## Worked example 4: Newton-like conservation laws from a single continuity equation, for a localized self-sustaining energy knot This example is deliberately concrete. We assume only: 1. A nonnegative energy density $u(\mathbf{x},t)\ge 0$. 2. A local energy-flux vector field $\mathbf{S}(\mathbf{x},t)$. 3. A single continuity equation (energy bookkeeping under continuous transport): $$ \partial_t u + \nabla\cdot \mathbf{S}=0. $$ We do *not* assume Newton’s laws. We do *not* assume momentum conservation as an axiom. We do *not* assume “force” as a primitive. We show how the *form* of Newtonian-looking statements emerges as identities once: - you define a localized object as a region of concentrated energy, - you track the motion of that region using its center-of-energy, - and you interpret changes as flux across its boundary. The key idea is: “laws of motion” are accounting identities for localized conserved flow. --- # 1. Localized object definition: an energy knot as a moving concentration Let $V$ be a (possibly time-dependent) region enclosing a localized concentration of energy (“knot”). Define its total energy: $$ E_V(t) := \int_V u(\mathbf{x},t)\,d^3x. $$ To talk about “where it is,” define its center-of-energy: $$ \mathbf{X}(t) := \frac{1}{E_V(t)}\int_V \mathbf{x}\,u(\mathbf{x},t)\,d^3x. $$ This is purely definitional. It introduces no mechanics. --- # 2. Energy change is boundary flux Differentiate $E_V(t)$ in time. If $V$ is fixed in space (for now): $$ \frac{dE_V}{dt}=\int_V \partial_t u\,d^3x. $$ Use continuity: $$ \frac{dE_V}{dt}=-\int_V \nabla\cdot\mathbf{S}\,d^3x. $$ Apply divergence theorem: $$ \frac{dE_V}{dt}=-\oint_{\partial V}\mathbf{S}\cdot d\mathbf{A}. $$ So: energy of the object changes only by energy crossing its boundary. This is the first “conservation law” and it is not assumed beyond the single continuity equation. --- # 3. Velocity of the center-of-energy from continuity Differentiate the numerator of $\mathbf{X}(t)$: $$ \frac{d}{dt}\int_V \mathbf{x}\,u\,d^3x = \int_V \mathbf{x}\,\partial_t u\,d^3x = -\int_V \mathbf{x}\,\nabla\cdot\mathbf{S}\,d^3x. $$ Now use the identity (componentwise): $$ x_i\,\partial_j S_j = \partial_j(x_i S_j) - S_i. $$ Integrate over $V$: $$ -\int_V x_i\,\partial_j S_j\,d^3x = -\int_V \partial_j(x_i S_j)\,d^3x + \int_V S_i\,d^3x = -\oint_{\partial V} x_i\,\mathbf{S}\cdot d\mathbf{A} + \int_V S_i\,d^3x. $$ Thus, in vector form: $$ \frac{d}{dt}\int_V \mathbf{x}\,u\,d^3x = \int_V \mathbf{S}\,d^3x - \oint_{\partial V} \mathbf{x}\,(\mathbf{S}\cdot d\mathbf{A}). $$ Now expand: $$ \frac{d}{dt}\left(E_V \mathbf{X}\right) = \int_V \mathbf{S}\,d^3x - \oint_{\partial V} \mathbf{x}\,(\mathbf{S}\cdot d\mathbf{A}). $$ Use product rule: $$ E_V \dot{\mathbf{X}} + \dot{E}_V\,\mathbf{X} = \int_V \mathbf{S}\,d^3x - \oint_{\partial V} \mathbf{x}\,(\mathbf{S}\cdot d\mathbf{A}). $$ Substitute $\dot{E}_V = -\oint_{\partial V}\mathbf{S}\cdot d\mathbf{A}$: $$ E_V \dot{\mathbf{X}} = \int_V \mathbf{S}\,d^3x - \oint_{\partial V} (\mathbf{x}-\mathbf{X})\,(\mathbf{S}\cdot d\mathbf{A}). $$ This is an exact identity derived from continuity alone. Interpretation: - $\int_V \mathbf{S}\,d^3x$ is the bulk transport tendency of energy within $V$. - The boundary term corrects for energy leaking in/out at different positions relative to $\mathbf{X}$. Now take the “closed object” regime: Assume the knot is self-sustaining and well-separated so that there is negligible net flux through $\partial V$: $$ \oint_{\partial V}\mathbf{S}\cdot d\mathbf{A}\approx 0, \qquad \oint_{\partial V} (\mathbf{x}-\mathbf{X})(\mathbf{S}\cdot d\mathbf{A})\approx 0. $$ Then the identity reduces to: $$ E_V \dot{\mathbf{X}} \approx \int_V \mathbf{S}\,d^3x. $$ So the center-of-energy velocity is: $$ \dot{\mathbf{X}} \approx \frac{1}{E_V}\int_V \mathbf{S}\,d^3x. $$ This is the Newtonian-looking statement “velocity is total flux divided by total content,” but it is an identity from continuity plus localization. No “inertia postulate” occurred. --- # 4. Momentum as a derived bookkeeping variable Define a derived “momentum-like” quantity: $$ \mathbf{P}(t) := \frac{1}{c^2}\int_V \mathbf{S}\,d^3x. $$ This is motivated by Maxwell kinematics (when applicable), but here it can be taken as a definition: a rescaled integrated energy flux. Combine with the previous relation: $$ \dot{\mathbf{X}} \approx \frac{c^2}{E_V}\mathbf{P}. $$ If the knot has approximately constant total energy $E_V\approx E$, define an effective mass: $$ m := \frac{E}{c^2}. $$ Then: $$ \mathbf{P} \approx m\,\dot{\mathbf{X}}. $$ This is “$p=mv$” as an emergent relation: it is a definition of $m$ from energy, plus the continuity-derived expression for $\dot{\mathbf{X}}$. Nothing mechanical has been assumed. --- # 5. Acceleration arises from *flux exchange* across the boundary Now differentiate $\mathbf{P}$: $$ \dot{\mathbf{P}}=\frac{1}{c^2}\frac{d}{dt}\int_V \mathbf{S}\,d^3x = \frac{1}{c^2}\int_V \partial_t \mathbf{S}\,d^3x, $$ if $V$ is fixed. (If $V$ moves with the knot, add the standard Reynolds transport terms; the conclusion is the same: boundary flux terms appear.) At this stage, continuity alone does not tell you $\partial_t \mathbf{S}$. So how do we proceed without “adding mechanics”? The program’s move is: - identify that *any* change of the knot’s net transport must come from exchange with the outside through its boundary, - and therefore define “force” as the boundary flux of transport content. To do this cleanly, introduce a symmetric second-rank flux object $\mathbf{T}$ such that $$ \partial_t\left(\frac{\mathbf{S}}{c^2}\right)+\nabla\cdot \mathbf{T}=0 $$ inside regions where there is no external exchange. This is not a new postulate if you are in Maxwell theory: $\mathbf{T}$ is the Maxwell stress tensor and the identity is a theorem. But even in a continuity-first setting, the structural point is: - if a conserved density exists, it has an associated flux. - energy has density $u$ and flux $\mathbf{S}$. - transport-content $\mathbf{S}/c^2$ plays the role of another density and therefore has a flux $\mathbf{T}$. Assuming such a local balance law holds for the transport-content inside the closed region, integrate: $$ \frac{d}{dt}\int_V \frac{\mathbf{S}}{c^2}\,d^3x = -\int_V \nabla\cdot\mathbf{T}\,d^3x = -\oint_{\partial V}\mathbf{T}\cdot d\mathbf{A}. $$ Thus: $$ \dot{\mathbf{P}} = -\oint_{\partial V}\mathbf{T}\cdot d\mathbf{A}. $$ Define the right-hand side as the net external action on the knot: $$ \mathbf{F}_{\mathrm{ext}} := -\oint_{\partial V}\mathbf{T}\cdot d\mathbf{A}. $$ Then we have: $$ \dot{\mathbf{P}} = \mathbf{F}_{\mathrm{ext}}. $$ This is Newton’s “second law” in its most honest form: - it is not a postulate about forces, - it is a definition of force as boundary flux of transport-content, - and it becomes a theorem once $\mathbf{T}$ is specified (in Maxwell, it is). If additionally $\mathbf{P}\approx m\dot{\mathbf{X}}$ and $m$ is constant, then: $$ m\,\ddot{\mathbf{X}} = \mathbf{F}_{\mathrm{ext}}. $$ So “$F=ma$” is a derived identity: acceleration is caused by net boundary flux of stress-like transport. --- # 6. Newton’s third law as action–reaction from flux continuity Consider two disjoint localized knots in volumes $V_1$ and $V_2$. Let $V=V_1\cup V_2$ enclose both. If the combined region is closed (no net flux through its outer boundary), then: $$ \dot{\mathbf{P}}_{\mathrm{total}} = 0. $$ But $$ \mathbf{P}_{\mathrm{total}} = \mathbf{P}_1+\mathbf{P}_2, \qquad \dot{\mathbf{P}}_1=\mathbf{F}_{2\to 1},\quad \dot{\mathbf{P}}_2=\mathbf{F}_{1\to 2} $$ where each $\mathbf{F}$ is defined by the stress flux across the respective boundaries. Then $$ 0=\dot{\mathbf{P}}_1+\dot{\mathbf{P}}_2=\mathbf{F}_{2\to 1}+\mathbf{F}_{1\to 2}. $$ So $$ \mathbf{F}_{2\to 1}=-\mathbf{F}_{1\to 2}. $$ This is Newton’s third law, obtained from: - closure of the combined region, - and the fact that changes can only be mediated by boundary exchanges. No separate “third law axiom” is needed. --- # 7. Newton’s first law as the closed-knot limit If a knot is fully isolated so that the stress flux across its boundary vanishes: $$ \oint_{\partial V}\mathbf{T}\cdot d\mathbf{A}=0, $$ then $$ \dot{\mathbf{P}}=0. $$ If $m$ is constant and $\mathbf{P}=m\dot{\mathbf{X}}$, then: $$ \ddot{\mathbf{X}}=0. $$ This is inertia as the no-exchange limit. So “an object in motion stays in motion” is the statement: - if no boundary exchange occurs, integrated transport-content stays constant. --- # 8. What has and has not been assumed ### What was assumed - The continuity equation for energy (observed continuous transport accounting). - The existence of localized concentrations (knots) so one can define $E_V$ and $\mathbf{X}$. - For the Newton 2/3 pieces: the existence of a flux $\mathbf{T}$ for the transport-content density $\mathbf{S}/c^2$. ### What was not assumed - No primitive mass. - No primitive momentum. - No primitive force. - No primitive Newton laws. Mass, momentum, force appear as *derived bookkeeping constructs* for localized continuous transport. ### Where Maxwell enters If one now specializes to source-free Maxwell theory, then: - $u$ and $\mathbf{S}$ are given by field expressions, - $\mathbf{T}$ is the Maxwell stress tensor, - and the balance law for $\mathbf{S}/c^2$ is a theorem of Maxwell equations. So in a Maxwell universe, the above “Newton identities” are not just structural possibilities; they become explicit field-theoretic consequences. --- # 9. The core message of this worked example Newton-like mechanics is not an independent layer. For localized, self-sustaining concentrations of continuously transported energy: - “inertia” is the persistence of integrated flux content under zero boundary exchange, - “force” is the net boundary flux of stress-like transport, - “action–reaction” is closure of the combined region, - “mass” is $E/c^2$ once local transport is tied to a maximal rate. Everything is continuity plus localization plus boundary accounting.