{"id":"22c8a309-077a-47f0-9794-b5c292be7c3d","arxiv_id":"2506.07317","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The oblique-incidence skin temperature equals the perpendicular-incidence solution with depth and time stretched by one over the cosine of the refracted angle, and amplitude scaled by that cosine.","lead":"This paper derives a formula for how skin temperature rises when a millimeter-wave beam hits the skin at an angle, not just perpendicularly. The formula reduces an angled exposure to an equivalent straight-on exposure, which can simplify safety tests and infrared-camera measurements.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Angle-independent surface transmission fraction α is the load-bearing weak point: for p-polarized beams near the Brewster angle, the ordering in (44) and conclusion 4 can reverse.","rationale":"The reader's weakest-assumption analysis and my independent reading converge on the same concern: α is treated as a constant independent of incident angle, which enters at Eq. (14) and is then propagated through all the applied comparisons in Section 4. The central asymptotic result (28) is derived cleanly from the model, and the scaling law connecting the original and projected beams is mathematically sound under the stated model. The unstated angle-dependence of α is not a flaw in the asymptotic expansion itself; it is a flaw in the physical interpretation of the comparison theorems. Because the paper's two main applied conclusions—the ordering of surface temperatures in (44) and the ordering of activation times in (50)—are presented as analytical results about practical beam setups, this omission is consequential. The issue is concrete and testable: a simple Fresnel calculation for a planar skin-air interface will show whether α(θ1) can exceed α(0) by enough to reverse the inequalities. The reader's other concern, that assertions (A1) and (A2) are only verified numerically, is real but less load-bearing, since those monotonicity statements can be checked analytically and are not the source of the potential reversal. My recommendation is to keep the reader's CONDITIONAL verdict unchanged: the paper should state the angular dependence of α or restrict the comparison conclusions to cases where it is negligible.","tokens_in":19635,"tokens_out":5011,"duration_ms":42357,"concrete_test":"Compute the Fresnel power transmission coefficient for a planar skin-air interface with n_skin=1.4 (or the complex refractive index at 95 GHz) for both s- and p-polarization over θ1 ∈ [0°, 80°]. Insert the resulting α(θ1) into the three surface-temperature expressions behind (44): T_1v = α(0) P_i h(t), T_1 = α(θ1) cosθ1 P_i λ h(t/λ²), and T_2 = α(0) cosθ1 P_i h(t), with λ = cos(arcsin(sinθ1/1.4)). Check whether T_1v > T_1 > T_2 holds at representative nondimensional times such as t=1 and t=10. If p-polarization at θ1≈55° produces α(θ1) > α(0)/λ and reverses the left inequality, conclusion 4 is false as stated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The derivation of the leading-order solution (28) is internally consistent: the scaling (26)-(27) exactly maps (24) to the λ=1 case, and dropping the ε² lateral conduction terms in (19) is justified by ε≪1. The load-bearing weak point is the angle-independent transmission fraction α introduced in (14). The paper defines α as the fraction of incident beam power that passes through the skin surface and then uses the same numerical value for all three beams in Section 4.2: the original beam at θ1, the same intrinsic beam at θ1=0, and the projected beam at θ1=0. Inequalities (40)-(43), the surface-temperature ordering (44), and the activation-time ordering (50) all carry a common factor α through the comparison. But for a real skin-air interface, the Fresnel power transmission coefficient depends on θ1 and on polarization. For p-polarization with n_skin≈1.4, α(θ1) rises from about 0.972 at normal incidence to about 1 at the Brewster angle near 54°, then falls; for s-polarization it decreases monotonically. With p-polarization near the Brewster angle, the factor λh(t/λ²) that makes the oblique beam cooler than the normal beam in (42) is offset by α(θ1)>α(0), so T_surf(θ1)>T_surf(0) can occur, breaking conclusion 4. The same issue affects the comparison between the original beam and the projected beam, because the latter is evaluated with α(0) while the former should use α(θ1). The paper never flags this as a limitation or restricts the conclusions to s-polarization or small angles. This hidden assumption does not invalidate the formal asymptotic solution (28), but it is load-bearing for the paper's applied comparison claims.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies skin heating by an obliquely incident Gaussian millimeter-wave beam. Using the small ratio ε of electromagnetic penetration depth to lateral beam scale, the authors nondimensionalize the heat equation, drop O(ε²) lateral conduction and O(ε) lateral beam drift at leading order, and separate variables to obtain Eq. (28): the oblique-beam temperature is expressed as P_d^a f(x,y) λ U^(0)(z/λ, t/λ²), where λ = cos θ2 and U^(0) is the known normal-incidence solution from the authors' earlier work. They then derive scaling laws for the three-dimensional temperature, the skin surface temperature, and the activated skin volume, and compare three beam configurations: the original oblique beam, the same intrinsic beam at normal incidence, and the projected beam at normal incidence.","tokens_in":19928,"tokens_out":13991,"duration_ms":98714,"significance":"If the result holds, it reduces a genuinely three-dimensional oblique-incidence heating problem to the known one-dimensional normal-incidence solution through an exact scaling in depth, time, and amplitude, with no fitted parameters. This is a useful and clean contribution for millimeter-wave dosimetry. The central asymptotic derivation is internally consistent: the scaling (26)-(27) exactly maps the λ case to the λ=1 case, and the size of ε is realistically tiny for 95 GHz beams with centimeter-scale spots. The paper's main weakness is not in the asymptotic solution itself but in the comparative conclusions, which depend on treating the surface transmission fraction α as an angle-independent constant.","major_comments":[{"comment":"The surface transmission fraction α is introduced as a single constant independent of incidence angle and polarization, and the same value is used for beams at θ1 and at θ1=0 in the comparisons of Section 4.2. For a real skin-air interface, however, the Fresnel power transmission coefficient depends on both. This is load-bearing: for s-polarization at large incidence angles the transmission coefficient drops substantially (for n_skin≈1.4 at θ1≈85°, α(θ1)/α(0) is roughly 0.5), so the right inequality in (44), comparing the original oblique beam with the projected perpendicular beam, becomes α(θ1)λh(t/λ²) > α(0)h(t). With λ≈0.70 and λh(t/λ²)/h(t)≈1/λ at small t, this ordering is reversed. The left inequality in (44) appears robust because of the additional cosθ1 factor in P_d^a, but the conclusion as stated in conclusion 4, and the activation-time ordering (50), are not valid for arbitrary incident angle and polarization. The fix is to carry α(θ1) explicitly through the derivation or to state the polarization and angle restrictions under which the inequalities hold; the asymptotic solution (28) itself remains valid if α is interpreted as angle-dependent.","section":"§2.3, Eq. (14); §4.2, Eqs. (42)-(44) and (50)"},{"comment":"The surface-temperature ordering (44) and the activation-time ordering (50) rest on assertions (A1) and (A2), but the text says only that they 'can be derived analytically' and then demonstrates them numerically in Figure 6. Since these monotonicity statements are load-bearing for the paper's main comparative claims, the analytic proof should be included or a reference supplied. The explicit formula (34) and the small- and large-time expansions (37) make this a short exercise, but as written the comparison is not fully analytic.","section":"§4.2, assertions (A1)-(A2), Eqs. (40)-(43)"}],"minor_comments":[{"comment":"The sentence 'We express (σξ, ση, ϕ2) in terms of (σξ, ση, ϕ2)' appears to be a typo; the second set should presumably be (σ1, σ2, ϕ, θ1).","section":"§2.2, after Eq. (9)"},{"comment":"There are two typos: 'nondimensional systrem' should be 'nondimensional system', and 'when the incident angle and/or the specific beam spot geometry are varies' should be 'are varied'.","section":"§3.1"},{"comment":"'Appying the Beer-Lambert Law' should be 'Applying the Beer-Lambert Law'.","section":"§2.3"},{"comment":"The labels in Figure 5 such as 't62' are unclear and appear to be a rendering artifact; they should be typeset as t/λ² to match the notation in Eq. (32).","section":"Figure 5"},{"comment":"It would help to state explicitly that the asymptotics in (37) are for fixed λ and are used only in the discussion of the estimation formulas (38) and (39); the present wording is slightly ambiguous about the t-range.","section":"§4.2, Eq. (37)"}],"recommendation":"major_revision","confidential_remarks":"The paper is well within the scope of the journal and the central asymptotic derivation is sound. The stress-test concern about α is partially valid: the left inequality in (44) appears protected by the cosθ1 factor, but the right inequality can reverse at large angles for s-polarization, so the comparative conclusions need to be revised. The missing analytic proofs of (A1) and (A2) should also be supplied. These are fixable within the manuscript's scope, hence major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a real extension of the authors' earlier normal-incidence work, not a repackaging. For arbitrary incident angle they derive a leading-order solution by a lambda-scaling of the lambda=1 solution, get a clean mapping between the oblique beam and the projected perpendicular beam, and pull out scaling laws for surface temperature and activation volume. The projection geometry in (9) is new and useful. The math is internally consistent; dropping O(epsilon^2) lateral conduction is justified for epsilon around 0.003, and the separation of variables is clean. Credit where due: the final solution is expressed in terms of a parameter-free U^(0), and the comparison to the prior lambda=1 case is a genuine derivation, not a fit.\n\nThe main soft spot is the surface transmission fraction alpha in (14). It is used as a single constant for all beams, including comparisons at different incident angles in (44) and (50). Fresnel transmission depends on angle and polarization. For p-polarized illumination near Brewster's angle (about 54 degrees for n_skin near 1.4), alpha(theta1) is meaningfully larger than alpha(0), and the ordering in (44) can reverse. The paper never flags this. It does not invalidate the formal asymptotic solution (28), since alpha only multiplies both sides, but it is load-bearing for the applied conclusions about which beam heats the surface more. The fix is straightforward: state the assumption, restrict conclusions to s-polarization or small angles, or carry the angular dependence through the inequalities.\n\nSecond soft spot: assertions (A1) and (A2) are said to be analytically derivable but only demonstrated numerically. Given the rest of the paper is analytical, providing the proofs (or at least a clear statement that they are unproven conjectures) should be a condition of acceptance. Also the reviewer should check whether h(t) indeed satisfies both monotonicities; the paper's own figure looks convincing, but a proof is better.\n\nThe citation to the authors' own [19] is appropriate—that is the parameter-free normal-incidence solution the scaling builds on, and it is an analytical result, not a hidden input. No circularity.\n\nBottom line: for anyone working on MMW skin-heating dosimetry or experiment design, this gives a usable conversion between oblique and perpendicular exposures. It deserves a serious referee, and I would recommend acceptance after revision on the two points above.","headline":"The oblique-incidence asymptotic solution and scaling law are clean and worth publishing; the angle-independent transmission coefficient threatens the applied comparison claims, not the core math.","tokens_in":20491,"tokens_out":1762,"would_cite":true,"duration_ms":15887,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35B40","80A20"],"pacs":[],"model":"deepseek-v4-flash","headline":"An angled millimeter-wave beam heats skin exactly like the perpendicular projected beam after rescaling depth by $\\cos\\theta_2$, time by $\\cos^2\\theta_2$, and amplitude by $\\cos\\theta_2$.","keywords":["electromagnetic heating","skin tissue","incident angle","asymptotic solution","scaling laws","activated skin volume","millimeter-wave","Gaussian beam"],"falsifier":"Measure the beam-center surface temperature for the same intrinsic beam at two incident angles $\\theta_1$ and $0$ with an infrared camera, and compute $R = T_{\\mathrm{surf}}(t;\\theta_1)\\big/\\left[\\cos\\theta_2\\,T_{\\mathrm{surf}}(t/\\cos^2\\theta_2;0)\\right]$ using the model's own Snell angle $\\theta_2$ and the same $\\alpha$ for both. If $R$ differs from 1 by more than the model's $O(\\varepsilon)$ error over the reported time range, the leading-order rescaling law fails; if the discrepancy tracks an independently measured $\\alpha(\\theta_1)/\\alpha(0)$, then the constant-$\\alpha$ assumption is the failing part.","tokens_in":19354,"feed_emoji":"🔥","tokens_out":9030,"duration_ms":71091,"temperature":0.7,"pith_summary":"The paper derives a leading-order formula for the three-dimensional temperature rise in skin exposed to a millimeter-wave beam hitting the surface at an arbitrary angle. Its central claim is that the angled-beam temperature at time $t$ equals the temperature of a perpendicular reference beam evaluated at depth $z/\\cos\\theta_2$ and time $t/\\cos^2\\theta_2$, multiplied by $\\cos\\theta_2$, where $\\theta_2$ is the refracted angle inside the tissue. Because the argument rests on the large separation between penetration depth and lateral beam size, not on the angle being small, the formula is claimed to hold for any incident angle. This matters for practical exposure assessment: it turns angled exposures into one parameter-free solution, and it yields direct scaling rules for surface temperature, absorbed-power estimation, and the volume where heat-sensitive nerve endings are activated.","feed_headline":"One rescaling law predicts skin heat at any beam angle","feed_subtitle":"Depth, time, and amplitude all rescale by cos θ2, so a single perpendicular solution covers every incident angle.","key_machinery":"The load-bearing object is the rescaled depth-time map defined by Eqs. (26)-(27): the angled-beam depth profile $W^{(0)}(z,t;\\lambda)$ satisfies the same one-dimensional heat problem as the normal-incidence profile $U^{(0)}$ after the substitution $(z,t)\\to(z/\\lambda,t/\\lambda^2)$, and the two are related by $W^{(0)}(z,t;\\lambda)=\\lambda\\,U^{(0)}(z/\\lambda,t/\\lambda^2)$. This identity reduces the angled problem to the previously solved normal-incidence case. It works because the heat equation is linear, the lateral derivatives are suppressed at leading order by $\\varepsilon\\ll 1$, and the heat source separates into $f(x,y)$ times an exponential in $z$; the refracted angle only enters through $\\lambda=\\cos\\theta_2$ and the projected spot parameters. The paper's surface temperature results rest on the single-variable function $h(t)=\\operatorname{erfc}(\\sqrt{t})e^t-1+2\\sqrt{t/\\pi}$, which is $U^{(0)}$ evaluated at $z=0$.","core_discovery":"On the paper's own terms, the discovery is that obliquely incident millimeter-wave heating is not a new problem. The leading-order asymptotic solution of the heat equation, Eq. (28), separates as $T^{(0)} = P_d^a\\, f(x,y;\\sigma_\\xi,\\sigma_\\eta,\\phi_2)\\, \\lambda\\, U^{(0)}(z/\\lambda, t/\\lambda^2)$ with $\\lambda=\\cos\\theta_2$, $P_d^a = \\alpha\\cos\\theta_1 P_d^{(i)}$, and $U^{(0)}$ the parameter-free solution for normal incidence. Equivalently, the temperature field of an angled beam is obtained from the perpendicular projected beam by stretching depth by $\\lambda$, stretching time by $\\lambda^2$, and multiplying the whole field by $\\lambda$. The author asserts this is valid for arbitrary incident angle $\\theta_1$ because the only small parameter is $\\varepsilon$, the ratio of the sub-millimeter penetration depth to the multi-centimeter lateral beam scale; lateral heat conduction drops out at leading order. All later scaling laws, including surface-temperature ordering, activation-time ordering, and the activated-volume equivalence, are corollaries of this one identity.","pith_inferences":["The analysis assumes a single angle-independent surface transmission fraction $\\alpha$; if real Fresnel transmission varies with incidence angle and polarization, the ordering of activation times in Eq. (50) could reverse, and measuring $\\alpha(\\theta_1)$ for the relevant polarization would be the natural correction.","The rescaling identity only needs the heat source to factor into a lateral envelope times an exponential in depth, so the same argument should carry over to non-Gaussian beam profiles, with $f(x,y)$ replaced by the measured spot shape.","Because the equivalence maps angled exposure at time $t$ to perpendicular exposure at time $t/\\cos^2\\theta_2$, safety comparisons between different beam orientations could be reduced to comparing equivalent normal-incidence exposures at adjusted times."],"forward_implications":["At any fixed time, the beam-center surface temperature for an angled beam lies between the temperature of the same intrinsic beam at normal incidence (hotter) and the temperature of the projected beam at normal incidence (cooler).","The time to reach nociceptor activation is ordered the same way: the normal-incidence intrinsic beam activates first, the angled beam second, and the projected beam last.","Estimating absorbed power density from the early-time surface-temperature slope overestimates the true value by $1/\\cos\\theta_2$ if the incident angle is ignored, whereas the late-time estimator is angle-independent.","The activated skin volume of an angled beam at time $t$ equals that of a modified perpendicular beam at time $t/\\cos^2\\theta_2$, with power density scaled by $\\cos\\theta_2\\cos\\theta_1$ and intrinsic spot area scaled by $\\cos\\theta_2/\\cos\\theta_1$.","Lateral heat conduction is negligible at leading order, so the in-plane shape of the Gaussian spot enters only as a multiplicative factor $f(x,y)$, not through lateral diffusion."],"supporting_citations":[{"why":"supplies the parameter-free normal-incidence solution $U^{(0)}$ that the scaling identity extends.","marker":"[19]"},{"why":"supplies the previous nondimensional formulation of perpendicular-beam skin heating that this paper generalizes to arbitrary angles.","marker":"[9]"},{"why":"gives the millimeter-wave absorption coefficient and fixes the sub-millimeter penetration depth used throughout the analysis.","marker":"[14]"},{"why":"provides experimental millimeter-wave exposures with measured skin surface temperatures that motivate the beam-size range and the surface-temperature scalings.","marker":"[15]"},{"why":"defines the nociceptor activation threshold that sets the temperature scale and the activated-volume criterion.","marker":"[18]"},{"why":"supplies the refractive index of skin used in Snell's law to connect the refracted angle $\\theta_2$ to the incident angle $\\theta_1$.","marker":"[11]"}],"fun_headline_variants":["Cos θ2 rescaling unifies skin heating at any beam angle","One rescaling law covers all incidence angles for skin heat","Angled beams need only depth and time stretching by cos θ2","Skin heat from oblique beams: rescale to perpendicular case","Universal rescaling maps angled-beam skin heat to normal"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The fraction of beam power entering the skin, $\\alpha$, is treated as a single constant that does not change with incident angle or polarization.","fun_headline_variants_meta":{"raw":{"variants":["Cos θ2 rescaling unifies skin heating at any beam angle","One rescaling law covers all incidence angles for skin heat","Angled beams need only depth and time stretching by cos θ2","Skin heat from oblique beams: rescale to perpendicular case","Universal rescaling maps angled-beam skin heat to normal"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000375,"raw_usage":{"total_tokens":2023,"prompt_tokens":990,"completion_tokens":1033,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":606,"completion_tokens_details":{"reasoning_tokens":949}},"tokens_in":606,"tokens_out":1033,"duration_ms":9575,"temperature":1.0,"reasoning_tokens":949,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:35:49.952942+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the beam-center surface temperature for the same intrinsic beam at two incident angles $\\theta_1$ and $0$ with an infrared camera, and compute $R = T_{\\mathrm{surf}}(t;\\theta_1)\\big/\\left[\\cos\\theta_2\\,T_{\\mathrm{surf}}(t/\\cos^2\\theta_2;0)\\right]$ using the model's own Snell angle $\\theta_2$ and the same $\\alpha$ for both. If $R$ differs from 1 by more than the model's $O(\\varepsilon)$ error over the reported time range, the leading-order rescaling law fails; if the discrepancy tracks an independently measured $\\alpha(\\theta_1)/\\alpha(0)$, then the constant-$\\alpha$ assumption is the failing part.","supporting_citations":[{"cited_title":"and Zhou, H","cited_arxiv_id":null,"evidence_quote":"supplies the parameter-free normal-incidence solution $U^{(0)}$ that the scaling identity extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"gives the millimeter-wave absorption coefficient and fixes the sub-millimeter penetration depth used throughout the analysis."},{"cited_title":"and Beason, C.W","cited_arxiv_id":null,"evidence_quote":"provides experimental millimeter-wave exposures with measured skin surface temperatures that motivate the beam-size range and the surface-temperature scalings."},{"cited_title":"and Campbell, J,N","cited_arxiv_id":null,"evidence_quote":"defines the nociceptor activation threshold that sets the temperature scale and the activated-volume criterion."},{"cited_title":"and Parrish, J.A","cited_arxiv_id":null,"evidence_quote":"supplies the refractive index of skin used in Snell's law to connect the refracted angle $\\theta_2$ to the incident angle $\\theta_1$."}],"review_version":1}