One Ratio, Three Regimes: How Faster Adjustment Turns Damping Links into Amplifying Ones

Many systems pass work, goods or information along a chain of units. Each unit sees its own state with a delay T and adjusts with a time scale q. We ask what happens to a disturbance along the chain when q falls while T stays the same. For a unit that passes work on in proportion to its own delayed stock, the ratio r = T/q sets the regime of each link: for r <= 1/2 every nonzero frequency is attenuated; for 1/2 < r < pi/2 the unit is stable but low frequencies are amplified at each link; for r >= pi/2 the unit is not asymptotically stable. These thresholds are the classical ones of delayed car-following models; we apply them to flow chains. For a unit that reorders to restore its own stock (the anchoring-and-adjustment rule of the beer game) with lead time L > 0, slow fluctuations are amplified: |G|^2 = 1 + w^2 L (2q + beta L) + O(w^4); this does not hold at every frequency. At zero reading delay the unit is a known transfer function (Riddalls and Bennett, 2000); the new part is the reading delay, which does not enter the leading coefficient. Simulations agree with the gain formulas within 0.2% at the main settings, and within 2% across the regimes at small time steps; the Euler bias in the noise runs is largely explained by the exact discrete-time response of the scheme. In a chain of 60 links with noise of 10% of the mean flow, the variance stops growing after about 30 links; we report this as an observation in one setting, not as a general bound. The evidence is analysis and simulation; real data has not been used. The record contains the paper (PDF and Markdown), its four figures, and a code package with all scripts, their outputs and a reproduction check. Use of AI: AI agents (large language models) took part in drafting, coding, translation and review, together with the author. The author set the questions, reviewed the results, and takes responsibility for the paper.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-10-29
DOI
https://doi.org/10.5281/zenodo.23749812
Primary Topic
Supply Chain and Inventory Management
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

One Ratio, Three Regimes: How Faster Adjustment Turns Damping Links into Amplifying Ones

CK Hung
Zenodo (CERN European Organization for Nuclear Research)
Supply Chain and Inventory Management
article

One Ratio, Three Regimes: How Faster Adjustment Turns Damping Links into Amplifying Ones

CK Hung
article en

Abstract

Many systems pass work, goods or information along a chain of units. Each unit sees its own state with a delay T and adjusts with a time scale q. We ask what happens to a disturbance along the chain when q falls while T stays the same. For a unit that passes work on in proportion to its own delayed stock, the ratio r = T/q sets the regime of each link: for r <= 1/2 every nonzero frequency is attenuated; for 1/2 < r < pi/2 the unit is stable but low frequencies are amplified at each link; for r >= pi/2 the unit is not asymptotically stable. These thresholds are the classical ones of delayed car-following models; we apply them to flow chains. For a unit that reorders to restore its own stock (the anchoring-and-adjustment rule of the beer game) with lead time L > 0, slow fluctuations are amplified: |G|^2 = 1 + w^2 L (2q + beta L) + O(w^4); this does not hold at every frequency. At zero reading delay the unit is a known transfer function (Riddalls and Bennett, 2000); the new part is the reading delay, which does not enter the leading coefficient. Simulations agree with the gain formulas within 0.2% at the main settings, and within 2% across the regimes at small time steps; the Euler bias in the noise runs is largely explained by the exact discrete-time response of the scheme. In a chain of 60 links with noise of 10% of the mean flow, the variance stops growing after about 30 links; we report this as an observation in one setting, not as a general bound. The evidence is analysis and simulation; real data has not been used. The record contains the paper (PDF and Markdown), its four figures, and a code package with all scripts, their outputs and a reproduction check. Use of AI: AI agents (large language models) took part in drafting, coding, translation and review, together with the author. The author set the questions, reviewed the results, and takes responsibility for the paper.

Zenodo (CERN European Organization for Nuclear Research)
Openalex Percentile: Top 6%
Supply Chain and Inventory Management
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon