not invariant under the Galilean boost.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/340/1*4whiS3S8lASOy6RkcXMOGw.png" /></figure><p>In 1887, Michelson and Morley made use of the interference of light to assess the state of motion of the all-pervading <em>Aether</em> with respect to Earth’s rotation. The <em>Aether</em>, as an especulative rigid frame, was supposed to generate a “<em>Aether </em>wind”<em> </em>due to the motion of bodies through it. They obtained a negative result by finding no evidence of <em>Aether </em>as from the undeniable (now tested) constancy of the speed of light. The fate of the <em>Luminiferous Aether</em> was sealed.</p><p>In 1905, Einstein published his “On the Electrodynamics of moving bodies”, taking the <em>Aether</em> out from the <em>Zeitgeist</em> while killing the Absolute time of Newton. There, he proposed the two principles that govern Special Relativity Theory:</p><ol><li>The principle of relativity: the laws of physics are the same (invariant) in all inertial reference frames.</li><li>The constancy of the speed of light: the speed of light in vacuum is constant for all observers, regardless of the motion of the source or the observer.</li></ol><p>Now, it is relevant to mention that many great minds contributed to this theory, including Poincaré, Minkowski, Lorentz, FitzGerald, among others.</p><p>The time came to make quabla invariant. <em>Grosso modo</em>, in a Minkowski space, the Lorentzian group of transformations (subgroup of a Poincaré group) leaves the D’Alambertian invariant (amongst many other things).</p><p>At the end, the electromagnetic theory was right. Why? Because, instead of being a derivation of a set of speculations, it emerged from directly asking to nature (as my relativity professor, Francisco Nettel, in the spirit of Ilya Prigogine, said).</p><p><strong>References</strong></p><p>Bergmann, P. G. (1976). Introduction to the theory of relativity (Rev. ed.). Dover Publications.</p><p>De Andrade, E. M. P., Faber, J., &amp; Rosa, L. P. (2013). A spontaneous physics philosophy on the concept of ether throughout the history of science: Birth, death and revival.Foundations of Science, 18(3), 559 – 577.</p><p>Ginoux, J.-M. (2024). Poincaré, Einstein and the discovery of special relativity: An end to the controversy. Springer.</p><p>Michelson, A. A., &amp; Morley, E. W. (1887). On the relative motion of the earth and the luminiferous ether. American Journal of Science, 34, 333 – 345.https://doi.org/10.2475/ajs.s3-34.203.333.</p><p>Rahaman, F. (2022). The special theory of relativity: A mathematical approach (2nd ed.). Springer.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e1082cf54ae6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A glimpse into the realm of Autonomous Harmonic Oscillators, constants of the motion and Fosse’s…]]></title>
            <link>https://anacabreu.medium.com/a-glimpse-into-the-realm-of-autonomous-harmonic-oscillators-constants-of-the-motion-and-fosses-1015b3a857d2?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/1015b3a857d2</guid>
            <category><![CDATA[harmonic-oscillator]]></category>
            <category><![CDATA[jon-fosse]]></category>
            <category><![CDATA[mechanics]]></category>
            <category><![CDATA[mathematical-physics]]></category>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Wed, 09 Jul 2025 02:20:32 GMT</pubDate>
            <atom:updated>2025-07-09T15:38:00.496Z</atom:updated>
            <content:encoded><![CDATA[<h3>A glimpse into the realm of Autonomous Harmonic Oscillators, constants of the motion and Fosse’s Aliss at the Fire</h3><p>A couple of years ago, I was finishing Jon Fosse’s (2023 Nobel prize) “Aliss at the Fire”. Naturally, I could think of nothing else but harmonic oscillators and the undeniable, inherent beauty of periodicity.</p><p>Grief can trascend irreversibility. Signe’s mind broke the arrow of time. Her memory is caught in a cycle where pain precedes hope, and hope gives way to pain. This inescapable duality constitutes and irreducible reality, where longing dissolves the linearity of time: circle’s non-contractibility.</p><p>Asle’s disappearance is the alpha and omega of a universe reborn through its own iterative death: cosmic Dance of Shiva. That is “Aliss at the Fire”: repetition within resignation.</p><blockquote>… and she, lying there on the bench, sees herself standing and looking at the flames in the stove and then she sees herself go over to the window and she sees herself stand there and look out the window and then she looks, standing there in front of the window, at the bedroom door and it opens… <em>(Fosse, 2010)</em></blockquote><p>I confess that, between Laplace’s Demon and Heisenberg’s uncertainty, I would choose determinism. Between dissipative forces and constants of motion, I am more inclined toward conservative systems. I dream of an ergodic, static, analytic ultimate reality -an oniric, homogeneous container where order is preserved through a symmetry yet to be unveiled.</p><h3><strong>Autonomous Harmonic Oscillators</strong></h3><h4>Brief history</h4><p>The origins of dynamics traces back to the study of mechanical systems. One of the most studied classical system is the harmonic oscillator. The study of oscillatory phenomena began with observations of natural, periodic systems like pendulums and vibrating strings. The Pythagoreans, who laid the groundwork for understanding harmonic motion in music, investigated the relationship between the pitch of a sound and the length of a vibrating string.</p><p>Around 1602, during the scientific revolution, Galileo noticed that, for small perturbations, the period of a pendulum’s swing is almost independent of its amplitude. That is how he designed a pendulum clock. In 1656 Huygens built the first pendulum clock. This engineer also derived the formula for the period of a simple pendulum while introducing the concept of isochronism and contributing to the study of synchronization phenomena (coupled oscillators). In 1678, Hooke’s work on springs (Hooke’s law) was published.</p><p>The mathematical formalization followed: Netwon’s “Principia” (laws of motion), Euler’s “De Novo Genere Oscillationum” (differential equations for mass-spring systems, wave propagation, strings , etc.), Bernoulli’s superposition principle for normal -harmonic- modes, Lagrange’s “Mécanique Analytique”, Hamilton, Fourier, etc.</p><p>Since then, the study of oscillators has been expanded to wave phenomena, practical applications and quantum mechanics. The simple harmonic oscillator became a fundamental model in modern physics (even atoms are modelled as oscillators). Oscillators are relevant nowadays: nonlinear and coupled oscillators revealing chaotic behavior, fields being treated as collections of oscillators in Quantum Field Theory and, in general, the harmonic oscillator playing a <em>sine qua non</em> role as a foundational physics topic (even to introduce or give an intuition regarding differential equations).</p><h4>The model</h4><p>For small displacements, a pendulum experiencing a frictional force is an example of a damped harmonic oscillator. A spring with a mass attached at one end and fixed at the other is an even simpler one.</p><p>Using Newtonian mechanics, we can model the damped harmonic oscillator as an ideal particle of mass <em>m, </em>moving, to and fro, along the <em>x</em>-axis (let’s keep it simple: 1-D), from its equilibrium point. Let <em>x(t) </em>be the position of the mass at time <em>t </em>and, let <em>ẋ(t)</em> and <em>ẍ(t) </em>be, respectively, the velocity and the acceleration of the mass.</p><p>By the second Newton’s law, the force acting on the mass is equal to the rate of change of its linear momentum:<em> F = mẍ. </em>In addition, the spring is particularly known by the restorative force, proportional to its displacement: move the spring from its equilibrium and it will “try” to “return” with a force according to Hooke’s Law: <em>H = -kx, </em>where <em>k </em>is a constant that tells how stiff the spring is.</p><p>Adding a resistive (dissipative) force <em>R</em> leads to a damped motion. This force, always opposite to velocity’s direction -since it slows down the mass-, can be thought as the friction that impedes the system to oscillate forever. Thus, <em>R = -αẋ, </em>with <em>α</em>, a real greater or equal to zero, being the damping constant.</p><p>Hence, we can express the equation of motion of our system as follows:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/417/1*1az5GOx-2nPzE4HtCpXn4w.png" /><figcaption>Equation of motion of a damped harmonic oscillator</figcaption></figure><p>where Γ = α/(2m) is the damping parameter andω₀ = √{k/m} is the natural angular frequency of the oscillation, i.e.: the number of revolutions per second (inversely proportional to the period of the motion).</p><p>Notice that ẍ + 2Γẋ + ω₀² x = 0 is an homogeneous (autonomous or non-driven system), second order linear equation with constant coefficients and, therefore, with analytical solution. This equation can be easily solved by finding the transient state through the auxiliary quadratic equation λ² + 2Γλ + ω₀² = 0. If the discriminant ▵ &gt; 0 (overdamped oscillator), then x(t) = Be^{λ₁t} + Ae^{λ₂t}. If ▵ = 0 (critical damping), then x(t) = (Bt+C)e^ {−ω₀t}. Lastly, if ▵ &lt; 0 (underdamped oscillator), then x(t) = cos (α t + ϕ)Ae^{−2Γt}.</p><p>Now, if the dissipative force is zero (zero damping constant), then we get the equation of motion of a simple harmonic oscillator (purely imaginary solution):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/243/1*D3_KodGhxNfLn8yC8W6ipw.png" /><figcaption>Equation of motion of the simple harmonic oscillator and roots for the auxiliary equation</figcaption></figure><p>And x(t) = A cos (ω₀ t )+ B sin (ω₀ t) is the position of the mass that will perpetually oscillate.</p><p>The beauty of the simple harmonic oscillator is evident: the phase space will show closed paths around a single point, indicating a quantity conserved, a constant of the motion, thus, endless cycles.</p><h4>Constants of motion</h4><p>Let’s formally define the concept “constant of the motion”:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*p7xca5fi2QIr2sYEQVhjzw.png" /></figure><p>If we are provided with <em>K</em>, we know that it is a constant of the motion if, for all<em> x</em> and <em>y</em>, thet total derivative of <em>K</em> is zero (time invariant):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/394/1*PsJJOBArJ5HjiVvOvhAwyA.png" /><figcaption>zero total derivative</figcaption></figure><p>The Simple Harmonic Oscillator is a conservative system in which energy is a conserved quantity. Let’s show this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aGi8FQ7ibvXOY-lCPNeLkA.png" /></figure><p>Now,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/546/1*FZfe2L3qQvOrgOPlcEIbqg.png" /></figure><p>And, the total mechanical energy of our system is a constant of the motion.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/689/1*uO0pAMpH1NecLT24x4BmGQ.png" /></figure><p>Constants of motion are relevant when it comes to stability analysis of nonlinear systems in which fixed points are elliptic. By the Hartman-Grobman theorem, we know that, in the neighborhood of an hyperbolic fixed point, the behavior in the linearized system agrees with that of the nonlinear system. But, this is not true when it comes to centres. To find the nature of an elliptic point, it is useful to look for constants of the motion. If there is a constant of the motion, our equilibrium point is a centre.</p><p>It is your turn to show, through contants of the motion, that the total mechanical energy of a Damped Harmonic Oscillator is not time invariant.</p><p><strong>References:</strong></p><p>Dugas, R. (1988). A history of mechanics. Dover Publications.</p><p>Mach, E. (1919).The Science of Mechanics: A Critical and Historical Exposition of its Principles. (T. J. McCormack, Trans.). The Open Court Publishing Co</p><p>Taylor, J. R. (2005). Classical mechanics. University Science Books.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1015b3a857d2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Euler’s identity and Ogawa’s Professor beloved equation]]></title>
            <link>https://anacabreu.medium.com/eulers-identity-and-ogawa-s-professor-beloved-equation-ce5be5202746?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/ce5be5202746</guid>
            <category><![CDATA[number-theory]]></category>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[books]]></category>
            <category><![CDATA[euler]]></category>
            <category><![CDATA[pi]]></category>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Sat, 15 Mar 2025 00:58:40 GMT</pubDate>
            <atom:updated>2025-03-15T23:44:25.327Z</atom:updated>
            <content:encoded><![CDATA[<h3>Euler’s identity and Ogawa’s Professor’ beloved equation</h3><h4>Today is π day so, let’s celebrate it by deriving one of the most beautiful relationships in mathematics: Euler’s identity (the Professor’s beloved equation).</h4><p>Some -smooth and analytic- functions <em>f </em>are such that:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/429/1*WvnI4FUiOdJGmGO-W42P3g.png" /><figcaption>Taylor expansion</figcaption></figure><p>Where<em> T </em>is the funcion’s Taylor expansion. If <em>c = 0, </em>then we call it <em>f</em>’s<em> </em>Maclaurin series:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/226/1*YTWi0addAnZC2325LL1sOA.png" /><figcaption>Maclaurin expansion</figcaption></figure><p>Thus, the Maclaurin series for <em>f(x) = exp(x) </em>is given by:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/258/1*zYrrVxO4yUOnptdKtct-0g.png" /><figcaption>exponential Maclaurin expansion</figcaption></figure><p>Now, if we let <em>x = iθ </em>, then:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/441/1*EH-46BuuSnNErfOFi22HIA.png" /><figcaption>Rearranging and substituting</figcaption></figure><p>Since the Maclaurin series for the sine and cosine functions are:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/298/1*faivISIKgSCfE4G5_m2UHw.png" /><figcaption>Trigonometric expansions</figcaption></figure><p>And,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/510/1*zRsnWd67qTdgxDdRxlDF2w.png" /><figcaption>Rearranging</figcaption></figure><p>Then,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/161/1*s4cxtRSSmXT2Cc9ZBqXX8g.png" /><figcaption>Beautiful equality</figcaption></figure><p>For <em>θ = π</em>, we get:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/178/1*tr7Ow2qS42CPAQZK45U7pw@2x.jpeg" /></figure><p>Hence,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/589/1*-4ETC3CNwAZoWyS4TfbPTQ@2x.jpeg" /></figure><p>Which is the Professor’s beloved equation. The equation written by the Professor to end the argument between his sister-in-law and the houskeeper.</p><p>Yōko Ogawa’s “The Professor and the Houskeeper” is a heartwarming novel about an elderly number theorist whose memory lasts for only 80 minutes. After that lapse, he cannot remember anything but the memories he accumulated before the car accident.</p><p>This is not a story about mathematics. Of course, the love for number theory is inherent and omnipresent but, the beauty of the bond between the housekeeper, the Professor and the housekeeper’s son (called Root, by the Professor) exceeds every other element within the story.</p><blockquote>“He treated Root exactly as he treated prime numbers. For him, primes were the base on which all other natural numbers relied; and children were the foundation of everything worthwhile in the adult world.”</blockquote><p>The Professor, who despite his condition, never abandoned his mathematics, transmitted his love for discrete mathematics to both, the Houskeeper and Root. She -particularly- discovered the beauty of mathematics. In exchange, she granted him with true, uninterested friendship.</p><p>And, that’s all what this story is about: friendship, fraternal love, tenderness. About the joy of experiencing the happiness of our loved ones. About admiration, respect. About the virtue of never abandoning.</p><p>This book is an absolute must! Your heart will definitely melt.</p><p>Happy pi day!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ce5be5202746" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Conservation of mechanical energy -by Châtelet- from the fundamentals of Newtonian mechanics]]></title>
            <link>https://anacabreu.medium.com/conservation-of-mechanical-energy-by-ch%C3%A2telet-from-the-fundamentals-of-newtonian-mechanics-50bcd29bc7a1?source=rss-be6f053229d------2</link>
            <guid isPermaLink="false">https://medium.com/p/50bcd29bc7a1</guid>
            <category><![CDATA[energy]]></category>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[mechanics]]></category>
            <category><![CDATA[newton]]></category>
            <category><![CDATA[mathematics]]></category>
            <dc:creator><![CDATA[Γ(z)]]></dc:creator>
            <pubDate>Wed, 25 Dec 2024 23:32:46 GMT</pubDate>
            <atom:updated>2024-12-26T01:13:13.394Z</atom:updated>
            <content:encoded><![CDATA[<h4>In honor to the genius, on his birthday</h4><p>The word consistency, to me, is tightly intertwined with the consequences of Gödel’s Incompleteness Theorems (the dagger that killed Hilbert’s <em>Wir werden wissen</em>). Although I am skeptic about human’s capacity to understand physical reality as the complete whole, as the One it is, I shall confess that I believe in relative consistency as the most valuable tool we posses when it comes to overcome our human intellectual limitations.</p><p>Think about a dialectic process in which, to explain a set of phenomena, a first model is proposed. The model is locally tested. If it works, we keep it. The model is useful but, as a matter of fact, it fails under different conditions: it is not global, it is incomplete: since it does not explain the totality of phenomena. A second model emerges. The requirement for this second model to be as valid as the first can be seen as a kind of equivalence: it must explain (predict), at least, as much as the first model, i.e.: the true/real statements of the first model should be “trasnferrable”.</p><p>If we validate this second model based on its relative consistency with respect to the first model then it would be fair to think that the structure of our building of knowledge, although incomplete, is decently solid. Sounds familiar? The robustness of classical mechanics is undeniable. Is it the only way? Not at all. It is just the path we have taken.</p><p>In mathematics, there are several examples of relative consistency. For example, Hilbert showed that euclidean geometry is as consistent as plain algebra. Another example is the logical relation between the system of reals and any hyperreal number system. Today, we know that (by the transfer principle -Robinson’s nonstandard analysis-) any true sentence of real numbers is true for hyperreal numbers.</p><p>The Newtonian model is incomplete but the equivalence with the energy approach, derived by one of the most beautiful minds in humankind history, made modern analytic (Lagrangian and Hamiltonian formulations) mechanics’ emergence possible.</p><p>Within Definition III of his Principia, Newton characterizes inertia: the <em>vis insita</em> of matter (term coined by Kepler but redefined to include uniform linear motion) as a:</p><blockquote>… power of resisting by which ever body, as much as in it lies, endeavours to persevere in its present state, whether it be of rest or of moving uniformly forward in a right line. <em>(Newton, 1687)</em></blockquote><p>This idea of resistance to motion due to inherent properties of the system: either inertial mass (resistance to linear motion) or moment of inertia (resistance to rotation), is, in my opinion, the Newtonian mechanics’ heart of heart. Not only reflects the philosophical Zeitgeist, but also explains the non-triviality of stating both the first and the second laws of motion although, mathematically, the first law is not independent from the second law but a special case of this. Laws I and II follow from this idea blended with proportionality.</p><blockquote><em>LAW II</em></blockquote><blockquote>The alteration of motion is ever proportional to the motive force impressed; and is made in the direction of the right line in which that force is impressed. <em>(Newton, 1687)</em></blockquote><p>Which means that the state of motion of a body can be deduced from the resultant force acting on it divided by the total constant, inertial mass of it, i.e.: by solving the differential equation:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/462/1*qrT7swI28sIiFR-IoLXz7w.png" /><figcaption>Newton’s Second Law of motion as an ordinary differential equation</figcaption></figure><p>If the force depends only on position (non-dissipative force), then we have an ordinary differential equation that can be solved by variable separation:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/414/1*oxIXli9TRqjG7Ep5N1xXwQ.png" /><figcaption>Second law as ODE</figcaption></figure><p>If the force is conservative, then the potential energy:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/215/1*g4Guwwsk-VPPZ7FnFGpusQ.png" /><figcaption>Potential energy</figcaption></figure><p>Substituting and rearranging:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/184/1*V10BLVlB4Cib_8TW6DbuCw.png" /></figure><p>Which does not mean too much without the concept of kinetic energy worked by the inspiring mathematician and natural philosopher of the 18th century, Émilie Du Châtelet (tutored by Maupertuis -the mind behind the origins of the principle of least action-), who derived the energy conservation principle from the fundamentals of Newtonian mechanics while translating Newton’s Principia from Latin to French.</p><p>This derivation was the quid for later development of Lagrangian and Hamiltonian mechanics.</p><p>Thanks to Émilie, we know that the kinetic energy of the system is given by:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/175/1*icsk49_2g7tp5PKS4CdRyg.png" /><figcaption>Kinetic energy</figcaption></figure><p>Hence,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/167/1*3W30oRSzi9nFQxK4Ip2r2w.png" /><figcaption>Mechanical energy</figcaption></figure><p>Since the total mechanical energy of the system is the sum of both potential and kinetic energy, then <em>C</em>, here, is constant.</p><p>Another way to verify that a quantity is conserved is by observing the net change. If the sum of changes is equal to zero, then we have a constant of motion.</p><p>Let work<em> W</em> be the energy transferred by a force such that (work-energy theorem):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/159/1*Z5wLzeMlQuV6kLjxmJvCDg.png" /><figcaption>Work-energy theorem</figcaption></figure><p>If the force is conservative:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/165/1*xyiUGH8AHjZsAll51ha3Lw.png" /><figcaption>Work and potential energy</figcaption></figure><p>And that is how:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/401/1*DuDvNxQ2PNWAAX6-Yff_Kg.png" /></figure><p>Beautiful, isn’t it? One model served as the foundation of a second one: the powerful, equivalent energy approach. Now, try to solve the simple pendulum problem with both Newtonian and pre-Lagrangian models, just to feel the elegance of Du Châtelet’s derivation.</p><p>Aren’t you sure about the equivalence? Let us corroborate:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/722/1*02YRVkcJElWDZzTsfMWdqw.png" /><figcaption>Equivalence Newtonian-Energy approach</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=50bcd29bc7a1" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>