Skip to content

Twisted reduction to three dimensions

A circle reduction does more than remove a coordinate. If the circle is fibered over space, momentum around it behaves as charge under a three-dimensional connection. Supersymmetry also ties that connection to the fixed auxiliary sources. We will derive both effects and see why the scalar combination z−rHz-rH appears in curved transformation rules.

Use the four-dimensional new minimal equation and the three-dimensional background dictionary. The parent matter theory has four-dimensional N=1\mathcal N=1 supersymmetry, four real Lorentzian Poincaré generators, all manifest. Its ordinary reduction has three-dimensional N=2\mathcal N=2, also four real generators. Euclidean Killing-spinor solutions are counted over C\mathbb C; neither changing dimension nor choosing independent Euclidean columns doubles the physical algebra. No dynamical gauge group or gauge rank is specified here.

This is the classical source and transformation-law construction of Appendix D. It is not an assertion that integrating out every nonzero circle mode leaves an unchanged quantum action.

The metric determines which components are horizontal

Section titled “The metric determines which components are horizontal”

Let xμx^\mu be three-dimensional coordinates and yy a length coordinate along the fiber. Start with

ds42=gμνdxμdxν+ϑ2,ϑ=dy+cμdxμ.(B03.1)ds_4^2=g_{\mu\nu}dx^\mu dx^\nu+\vartheta^2, \qquad \vartheta=dy+c_\mu dx^\mu . \tag{B03.1}

All metric coefficients are independent of yy. The fiber has fixed unit coefficient in this expression; if it is a circle, its radius is specified by y∼y+2πRy\sim y+2\pi R. We have not included a varying radius scalar.

The vectors dual to the base one-forms are horizontal lifts: Ea=eaμ(∂μ−cμ∂y)E_a=e_a{}^\mu(\partial_\mu-c_\mu\partial_y), while Ef=∂yE_f=\partial_y. Consequently a one-form decomposes as

W=Wμdxμ+Wy dy=(Wμ−Wycμ)dxμ+Wyϑ.(B03.2)W=W_\mu dx^\mu+W_y\,dy =(W_\mu-W_yc_\mu)dx^\mu+W_y\vartheta . \tag{B03.2}

Its horizontal component is Wμ−WycμW_\mu-W_yc_\mu, not its coordinate component WμW_\mu. For a vector, the corresponding vertical coefficient is Uy+cμUμU^y+c_\mu U^\mu. These two rules are compatible with contraction and raising indices. In particular det⁡G=det⁡g\det G=\det g.

Under the change of fiber coordinate y′=y−λ(x)y'=y-\lambda(x), c′=c+dλc'=c+d\lambda keeps ϑ\vartheta fixed. A mode

Φ(x,y)=eizyϕ(x)\Phi(x,y)=e^{izy}\phi(x)

then has ϕ′=eizλϕ\phi'=e^{iz\lambda}\phi. Applying a horizontal derivative gives

(∂μ−cμ∂y)Φ=eizy(∂μ−izcμ)ϕ.(B03.3)(\partial_\mu-c_\mu\partial_y)\Phi =e^{izy}(\partial_\mu-izc_\mu)\phi . \tag{B03.3}

Thus cc is the graviphoton connection, and fiber momentum zz becomes its charge. Here [z]=1[z]=1 and [cμ]=0[c_\mu]=0 when xμx^\mu has length units. For an ordinary periodic scalar on a circle, z=n/Rz=n/R with integer nn. A formal local ansatz permits a continuous parameter; a genuine compactification additionally requires the chosen global boundary conditions.

To keep the matrix conventions explicit, use the four-dimensional frame ordering of Appendix D: the fiber is direction 22, and the ordered base directions are (1,3,4)(1,3,4). This is an even permutation of the orientation used in the preceding lesson. Its positive Euclidean blocks give

γ1=τ3,γ2=−τ1,γ3=−τ2,γaγb=δab+iεabcγc.\gamma_1=\tau_3,\qquad \gamma_2=-\tau_1,\qquad \gamma_3=-\tau_2,\qquad \gamma_a\gamma_b=\delta_{ab}+i\varepsilon_{abc}\gamma_c .

These are exactly the three-dimensional matrices of the paper. Identify the opposite-chirality four-dimensional column by ζ~4=Sζ~\widetilde\zeta_4=S\widetilde\zeta, where S=iτ2S=i\tau_2, and leave the undotted column unchanged. This fixed index identification gives

Σab=i2εabcγc,Σaf=i2γa,S−1Σ~abS=i2εabcγc,S−1Σ~afS=−i2γa.(B03.4)\begin{aligned} \Sigma_{ab}&=\frac i2\varepsilon_{abc}\gamma_c, &\Sigma_{af}&=\frac i2\gamma_a,\\ S^{-1}\widetilde\Sigma_{ab}S&=\frac i2\varepsilon_{abc}\gamma_c, &S^{-1}\widetilde\Sigma_{af}S&=-\frac i2\gamma_a . \end{aligned} \tag{B03.4}

It also implements the dotted-to-undotted contraction convention in (D.2). This is a component dictionary; it does not rephase a pair of adjoint Lorentzian supercharges. Tilded Euclidean fields remain independent.

Put f=dc=12fabea∧ebf=dc=\tfrac12f_{ab}e^a\wedge e^b and

va=−i2εabcfbc,fab=iεabcvc.v^a=-\frac i2\varepsilon^{abc}f_{bc}, \qquad f_{ab}=i\varepsilon_{abc}v^c .

For the ordinary Cartan connection, deA+ΩAB∧eB=0de^A+\Omega^A{}_B\wedge e^B=0 gives

Ωab(4)=Ωab(3)−12fabϑ,Ωfa(4)=12fabeb.(B03.5)\Omega^{(4)}_{ab}=\Omega^{(3)}_{ab}-\frac12 f_{ab}\vartheta, \qquad \Omega^{(4)}_{fa}=\frac12 f_{ab}e^b . \tag{B03.5}

For example, the second formula cancels dϑ=fd\vartheta=f in Cartan’s equation; the first cancels its induced term in the equation for eae^a. The paper’s input-first connection has the opposite all-lowered sign, as explained in the prerequisite.

Insert (B03.4)–(B03.5) into ∇=∂+ΩABΣAB/2\nabla=\partial+\Omega_{AB}\Sigma^{AB}/2, summing both ordered frame pairs. For a fiber-independent label,

∇f(4)ζ=14vaγaζ,∇h,a(4)ζ=∇a(3)ζ+14εabcvbγcζ,∇f(4)ζ~=14vaγaζ~,∇h,a(4)ζ~=∇a(3)ζ~−14εabcvbγcζ~.(B03.6)\begin{aligned} \nabla_f^{(4)}\zeta&=\frac14v^a\gamma_a\zeta,& \nabla_{h,a}^{(4)}\zeta &=\nabla_a^{(3)}\zeta+\frac14\varepsilon_{abc}v^b\gamma_c\zeta,\\ \nabla_f^{(4)}\widetilde\zeta&=\frac14v^a\gamma_a\widetilde\zeta,& \nabla_{h,a}^{(4)}\widetilde\zeta &=\nabla_a^{(3)}\widetilde\zeta-\frac14\varepsilon_{abc}v^b\gamma_c\widetilde\zeta . \end{aligned} \tag{B03.6}

The subscript hh means the horizontal lift. A mode eizye^{izy} adds iziz to the vertical derivative and −izcμ-izc_\mu to the horizontal one. The two signs in the horizontal spin rotations are consequential.

Solve the vertical equation before reducing the horizontal one

Section titled “Solve the vertical equation before reducing the horizontal one”

Denote the four-dimensional background one-forms by A,V\mathcal A,\mathcal V and the final three-dimensional sources by A,V,H,CA,V,H,C. The first four-dimensional equation is

(∇M(4)−iAM)ζ=−iVMζ+iVNΣMNζ.(\nabla_M^{(4)}-i\mathcal A_M)\zeta =-i\mathcal V_M\zeta+i\mathcal V^N\Sigma_{MN}\zeta .

The background and supersymmetry labels have zero fiber charge. Using (B03.6), its vertical component becomes

14v⋅γ ζ=i(Af−Vf)ζ+12Vh⋅γ ζ.(B03.7)\frac14v\cdot\gamma\,\zeta =i(\mathcal A_f-\mathcal V_f)\zeta +\frac12\mathcal V_h\cdot\gamma\,\zeta . \tag{B03.7}

The opposite chirality reverses the sign of the scalar difference. To make both vertical equations identities on all spinor components, choose Af=Vf=H\mathcal A_f=\mathcal V_f=H and Vh=v/2\mathcal V_h=v/2. This is the ansatz used in the reduction. It is sufficient to avoid an extra vertical projection; it is not a necessary condition for one specially selected four-dimensional spinor.

Define the remaining source by Ah=A−v/2\mathcal A_h=A-v/2. The complete dictionary is

C=c,V=v=−i∗3dC,A=A−12V+Hϑ,V=12V+Hϑ.(B03.8)\boxed{ C=c,\qquad V=v=-i*_3dC,\qquad \mathcal A=A-\frac12V+H\vartheta,\qquad \mathcal V=\frac12V+H\vartheta .} \tag{B03.8}

The Hodge star uses the base metric and orientation. In particular, Aμ=Aμ−Hcμ+Vμ/2A_\mu=\mathcal A_\mu-Hc_\mu+V_\mu/2 in coordinate components. The four-dimensional divergence constraint reduces to ∇μ(3)Vμ=0\nabla_\mu^{(3)}V^\mu=0, which follows from d2C=0d^2C=0 locally. The four-dimensional two-form source B(4)B^{(4)} is distinct from the metric connection CC. In the preceding lesson’s orientation its field strength can be written dB(4)=−H vol3−i2 dC∧ϑdB^{(4)}=-H\,\mathrm{vol}_3-\tfrac i2\,dC\wedge\vartheta. Taking its four-dimensional dual with (B02.1) gives exactly V\mathcal V in (B03.8). Local existence does not settle global potential patches or any chosen compact period lattice.

The horizontal equation before removing the extra spin rotation is

∇h,a(4)ζ=i(Aa−Va)ζ−12Hγaζ−14εabcVbγcζ.\nabla_{h,a}^{(4)}\zeta =i(A_a-V_a)\zeta-\frac12H\gamma_a\zeta -\frac14\varepsilon_{abc}V^b\gamma_c\zeta .

Subtract the +14+\tfrac14 term in (B03.6). For the other chirality, subtract its −14-\tfrac14 term. The result is

(∇a(3)−iAa)ζ=−12Hγaζ−iVaζ−12εabcVbγcζ,(∇a(3)+iAa)ζ~=−12Hγaζ~+iVaζ~+12εabcVbγcζ~.(B03.9)\begin{aligned} (\nabla_a^{(3)}-iA_a)\zeta &=-\frac12H\gamma_a\zeta-iV_a\zeta -\frac12\varepsilon_{abc}V^b\gamma_c\zeta,\\ (\nabla_a^{(3)}+iA_a)\widetilde\zeta &=-\frac12H\gamma_a\widetilde\zeta+iV_a\widetilde\zeta +\frac12\varepsilon_{abc}V^b\gamma_c\widetilde\zeta . \end{aligned} \tag{B03.9}

This restores the step between the paper’s (D.13) and (D.15). The coefficient one-half comes partly from the four-dimensional source term and partly from the fiber’s spin connection.

For a positive real four-dimensional metric, cc is real, so v=−i∗3dcv=-i*_3dc is imaginary. General Euclidean three-dimensional backgrounds allow complex CC. Accordingly Appendix D first reduces the equations and then complexifies their sources. A complex CC need not be the connection of a real Riemannian circle fibration. Nor does a locally defined CC automatically provide global circle transition functions with an integral charge lattice.

A nontrivial fibration with explicit surviving labels

Section titled “A nontrivial fibration with explicit surviving labels”

Take a flat base with coordinates (x1,x2,x3)(x^1,x^2,x^3) and a real nonzero constant bb of mass dimension one:

c=bx1dx2,V=−ib dx3,H=b,A=−3ib2 dx3.(B03.10)c=b x^1dx^2,\qquad V=-ib\,dx^3,\qquad H=b,\qquad A=-\frac{3ib}{2}\,dx^3 . \tag{B03.10}

The four-dimensional metric is real and positive, ds42=∑a(dxa)2+(dy+bx1dx2)2ds_4^2=\sum_a(dx^a)^2+(dy+b x^1dx^2)^2. The sources (B03.8) are complex, as Euclidean rigid supersymmetry permits. For this example one smooth two-form potential is B(4)=−bx1dx2∧dx3−ib2x1dx2∧dyB^{(4)}=-b x^1dx^2\wedge dx^3-\tfrac{ib}{2}x^1dx^2\wedge dy; it remains a globally defined form if the fiber is a circle.

Choose constant columns with γ3u±=±u±\gamma_3u_\pm=\pm u_\pm: u±=(1,∓i)T/2u_\pm=(1,\mp i)^T/\sqrt2. For the positive-charge equation, the three right-hand-side matrices acting on ζ\zeta are

M1=−b2(γ1+iγ2),M2=−iM1,M3=b2(1−γ3).M_1=-\frac b2(\gamma_1+i\gamma_2),\qquad M_2=-iM_1,\qquad M_3=\frac b2(1-\gamma_3).

They all annihilate u+u_+. The opposite-charge matrices annihilate u−u_-. Thus ζ=u+\zeta=u_+ and ζ~=u−\widetilde\zeta=u_- solve the complete equations, including their vertical components. Since [M1,M3]=bM1[M_1,M_3]=bM_1 has one-dimensional kernel, the curvature condition permits at most the u+u_+ line in the first sector; the other sector similarly has one line. The solution dimensions in this fiber-independent sector are exactly (1,1)(1,1) on this connected background. If yy is a circle, choose its periodic spin structure so these constant components are global. No sphere geometry has been used.

An independent geometric check probes the same connection with an R-neutral scalar. Substituting Φ=eizyϕ\Phi=e^{izy}\phi into the ordinary four-dimensional Laplacian gives

−Δ4⟶−∂12−(∂2−izbx1)2−∂32+z2.(B03.11)-\Delta_4\longrightarrow -\partial_1^2-(\partial_2-izb x^1)^2-\partial_3^2+z^2 . \tag{B03.11}

For ϕ=eikx2+ipx3f(x1)\phi=e^{ikx^2+ipx^3}f(x^1) and zb≠0zb\ne0, this is a harmonic oscillator centered at x1=k/(zb)x^1=k/(zb). Its eigenvalues are (2n+1)∣zb∣+p2+z2(2n+1)|zb|+p^2+z^2, n=0,1,…n=0,1,\ldots. This checks both the horizontal charge and the vertical mass term. It is a scalar geometry probe, not the full supersymmetric matter kinetic operator or a localization determinant.

Let a four-dimensional general multiplet have bottom scalar XX of R-charge rr, spinors χ,χ~\chi,\widetilde\chi, scalars M,M~M,\widetilde M, vector component BM\mathcal B_M, spinors λ,λ~\lambda,\widetilde\lambda, and top auxiliary D\mathcal D. The vector component is not the background R connection. Give every component the same fiber dependence eizye^{izy}; the supersymmetry labels remain independent of yy.

The new three-dimensional components are

bμ=Bμ−cμBy,σ=By,D=D−Hσ.(B03.12)b_\mu=\mathcal B_\mu-c_\mu\mathcal B_y,\qquad \sigma=\mathcal B_y,\qquad D=\mathcal D-H\sigma . \tag{B03.12}

Other components retain their names, with the dotted-column identification above. Their R-charges are r−1,r+1r-1,r+1 for χ,χ~\chi,\widetilde\chi; r−2,r+2r-2,r+2 for M,M~M,\widetilde M; r+1,r−1r+1,r-1 for λ,λ~\lambda,\widetilde\lambda; and rr for b,σ,Db,\sigma,D. They all carry central charge zz.

For a component of R-charge qq, the four-dimensional R-covariant derivative reduces to

Dh(4)=∇(3)−iq(A−12V)−izC,Df(4)=i(z−qH).(B03.13)\begin{aligned} \mathscr D_h^{(4)} &=\nabla^{(3)}-iq\left(A-\frac12V\right)-izC,\\ \mathscr D_f^{(4)}&=i(z-qH). \end{aligned} \tag{B03.13}

For spinors add the horizontal and vertical spin rotations in (B03.6); for a scalar these extra terms are absent. This directly explains the matter connection A−V/2A-V/2 and the effective scalar z−rHz-rH. The separate R-current source A−3V/2A-3V/2 discussed in the prerequisite does not replace the matter connection.

For example, the undotted rule in (C.2) contains σPMζ~4(iBM+DMX)\sigma_P^M\widetilde\zeta_4(i\mathcal B_M+\mathscr D_MX). The matrices satisfy σP,hS=−γ\sigma_{P,h}S=-\gamma and σP,fS=i1\sigma_{P,f}S=i1. Its horizontal and vertical parts therefore give

δχ=ζM−γμζ~(DμX+ibμ)−ζ~ [σ+(z−rH)X],Dμ=∇μ(3)−ir(Aμ−Vμ/2)−izCμ.(B03.14)\begin{aligned} \delta\chi&=\zeta M -\gamma^\mu\widetilde\zeta(\mathscr D_\mu X+ib_\mu) -\widetilde\zeta\,[\sigma+(z-rH)X],\\ \mathscr D_\mu&=\nabla_\mu^{(3)} -ir(A_\mu-V_\mu/2)-izC_\mu . \end{aligned} \tag{B03.14}

The same index identification yields δX=iζχ+iζ~χ~\delta X=i\zeta\chi+i\widetilde\zeta\widetilde\chi and

δχ~=ζ~M~−γμζ(DμX−ibμ)−ζ [σ−(z−rH)X].\delta\widetilde\chi =\widetilde\zeta\widetilde M -\gamma^\mu\zeta(\mathscr D_\mu X-ib_\mu) -\zeta\,[\sigma-(z-rH)X].

Here bilinears use the antisymmetric contraction of the three-dimensional spinors. The sign change in the tilded term of δX\delta X includes the dotted contraction convention, not just a renamed component.

The auxiliary shift in (B03.12) is a definition needed to match the three-dimensional general-multiplet convention; it is not an elimination of an auxiliary by its equation of motion. The four-dimensional gaugino term iζDi\zeta\mathcal D becomes iζ(D+Hσ)i\zeta(D+H\sigma). Another useful check is the curvature combination in the top-component rule. In the paper’s Ricci convention,

R4=R3−12V2,V2=H2+14V2,R4−6V2=R3−2V2−6H2.R_4=R_3-\frac12V^2,\qquad \mathcal V^2=H^2+\frac14V^2,\qquad R_4-6\mathcal V^2=R_3-2V^2-6H^2 .

No complex conjugation occurs in these squares. These are precisely the combinations in the complete reduced rules (D.17)–(D.18).

Finally let s=ζζ~s=\zeta\widetilde\zeta and Kμ=ζγμζ~K^\mu=\zeta\gamma^\mu\widetilde\zeta. The same matrix map gives the four-dimensional bilinear vector as K4=−Kh+is ∂yK_4=-K^h+is\,\partial_y. Applying its R-covariant Lie derivative to a scalar mode reduces the algebra (C.3) to

{δζ,δζ~}ϕ=−2i[KμDμϕ+s(z−rH)ϕ].(B03.15)\{\delta_\zeta,\delta_{\widetilde\zeta}\}\phi =-2i\left[ K^\mu\mathscr D_\mu\phi+s(z-rH)\phi \right]. \tag{B03.15}

Thus the central transformation is the vertical translation in the chosen mode. Tensor and spinor fields additionally carry the appropriate Lie derivative on their indices; the scalar calculation already fixes the charge term and its sign.

1. A flat connection can still change the modes

Section titled “1. A flat connection can still change the modes”

Set all R and auxiliary sources to zero and take a flat base with x1∼x1+Lx^1\sim x^1+L, y∼y+2πRy\sim y+2\pi R, and c=β dx1c=\beta\,dx^1. For an ordinary periodic scalar, find the eigenvalue of −Δ4-\Delta_4 on exp⁡(2πimx1/L+iny/R)\exp(2\pi i m x^1/L+i n y/R), with no dependence on the other coordinates. When can a globally defined fiber coordinate change remove cc? Explain the corresponding relabeling of modes.

Hint 1

The horizontal momentum is 2πm/L−βn/R2\pi m/L-\beta n/R, even though dc=0dc=0.

Hint 2

The proposed coordinate shift y′=y+βx1y'=y+\beta x^1 must change by an integer multiple of the fiber circumference when x1x^1 completes its circle.

Solution

The eigenvalue is

(2πmL−βnR)2+n2R2.\left(\frac{2\pi m}{L}-\frac{\beta n}{R}\right)^2+\frac{n^2}{R^2}.

A single-valued fiber redefinition removes cc if βL=2πRw\beta L=2\pi Rw for an integer ww. Reexpressing the same function in y′=y+βx1y'=y+\beta x^1 gives m′=m−nwm'=m-nw and the usual untwisted eigenvalue (2πm′/L)2+n2/R2(2\pi m'/L)^2+n^2/R^2. For generic β\beta the local connection is flat but its global holonomy cannot be removed this way. The integer fiber charge, the flat connection and the radius all affect a genuine mode spectrum.

2. A vertical kernel is weaker than the reduction ansatz

Section titled “2. A vertical kernel is weaker than the reduction ansatz”

Retain a given fibration but change Af−Vf=δ\mathcal A_f-\mathcal V_f=\delta and Vh=v/2+u\mathcal V_h=v/2+u, with complex scalar δ\delta and vector uu. Derive the two vertical constraints. For δ≠0\delta\ne0 and u=2iδ e3u=2i\delta\,e^3, determine their kernels. Have you thereby proved that this changed background preserves supersymmetry?

Hint 1

Cancel the common v⋅γ/4v\cdot\gamma/4 term on both sides of the vertical equation.

Hint 2

The remaining matrices are linear combinations of 11 and γ3\gamma_3. The horizontal differential equations still have to be solved.

Solution

The constraints are (iδ+u⋅γ/2)ζ=0(i\delta+u\cdot\gamma/2)\zeta=0 and (−iδ+u⋅γ/2)ζ~=0(-i\delta+u\cdot\gamma/2)\widetilde\zeta=0. Both determinants equal −δ2−u2/4-\delta^2-u^2/4. With the assigned complex uu, these become iδ(1+γ3)ζ=0i\delta(1+\gamma_3)\zeta=0 and iδ(−1+γ3)ζ~=0i\delta(-1+\gamma_3)\widetilde\zeta=0. Their respective kernels have γ3=−1\gamma_3=-1 and +1+1.

The vertical equations now permit only lines, instead of being identities on both components. This is a necessary local test, not a solution of the horizontal equations or their global holonomy; the changed sources must also satisfy their divergence constraint. In particular it does not establish surviving supercharges. Setting δ=u=0\delta=u=0 was a deliberate way to avoid imposing such an extra projection, not a classification of all possible reductions.

Reduce the preceding lesson’s Sℓ3×S1S^3_\ell\times S^1 background: c=0c=0, A=V=−i dy/ℓ\mathcal A=\mathcal V=-i\,dy/\ell. Consider an Abelian vector multiplet with constant vertical component By=σ\mathcal B_y=\sigma, vanishing horizontal gauge field and gauginos, and zero central charge. Determine the three-dimensional sources and the auxiliary needed to make both gaugino variations vanish. Why is simply setting the three-dimensional auxiliary to zero incorrect?

Hint 1

The R-neutral vector multiplet has iζD=iζ(D+Hσ)i\zeta\mathcal D=i\zeta(D+H\sigma) in its gaugino variation.

Hint 2

All field-strength and derivative terms vanish for this configuration. Use both signs of the tilded and untilded variations.

Solution

The sources are A=V=C=0A=V=C=0, H=−i/ℓH=-i/\ell. The four-dimensional gauge field has zero field strength, so the gaugino equations give D=0\mathcal D=0. In three-dimensional variables this is

D=−Hσ=iσℓ.D=-H\sigma=\frac{i\sigma}{\ell}.

Indeed δλ=iζ(D+Hσ)\delta\lambda=i\zeta(D+H\sigma) and δλ~=−iζ~(D+Hσ)\delta\widetilde\lambda=-i\widetilde\zeta(D+H\sigma) both vanish. Setting D=0D=0 would fail for nonzero σ\sigma. These equations select a supersymmetric fixed configuration; they are not the matter theory’s auxiliary equation of motion. The imaginary Euclidean value is consistent with the source convention and does not prescribe a path-integral contour for an unspecified theory.

Ordinary untwisted reduction is recovered when the fibration and relevant source shifts vanish. The more general construction preserves a specified fiber-invariant set of supersymmetry labels, assigns consistent charges to the matter modes, and reorganizes their transformations. Interactions must respect those charge assignments. A finite-radius quantum compactification also has a tower of modes, threshold effects and global data; none is computed merely by this local dictionary.

The Appendices C/D paper guide maps every reduction equation, tracks each component charge, and tests what changes when the fiber radius varies.