DescriptionThis preprint proposes a minimal “three-layer 137” closure based on a uniform, compressible mechanical ether. A Bernoulli-driven thin-shell structure naturally introduces a hierarchy N\sim 137, and the same geometric–mechanical rules are used to close local constants and cosmological scales within one consistent framework. The guiding principle is that most physical systems are globally uniform and locally random, leading to a universal mixture of coherent modes (sin/exp) and statistical Gaussian attractors (normal). All interactions are treated as contact-mediated processes through ether dynamics rather than action-at-a-distance. The laboratory layer is anchored byN_3 \equiv \alpha^{-1},from which the electron shell radius R_e, thickness \delta, rotational speed v_3, endpoint force F_{\rm end}, ether density \rho, effective modulus B_{\rm eff}, and viscosity \eta are derived through Bernoulli closure. Gravity is closed via a 1/N_3^3 leakage mechanism with a lever-arm correction, allowing the endpoint cosmic acceleration a_{\rm end} to be inferred from G_{\rm lab}. A shockwave phase engine then links a_{\rm end} to the Hubble horizon R_H and the Hubble time. A key observational bookkeeping is adopted: acceleration is integrated over [\theta,2\pi] while time is integrated over [0,\theta], reflecting the radial distance from the explosion center to the current phase position. This defines \kappa(\theta) and yieldsR_H^{(0)}=\frac{c^2}{a_{\rm end}},\qquadR_H^{(1)}=\frac{c^2}{\kappa(\theta)a_{\rm end}}.For \theta\approx 185^\circ, \kappa(\theta) deviates from unity by <2%, supporting “endpoint ≈ average” as a controlled approximation. The model is effectively governed by two core parameters, while a third quantity serves as a nuisance perturbation associated with shock-profile or geometric projection. A table of ~60 constants is provided, separating definition anchors, experimental anchors, and derived quantities. Under a consistent reference convention, most derived constants reach zero-error agreement, with a few ~2 ppb shifts attributable to different \alpha^{-1} conventions (e.g., 084 vs 177) and power-law amplification. 本文提出一种以**均匀可压缩机械以太(ether)**为底层介质的三层 137 闭合框架:由介质波速、伯努利压差与薄壳结构给出壳厚比与层级计数 N\sim 137,并在同一套“几何—介质—泄露”规则下,把微观基本常数与宏观宇宙量联立闭合。 本工作强调一个极简思想:真实世界的大部分区域呈“总体均匀、局部随机”,因此宏观扰动自然表现为**相干模态(sin/exp)与统计吸引子(normal)**的叠加;四种力的表现最终都应回到介质接触传递与伯努利梯度的统一解释,而不是“非接触的玄学力”。 核心闭合链(本地常数) 以实验室层固定为 N_3 \equiv \alpha^{-1} 并由电子壳几何与伯努利动压建立本地参数包:电子真实壳半径 R_e、壳厚 \delta、旋转面速度 v_3、端点力 F_{\rm end}、以太密度 \rho、等效模量 B_{\rm eff}、粘度 \eta 等可由同一链条推出。随后引力常数通过 1/N_3^3 级泄露与力臂修正闭合,从而可以反求宇宙端点加速度 a_{\rm end},并导出哈勃半径与宇宙年龄量级。 宏观相位引擎(宇宙学) 宇宙在本文中被描述为一次球对称爆发形成的冲击波剖面,“密部压缩 + 疏部拉伸”的差力对应暗能量效应。关键的观测积分口径为: • **加速度积分:**相位区间 [\theta,2\pi](波后疏部重排主导) • **时间积分:**相位区间 [0,\theta](观测距离为爆炸中心到当前相位位置) 从而引入平均—端点映射因子 \kappa(\theta),并给出两级哈勃视界: R_H^{(0)}=\frac{c^2}{a_{\rm end}},\qquad R_H^{(1)}=\frac{c^2}{\bar a(\theta)}=\frac{c^2}{\kappa(\theta)a_{\rm end}} 在本文采用的参数区间(\theta\approx 185^\circ)下,\kappa(\theta) 的偏离被约束在 <2%,因此“端点≈平均”可作为快速估算,而一级表达给出可复算精度。 两参数与“打扰量” 本文强调:若以太的微观性质足够单纯,则宏观引力(或等效压强尺度)几乎决定一切常数。在三层闭合里,主导可归结为两个关键参数决定整体尺度与闭合结构;第三个量仅作为冲击波剖面/几何投影引入的打扰项(nuisance),用于刻画观测位置或剖面形状的微偏离。 常数表与“零误差” 文末给出一张包含约 60 项的常数总表,区分: • 定义锚点(D):SI 精确常数(误差不统计) • 实验锚点(I):作为输入锚点锁定 • 物理推导(P):由本模型闭合推出并与参考值对照 多数派生量在统一口径下可达到**“零误差”一致性**;少数出现约 ~2 ppb 的系统差异,来源于 \alpha^{-1} 不同版本(如 084 与 177)的幂次放大效应,而非推导链条本身的不自洽。