scientific-figure-studio
UnverifiedFigureCraft 根据科学对象、关系和数据制作科研图与语义配色。适用于机制、硬件、分层材料、定量图及整套配图,交付可编辑图源和分别记录的技术、科学与视觉审阅,保留数值、拓扑和修改范围。
examples/bin-latch/input/draft.md
View on GitHub# Loose manuscript notes: removable retention at a returnable-bin corner
**DEMO FICTION. Every device, observation, specimen count, and number below is invented for a writing exercise. The values are preset constructions, not measurements or simulation outputs. There are no real references.**
Returnable bins are often opened for cleaning, and their wall corners need to stay together when the wall is pulled outward. We considered a small removable retention unit near the upper rim of one corner. The corner housing joins two perpendicular wall edges, but only one edge is detachable in this exercise. The other edge is permanently secured to the housing. Releasing a full loaded bin was not a requirement. The intended operation is to empty the bin, unload its wall, and then release the detachable edge. Dirt at a joint is relevant because an opening that is easy to move on a clean bench can become a particle trap.
Four configurations were assigned letters. A was a flexible polymer snap key whose tooth holds the wall tab. B was a removable steel pin crossing the housing and a hole in the tab; its small retaining loop was included in the part count. C was a thicker snap key, with a shorter bending region than A. D used a separate sliding shoe and an external release sled. The common outer housing envelope was 28 by 22 by 18 millimetres. Internal pockets differed, so these are matched-envelope alternatives rather than interchangeable inserts in one identical housing. The detachable polypropylene wall tab was 2.8 millimetres thick. This is probably better placed before the testing paragraph.
For D the tab enters along positive x and is pulled out along negative x. The tab lies in the x-z plane; y is its thickness direction. The shoe tooth enters its window along negative y. An x-facing surface of the shoe rests against a fixed housing stop when the tab is pulled. The release sled translates upward along positive z. A ramp on that sled pushes the shoe along positive y so that the tooth leaves the window. A thin return tongue biases the shoe back toward engagement. The tongue and shoe are one polymer part. With no sled displacement, tooth overlap is 1.0 millimetre. Three millimetres of sled travel produces 1.2 millimetres of transverse shoe travel. The stop constrains motion along x, while the guide permits motion along y.
The earliest D pocket had only 0.15 millimetre nominal clearance around the moving shoe. An intentionally placed 0.20 millimetre grain blocked its motion in a scratch trial. This scratch trial was outside the comparison set and is not another replicated result. The revised pocket used 0.45 millimetre nominal side clearance and a slot open through its underside. The slot was 1.2 millimetres wide. There is no claim that all particles leave this slot; no particle flow or removal fraction was recorded. The extra clearance also permitted more lateral shoe play. A minimum remaining tooth overlap of 0.7 millimetre at the adverse tolerance position was required in the dimensional check. No statistical tolerance analysis was performed.
The comparison data in the accompanying CSV describe two conditions per configuration, clean and dosed with grit. Each cell has twenty release fixtures and six different pullout fixtures. There are 208 fictional fixtures altogether, with no reuse between cells or measurement groups. All fixtures first undergo 300 tensile load cycles between 20 and 120 newtons while remaining engaged. The cycles are not repeated assembly and disassembly. Grit means dry spherical glass granules between 150 and 300 micrometres in diameter. A 0.15 gram dose is applied before insertion and again after cycles 100 and 200, for 0.45 gram total. Nothing is rinsed or blown out. This is a narrow artificial contamination condition.
At the end of cycling, the wall is unloaded to below two newtons for the release test. Each fixture receives one actuator attempt, with a sixty newton force ceiling and ten millimetres maximum actuator travel. The motion follows the mechanism's intended release direction. For B the actuator withdraws the pin; for D it moves the sled. Success means that the locking member fully clears the wall tab. The release force reported in the table is the median peak force among successful attempts only. Failed attempts are retained in the success count. Separate fixtures are then pulled apart at five millimetres per minute to obtain retention strength. Those medians describe destructive pullout after cycling, not operating loads.
I first noticed the 405 newton clean median for C, which is the largest retention value. In grit the corresponding value is 391 newtons, but only eight of twenty release attempts succeed. D reaches 318 newtons clean and 306 with grit; its success counts are twenty and eighteen. A has twenty clean successes and fourteen with grit. B has twenty clean successes and eleven with grit. Among the successful gritty releases, B needs a median thirteen newtons and D eighteen. Thus the force statistic alone misses some stuck units, and D does not minimize successful release force. The ranges and counts should stay with these numbers when they are condensed.
The components were specified as machined stock for this exercise. Polymer properties were not measured, and no moulded samples were included. D weighs 12.6 grams for the complete retention hardware, including its housing, compared with 8.4 grams for A. The wall tab is excluded from every hardware mass. D has three discrete parts, whereas A and C have two and B has three. More machining and registration during assembly are needed for D. No assembly time or price was collected. These costs should not disappear from a revised account. The simple snap also has a clean successful-release force of nineteen newtons, so an argument for every application would go beyond these data.
The tab-window edge tore in the constructed pullout observations for C; A's key bent away first. D's eventual pullout involved the housing stop lip. No microscopy, strain map, lifetime distribution, or independent failure analysis exists. The stop can only react the specified outward wall load, and the moving surfaces can still jam. Reverse loading, impact, washing chemicals, temperature, creep over months, and service under wall preload are outside the exercise. The evidence here cannot establish field durability or novelty over real connectors. It only provides a finite, fictional comparison that still needs a clear technical explanation.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |