AGENT SKILL

mfg-reference

Use when correcting a manufacturing or mechanical engineering misconception, looking up a strength, tolerance, cycle-time or cost figure, finding the books, or needing a quick-reference picker — plus the current state of additive manufacturing after its correction and industrial robotics adoption. Companion to the other manufacturing and mechanical engineering skills.

License 许可
MIT
最近更新
2026-11-15
Stars
★ 5
Forks
3

01 前言 Front Matter

name
mfg-reference
description
Use when correcting a manufacturing or mechanical engineering misconception, looking up a strength, tolerance, cycle-time or cost figure, finding the books, or needing a quick-reference picker — plus the current state of additive manufacturing after its correction and industrial robotics adoption. Companion to the other manufacturing and mechanical engineering skills.

03 SKILL.md 正文

Manufacturing and Mechanical Engineering: What's Live, Misconceptions, Numbers, and Books

Part 5 of 5 of the Manufacturing and Mechanical Engineering for Software Devs reference (plugin manufacturing-mechanical-engineering-for-software-devs), covering §25–§30. Sibling skills: mfg-mechanics-stress-fatigue-and-materials (§0–§5), mfg-machine-elements-mechanisms-and-tolerances (§6–§8), mfg-process-families-machining-additive-and-moulding (§9–§16), mfg-dfm-metrology-plm-npi-and-what-transfers (§17–§24). Section numbers are shared across the set; a reference written as §N → skill points into that sibling skill.

Currency: The mechanics is settled. Two areas moved. See §25 for additive manufacturing after its correction, and industrial robotics adoption.

⚠️ Companion to a civil/industrial engineering reference. ⚠️ That one covers flow, constraints and safety systems; this one is about physically MAKING things — and the lessons are different.

⚠️ The governing asymmetry: in software the marginal cost of a change is near zero and in manufacturing it is enormous. ⚠️ Almost every practice in this document is a response to that one fact — tolerances, DFM, PLM, EVT/DVT/PVT gates, and the reason mechanical engineers appear conservative to software people.

⚠️ GOTCHA boxes mark the intuitions that don't survive contact with atoms.

The three ideas that organize this document:

  1. ⚠️ NOTHING IS EXACT — everything is a tolerance (§8 → mfg-machine-elements-mechanisms-and-tolerances). The single biggest mental shift for software people: there is no equality in the physical world, only distributions, and a design that requires exactness is a design that fails.
  2. ⚠️ The PROCESS determines the design, not the other way round (§9 → mfg-process-families-machining-additive-and-moulding, §17 → mfg-dfm-metrology-plm-npi-and-what-transfers). Geometry that's free in CAD can be impossible, or absurdly expensive, to make.
  3. ⚠️ FATIGUE kills things that were never overloaded (§4 → mfg-mechanics-stress-fatigue-and-materials). Most mechanical failures happen far below the static strength, after many cycles — the physical analogue of a bug that only appears after a million requests.

§25. What's Live — verified August 2026

25.1 ⚠️ Additive manufacturing after its correction

⚠️ The clearest recent example of a technology completing the hype cycle, and the post-correction picture is more useful than either the boom or the bust framing.

  • ⚠️ What happened: ⚠️ "valuations adjusted, consolidation increased, and several organizations were forced to retrench, reduce capacity, or narrow their scope," with the resulting correction producing today's more restrained conditions. ⚠️ The honest diagnosis offered is a mismatch between investment timelines and actual adoption rates — the responses "were not irrational; they reflected prevailing signals from capital markets" at the time.
  • ⚠️ The numbers now: ⚠️ AMPOWER reports the industrial AM sector returning to growth of 5.6%; ⚠️ Wohlers reports metal AM revenues growing 15.3% year-over-year in 2025. ⚠️ Metal PBF is simultaneously CONSOLIDATING AND FRAGMENTING — the ten largest system makers hold about 78.3% of global revenue, down from 80%, as new entrants (particularly Chinese) arrive.
  • ⚠️ Where the growth actually is: ⚠️ defence, space and consumer are named as central drivers, with metal AM growth exceeding 20% over four years as military OEMs use it for supply chain problems, drones and next-generation components. ⚠️ Apple's adoption of LPBF for watch cases is described as the largest metal AM application to date and as validating LPBF as a genuine production technology.

⚠️ GOTCHA — the durable lesson is the one §9 → mfg-process-families-machining-additive-and-moulding already implies, and it survived the hype cycle intact. ⚠️ AM wins on GEOMETRY, PART CONSOLIDATION and LEAD TIME, and loses on volume. ⚠️ Stated plainly in the current coverage: if you need millions of simple identical plastic parts, injection moulding is the better answer; if a part is easily stamped, cast or machined at scale, AM is not competitive. ⚠️ The consolidation win is real and quantified — assemblies of 20–30 components printed as one unit, with reported weight reductions of 40–60% in topology-optimized parts — which is §17 → mfg-dfm-metrology-plm-npi-and-what-transfers's "delete the part" achieved by a different route.

⚠️ The remaining bottlenecks are not the printers: ⚠️ qualification pathways, a persistent skills gap (with many institutions lacking industrial-grade equipment, creating a circular hesitation), and post-processing. ⚠️ Note also the strategic self-criticism from within the industry: executives at Formlabs, Carbon, Stratasys and Materialise reportedly converged on the view that "the industry took too long to focus on customer outcomes rather than on technology descriptions" — a failure mode software people will recognize immediately. ⚠️ Sourcing caution: this space is covered almost entirely by AM trade press with obvious interests; I've anchored on the AMPOWER and Wohlers figures and on the sceptical framing, which is more informative than the promotional material.

25.2 ⚠️ Industrial robotics: steady, large, and not what the headlines say

⚠️ Included because the automation discourse and the installation data have diverged.

  • ⚠️ The scale, from the IFR's World Robotics data: ⚠️ 542,076 industrial robots installed in 2024 — more than double the number ten years earlier, with annual installations topping 500,000 for the fourth consecutive year. ⚠️ Asia accounted for 74% of new deployments, Europe 16%, the Americas 9%. ⚠️ Around 4 million robots are operating in factories worldwide.
  • ⚠️ Market value reached an all-time high of US$16.7 billion for industrial robot installations, per IFR's January 2026 trends release.
  • ⚠️ US installations rose 11% year-on-year to 38,000 units in 2025, ⚠️ driven by food and other NON-manufacturing sectors — while automotive, still the largest single adopter at 13,500 units, was 1% below the previous year.
  • ⚠️ Robot density (robots per 10,000 manufacturing employees) ⚠️ is the IFR's comparison metric: Korea leads at 1,220, roughly four times the global average, with Singapore second at 818 — and the sources note Singapore's figure is partly an artefact of a small manufacturing workforce.
  • ⚠️ Automotive's share has fallen from a historical 35–40% of installations to around 28%, ⚠️ as electronics, logistics and general industry adoption broadened.

⚠️ GOTCHA — the humanoid story is where expectation and evidence diverge most, and the IFR's own framing is notably restrained. ⚠️ The IFR notes companies are "moving beyond prototypes to deploy humanoids in real life" while stating the conditions plainly: to compete with traditional automation, humanoids "need to match stringent requirements for cycle times, energy consumption, and maintenance costs," meet industry standards for safety and durability, and "achieve human-level dexterity and productivity" to fill labour gaps. ⚠️ The IFR has separately said mass adoption of humanoids, particularly as household helpers, may not happen in the near or medium term. ⚠️ Read that as the industry body being deliberately careful about a category attracting enormous capital — the specific, boring requirements list is the tell.

⚠️ The genuine driver is labour, not capability: ⚠️ employers struggling to find people, with over 600,000 unfilled US manufacturing jobs cited. ⚠️ And cobots — collaborative robots that work safely alongside people — are reported at entry prices around $25,000 with payback periods under 18 months, which is what actually moves adoption at the small-manufacturer end. ⚠️ Sourcing note: the IFR figures are primary and consistent across outlets; the cobot pricing and payback claims come from a lead-generation site and should be treated as indicative.


§26. Misconceptions

Misconception Correction
Stiffness and strength are the same ⚠️ Steel and stainless share E; strengths differ hugely (§2 → mfg-mechanics-stress-fatigue-and-materials)
Parts fail when overloaded ⚠️ Fatigue kills below yield, after cycles (§4 → mfg-mechanics-stress-fatigue-and-materials)
Aluminium has an endurance limit ⚠️ It doesn't. It accumulates damage at any amplitude (§4 → mfg-mechanics-stress-fatigue-and-materials)
Sharp corners are fine ⚠️ Stress concentration. Add a radius (§2 → mfg-mechanics-stress-fatigue-and-materials, §4 → mfg-mechanics-stress-fatigue-and-materials)
More bolts and pins make it stronger ⚠️ Over-constraint makes parts fight each other (§7 → mfg-machine-elements-mechanisms-and-tolerances)
Lock washers prevent loosening ⚠️ Split lock washers are largely ineffective (§6 → mfg-machine-elements-mechanisms-and-tolerances)
Tighter tolerances are better engineering ⚠️ They cost non-linearly. Tolerance the FUNCTION (§8 → mfg-machine-elements-mechanisms-and-tolerances)
Worst-case stack-up is the right method ⚠️ Often absurdly conservative; RSS is realistic (§8 → mfg-machine-elements-mechanisms-and-tolerances)
A dimension is a number ⚠️ It's a distribution. GD&T states intent (§8 → mfg-machine-elements-mechanisms-and-tolerances)
CAD geometry can be made ⚠️ Round cutters can't make sharp internal corners (§11 → mfg-process-families-machining-additive-and-moulding)
3D printing will replace manufacturing ⚠️ It loses badly at volume (§13 → mfg-process-families-machining-additive-and-moulding, §25.1)
Casting and forging are interchangeable ⚠️ Forging's grain flow gives better fatigue life (§10 → mfg-process-families-machining-additive-and-moulding)
Thick walls are stronger ⚠️ They sink in moulding and cast porous (§10 → mfg-process-families-machining-additive-and-moulding, §14 → mfg-process-families-machining-additive-and-moulding)
Welded joints are as strong as the parent ⚠️ The heat-affected zone is where it fails (§12 → mfg-process-families-machining-additive-and-moulding)
Adhesives are weak ⚠️ Strong in shear, poor in peel. Design the joint (§12 → mfg-process-families-machining-additive-and-moulding)
FEA results are answers ⚠️ Hypotheses. Validate against hand calc and test (§20 → mfg-dfm-metrology-plm-npi-and-what-transfers)
Tooling is a detail ⚠️ It's the capital cost and the lead time (§14 → mfg-process-families-machining-additive-and-moulding, §19 → mfg-dfm-metrology-plm-npi-and-what-transfers)
A revision is like a git commit ⚠️ Non-interchangeable means a NEW part number (§21 → mfg-dfm-metrology-plm-npi-and-what-transfers)
You can roll back a change ⚠️ Old revisions physically exist, for decades (§21 → mfg-dfm-metrology-plm-npi-and-what-transfers)
Just order a few for testing ⚠️ Setup and MOQ dominate at low volume (§19 → mfg-dfm-metrology-plm-npi-and-what-transfers)
Inspection ensures quality ⚠️ Capable processes do. Also check gauge R&R (§18 → mfg-dfm-metrology-plm-npi-and-what-transfers)
Robots are replacing manufacturing labour ⚠️ The driver is unfilled vacancies (§25.2)
Humanoids are about to transform factories ⚠️ The IFR lists the unmet requirements plainly (§25.2)
Mechanical engineers are just conservative ⚠️ Change costs six figures. It's rational (§1 → mfg-mechanics-stress-fatigue-and-materials, §24 → mfg-dfm-metrology-plm-npi-and-what-transfers)

§27. Numbers

⚠️ Steel E ≈ 200 GPa · Aluminium ≈ 70 GPa · ⚠️ same for most steels
⚠️ Aluminium density ≈ 1/3 of steel — ⚠️ and ~1/3 the stiffness
⚠️ Stress concentration Kt  ⚠️ typically 2–3× at holes and notches
⚠️ Bearing L10  ⚠️ life at which 10% have failed — statistical
⚠️ Rib thickness  ⚠️ ~50–60% of wall (injection moulding)
⚠️ Volume crossovers  ⚠️ 1–10 machine/AM · 100–1,000 soft tooling ·
                      10,000+ hard tooling
⚠️ AM sector growth   ⚠️ 5.6% (AMPOWER) · metal AM +15.3% YoY (Wohlers)
⚠️ Metal PBF top-10 share  ⚠️ 78.3%, down from 80%
⚠️ AM part consolidation   ⚠️ 20–30 parts → 1; 40–60% weight reduction
⚠️ Industrial robots installed 2024  ⚠️ 542,076 — 2× a decade earlier
⚠️ Asia share of installs  ⚠️ 74% · Europe 16% · Americas 9%
⚠️ Robot installation market value  ⚠️ US$16.7 bn (all-time high)
⚠️ US installs 2025  ⚠️ 38,000, +11% · automotive 13,500, −1%
⚠️ Robot density  ⚠️ Korea 1,220 · Singapore 818 per 10,000 employees
⚠️ Automotive share of installs  ⚠️ ~28%, down from 35–40%

§28. Books

Author Work Why
Shigley (Budynas & Nisbett) Mechanical Engineering Design ⚠️ The standard. §2–§6 → mfg-mechanics-stress-fatigue-and-materials, mfg-machine-elements-mechanisms-and-tolerances
Ashby Materials Selection in Mechanical Design ⚠️ §5 → mfg-mechanics-stress-fatigue-and-materials. Genuinely elegant method
Boothroyd, Dewhurst & Knight Product Design for Manufacture and Assembly ⚠️ §17 → mfg-dfm-metrology-plm-npi-and-what-transfers. THE DFA reference
Krulikowski Fundamentals of GD&T ⚠️ §8 → mfg-machine-elements-mechanisms-and-tolerances, accessible
ASME Y14.5 — ⚠️ The GD&T standard itself
Kalpakjian & Schmid Manufacturing Engineering and Technology ⚠️ §9–§16 → mfg-process-families-machining-additive-and-moulding, comprehensive
Slocum Precision Machine Design ⚠️ §7 → mfg-machine-elements-mechanisms-and-tolerances's exact constraint, done properly
Skakoon The Elements of Mechanical Design ⚠️ Short, principled, excellent
Adams Bearing Design in Machinery / manufacturer catalogues ⚠️ SKF and similar catalogues are real references
Machinery's Handbook — ⚠️ The trade's single-volume reference
Wohlers Report / AMPOWER Report — ⚠️ §25.1's primary data
IFR World Robotics — ⚠️ §25.2's primary data

§29. Quick Reference

29.1 Picker

Question Where
Which process should I use? ⚠️ Volume and geometry decide (§9 → mfg-process-families-machining-additive-and-moulding, §19 → mfg-dfm-metrology-plm-npi-and-what-transfers)
Why is this part so expensive? ⚠️ Tolerances, setups, or low volume (§8 → mfg-machine-elements-mechanisms-and-tolerances, §11 → mfg-process-families-machining-additive-and-moulding, §19 → mfg-dfm-metrology-plm-npi-and-what-transfers)
It's strong enough — why did it break? ⚠️ Fatigue, or a stress concentration (§2 → mfg-mechanics-stress-fatigue-and-materials, §4 → mfg-mechanics-stress-fatigue-and-materials)
Assembly binds or warps ⚠️ Over-constrained (§7 → mfg-machine-elements-mechanisms-and-tolerances)
Parts measure fine but don't fit ⚠️ Datum mismatch, or stack-up (§8 → mfg-machine-elements-mechanisms-and-tolerances)
Moulded part has sink marks ⚠️ Wall thickness. Core it out, rib it (§14 → mfg-process-families-machining-additive-and-moulding)
Machined part can't be made ⚠️ Internal corners, tool reach, or setups (§11 → mfg-process-families-machining-additive-and-moulding)
Should we 3D print it? ⚠️ Geometry, consolidation or lead time — else no (§13 → mfg-process-families-machining-additive-and-moulding, §25.1)
How do I cut cost? ⚠️ Delete parts first (§17 → mfg-dfm-metrology-plm-npi-and-what-transfers)
Can I just change the drawing? ⚠️ Is it still interchangeable? (§21 → mfg-dfm-metrology-plm-npi-and-what-transfers)
FEA says it's fine ⚠️ Check boundary conditions; validate physically (§20 → mfg-dfm-metrology-plm-npi-and-what-transfers)
Supplier wants an MOQ ⚠️ Setup costs are fixed per order (§19 → mfg-dfm-metrology-plm-npi-and-what-transfers, §22 → mfg-dfm-metrology-plm-npi-and-what-transfers)
What should software steal from this? ⚠️ §24 → mfg-dfm-metrology-plm-npi-and-what-transfers, and tolerance thinking first

29.2 Design review checklist

  • [ ] ⚠️ Tolerances justified by FUNCTION, not habit (§8 → mfg-machine-elements-mechanisms-and-tolerances)
  • [ ] ⚠️ Stack-up analysed for every critical fit (§8 → mfg-machine-elements-mechanisms-and-tolerances)
  • [ ] Datums consistent between design, manufacture and inspection (§8 → mfg-machine-elements-mechanisms-and-tolerances)
  • [ ] ⚠️ Fatigue considered where loading is cyclic (§4 → mfg-mechanics-stress-fatigue-and-materials)
  • [ ] Stress concentrations filleted (§2 → mfg-mechanics-stress-fatigue-and-materials)
  • [ ] ⚠️ Exactly constrained, not over-constrained (§7 → mfg-machine-elements-mechanisms-and-tolerances)
  • [ ] ⚠️ Manufacturable by the intended process — draft, walls, reach (§17 → mfg-dfm-metrology-plm-npi-and-what-transfers)
  • [ ] Assembly possible in one direction, with access (§17 → mfg-dfm-metrology-plm-npi-and-what-transfers)
  • [ ] ⚠️ Part count challenged: can any of these be deleted? (§17 → mfg-dfm-metrology-plm-npi-and-what-transfers)
  • [ ] Standard parts used wherever possible (§6 → mfg-machine-elements-mechanisms-and-tolerances, §22 → mfg-dfm-metrology-plm-npi-and-what-transfers)
  • [ ] ⚠️ Dissimilar metals checked for galvanic contact (§5 → mfg-mechanics-stress-fatigue-and-materials)
  • [ ] ⚠️ Interchangeability decided before the revision is released (§21 → mfg-dfm-metrology-plm-npi-and-what-transfers)

§30. Method

§1–§24 → mfg-mechanics-stress-fatigue-and-materials, mfg-machine-elements-mechanisms-and-tolerances, mfg-process-families-machining-additive-and-moulding, mfg-dfm-metrology-plm-npi-and-what-transfers rests on settled mechanics and standard manufacturing practice — stress and fatigue theory, materials science, GD&T as codified in ASME Y14.5, the process families, DFM/DFA, metrology, and PLM conventions. ⚠️ None of it needed verification; Wöhler characterized fatigue in the 1860s and the volume-versus-process economics has not changed.

Two searches were run in August 2026, on additive manufacturing and industrial robotics — ⚠️ both chosen because they are the two areas where software people's expectations diverge most from manufacturing reality, and both now have data rather than projections.

Confidence. High in §8 → mfg-machine-elements-mechanisms-and-tolerances, §7 → mfg-machine-elements-mechanisms-and-tolerances and §4 → mfg-mechanics-stress-fatigue-and-materials, which are the sections I'd most want read. ⚠️ Tolerance thinking is the single best import from this field into software: nothing is exact, everything is a distribution, and a design requiring exactness fails. ⚠️ Exact constraint (§7 → mfg-machine-elements-mechanisms-and-tolerances) is the second — over-constraint produces conflict rather than robustness, and the redundant-sources-of-truth analogy is exact. §4 → mfg-mechanics-stress-fatigue-and-materials's fatigue framing gives software a vocabulary for defects that only appear after many cycles.

High in §21 → mfg-dfm-metrology-plm-npi-and-what-transfers's PLM section, and ⚠️ the interchangeability rule is the part worth carrying: if the change makes the part non-interchangeable it needs a new identifier, not a revision — because you cannot recall what already exists. That is semantic versioning with genuinely irreversible consequences.

Moderate-to-high on §25.1. ⚠️ The AMPOWER 5.6% and Wohlers 15.3% figures and the 78.3% market-share number are attributed to named industry reports and recur across coverage. ⚠️ But essentially all of this space is covered by AM trade press with commercial interests, so I weighted the sceptical framing — the retrenchment narrative, the "took too long to focus on customer outcomes" self-criticism, and the plain statement that AM loses to moulding at volume — over the promotional material. ⚠️ The 40–60% weight reduction and 20–30 part consolidation figures come from a vendor blog and are marked as reported.

High on §25.2's IFR data, which is the field's authoritative source and consistent across outlets: ⚠️ 542,076 installations in 2024, Asia 74%, US 38,000 in 2025 at +11%, Korea's density of 1,220, and the US$16.7 billion market value. ⚠️ The humanoid framing is deliberately drawn from the IFR's own careful language rather than from vendor claims — the fact that the industry body lists cycle times, energy consumption, maintenance costs and human-level dexterity as unmet requirements is more informative than any forecast. ⚠️ The cobot price and payback figures come from a marketing site and are indicative only.

04 同仓库更多 Skills

the-vibey-project
the-vibey-project/vibey

white-paper-writing

Use when writing white papers, technical reports, policy documents, thought leadership pieces, or any long-form authoritative document. Triggers on requests to write, structure, outline, or review white papers, technical reports, position papers, policy briefs, or evidence-driven arguments. Also use for B2B content marketing strategy, technical documentation requiring authority, academic-style reports, or any document that must educate while persuading. Use when someone asks "how should I structure this" for a long-form document.

★ 5
the-vibey-project
the-vibey-project/vibey

prose-technical-mechanics

Use when editing or writing prose at a technical/mechanical level: choosing between Anglo-Saxon and Latinate diction, placing information in topic vs stress positions, writing periodic vs cumulative sentences, handling sentence rhythm, punctuation as logic, transitions, paragraph structure, or phonesthetics. Use when someone wants to understand WHY prose works or doesn't, not just surface-level style advice. Triggers on word choice, sentence structure, emphasis, rhythm, diction, or any deep writing mechanics question.

★ 5
the-vibey-project
the-vibey-project/vibey

prose-mechanics

Use when writing, editing, or analyzing prose style; when trying to avoid generic AI-sounding writing; when applying specific prose styles such as journalistic, academic, narrative, KJV biblical, Hemingway plain, Latinate/Jamesian, Orwellian, Augustan, Victorian, Early Modern (Bacon, Browne, Milton), or free indirect discourse; or when coaching someone on writing voice and mechanics. Triggers on any request to improve, analyze, or write in a specific prose style, to fix "AI-sounding" writing, to make prose more vivid or concrete, to vary sentence rhythm, or to understand how a particular writer achieves their effects.

★ 5
the-vibey-project
the-vibey-project/vibey

narrative-structure

Use when structuring stories, pitches, presentations, white papers, or any persuasive narrative; analyzing the hero's journey in fiction or non-fiction; applying StoryBrand for business communication; designing narrative arcs for thought leadership; or understanding Vogler, Snyder, Propp, or Campbell frameworks. Triggers on any narrative structure, storytelling, pitch structure, StoryBrand, hero's journey, three-act structure, Freytag, Save the Cat, narrative arc, or "how should I structure this story/pitch/presentation" request.

★ 5
the-vibey-project
the-vibey-project/vibey

legal-brief-writing

Use when writing legal briefs, motions, or any court filing; when analyzing judicial cognition and persuasion; when applying elite brief-writing techniques from Roberts, Clement, Kagan, Garner; or when structuring legal arguments for maximum persuasion. Also use for autistic/neurodivergent advocates who need to understand how their natural communication style maps to elite legal advocacy. Triggers on any legal writing, brief drafting, motion preparation, or legal argument structuring request.

★ 5
the-vibey-project
the-vibey-project/vibey

journal-publishability-criteria

Use whenever a research paper, article or preprint is being written, revised, reviewed or published, and specifically before every publication of docs/paper.md. Covers what legitimate journals actually gate on, in order of gating — the editor's desk screen (scope, novelty, package problems), what reviewers judge (methodological soundness, claims the evidence can carry, statistical and qualitative reporting), EQUATOR reporting guidelines, the ICMJE authorship criteria, COPE's view of text recycling, the near-universal generative-AI policy (no AI author, disclosure, author responsibility), TOP 2025 transparency standards, how to verify a journal is legitimate, submitting without an institution, and a pre-submission checklist. Triggers on requests to evaluate, review, prepare or submit a paper, to judge publishability, to write a declarations or data-availability section, to disclose AI use, or to check a journal.

★ 5