cad-agent.References · Design process in the catalogue · Markdown · JSON
References · Practice
Design process
34 records: 1 read source that cad-agent's rules and checks may cite, and 33 records catalogued from the Reference Map, not yet read. Requirements and the dossier that holds them, integrity gates and section reviews, through to release.
Catalogued records are search snippets: their numbers are the authors' claims and their licences are as the map states them, unread. Cite one only after reading the source and promoting it (how).
Coverage
What cad-agent checked for these topics when the map was made, and the gap it found. The issue it drafted for each gap was never posted.
MissingTolerance stack-up (worst case, RSS)
Gap: No loop-sum helper: worst-case and RSS over a named direction, Cpk, tolerance-class defaults; src/cadagent/review/dossier/stackup.py does not exist.
Drafted issue, never posted: “Implement tolerance stack-up helper (worst case and RSS)”
MissingCost drivers and process break-even
Gap: No cost model: setup vs unit cost, tooling amortisation, FDM-vs-MJF-vs-CNC-vs-moulding break-even quantity, cost multipliers (tight tolerance, deep pocket).
Drafted issue, never posted: “Add process cost estimator and break-even helper”
Records
34 records
| Record | Kind | Evidence | Confidence | Licence | Process | Role |
|---|---|---|---|---|---|---|
| Jawstec: Why Heat-Set Insert Specifications Need ScrutinyJawstec Retells CNC Kitchen's insert tests: oversized holes lose pull-out, undersized ones are hard to install and may crack; size the hole for your insert and material and test it. No numbers. Read: 2026-11-04: opened the article (2026-09-04; the old ?p= link now redirects to its slug); read it whole. It retells CNC Kitchen's tests and gives no numbers of its own. | Guide | read | low 0.4 | proprietarynot read | Inserts and fastenersGeneral DFM | not assessed |
| 3D Printing vs Injection Molding Breakeven (Xometry)Xometry Breakeven comparison of printing against moulding by volume; only title seen in results. Note: Content unseen; sibling sources put breakeven at 200-2,000 units. | Guide | snippet only | low 0.4 | proprietarynot read | Moulding and castingGeneral DFM | not assessed |
| 5 Easy Ways to Reduce 3D Printing Costs (Xometry)Xometry Cost levers: machine time and material; smaller size, FDM infill, fewer supports, self-supporting angles near 45 degrees. Note: Qualitative; angle rule from combined search summary. | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Ashby materials selection method (CVUT course PDF)CTU Prague course notes Ashby charts: modulus vs density, strength vs density, and performance indices for light stiff design. Note: Lecture notes; CES EduPack and Ashby's book are primary | Paper or thesis | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| ASME Y14.5-2018 dimensioning and tolerancingSFS standards store US GD&T standard covering datums, form, orientation, position, profile and runout. Note: Store listing; full text paid | Standard | snippet only | low 0.4 | proprietarynot read | General DFMCNC machining | not assessed |
| Bearing shaft and housing tolerance (NSK Technical Insight)NSK Fit guidance: heavier loads need more interference; avoid tight outer-ring fit in split housings. Note: Figure from a related NSK insight page unconfirmed Keep max interference below about 7/10,000 of shaft diameter (Americas insight) | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Boothroyd Dewhurst DFMA presentation (Dewhurst, DFMA forum 2019)Boothroyd Dewhurst Inc. DFMA method from its originators; part-count reduction and assembly time estimation. Note: Seen as search result only; content not read | Paper or thesis | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| cq_gearsmeadiode (GitHub) CadQuery-based parametric involute gear modelling. Note: Apache-2.0, 166 stars, last push 2024-12-27 | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Design Tips for SLS 3D Printing wall thickness (Sinterit)Sinterit Printer-maker guide page on wall thickness across processes; contents not shown. Note: Only title and URL seen. | Guide | snippet only | low 0.4 | proprietarynot read | 3D printingGeneral DFM | not assessed |
| dfm-checkartprinting3d-web (GitHub) DFM review as CI: meshes, OpenSCAD, CadQuery, KiCad and Klipper configs, findings cited to rule ids. Note: MIT, 0 stars, last push 2026-08-28; unproven | Guide | snippet only | low 0.4 | proprietarynot read | 3D printingPCB and electronicsGeneral DFM | not assessed |
| DFMA explained (Wevolver)Wevolver DFMA overview: theoretical minimum part count Nmin, 2.93 s ideal assembly time per part, typical savings. Note: Secondary figures; primary source not found unconfirmed Published DFMA results: 20-50% fewer parts, 15-40% lower product cost (secondary). | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| DuPont General Design Principles for Engineering Polymers (mirror)DuPont (Stellenbosch-hosted copy) DuPont general design principles for engineering polymers; only the title appeared in results. Note: Third-party hosted; content unseen. | Guide | snippet only | low 0.4 | proprietarynot read | Moulding and castingGeneral DFM | not assessed |
| freecad.gearslooooo (GitHub) Gear module for FreeCAD. Note: GPL-3.0, 366 stars, last push 2026-09-15 | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Fundamentals of GD&T ASME Y14.5-2018 Foundational LevelSAE International SAE training course on Y14.5-2018 GD&T basics; paid course listing. Note: Course listing, not a free guide | Standard | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Geometric dimensioning and tolerancing (Wikipedia)Wikipedia Defines datum features, position, form controls at overview level. Note: Encyclopedic overview | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Hardox plasma/cutting guide (SSAB)SSAB Plasma kerf about 2 to 6.5 mm, HAZ 2 to 5 mm, tolerance about +/-1 mm; for thick plate. Note: Hardox plate focus; poor fit for thin plates. unconfirmed Plasma kerf roughly 2-6.5 mm; HAZ 2-5 mm | Guide | snippet onlyAuthors' claim: numbers stated, not measured | low 0.4 | proprietarynot read | General DFM | not assessed |
| IEC 62133-2 safety of portable sealed lithium cellsbatterydesign.net Lithium cell and battery safety requirements: overcharge, short circuit, forced discharge, mechanical and temperature tests. Note: Site has claims conflicting with publisher listings; use IEC text | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| ISO 2768 (Xometry)Xometry Xometry page on ISO 2768 general tolerances; appeared in results, content not shown in snippet. Note: Only title seen; content unconfirmed | Guide | snippet only | low 0.4 | proprietarynot read | CNC machiningGeneral DFMSheet metal and laser | not assessed |
| ISO 286 limits and fits: H7/g6 (Machining Doctor)Machining Doctor ISO 286 fit explanation: H7/g6 sliding clearance; H7/p6 and r6 interference. Note: Calculator pages; some errors seen across sources unconfirmed Dia 25 H7/g6 clearance 7-41 um; H7/p6 interference about 1-35 um. | Guide | snippet onlyAuthors' claim: numbers stated, not measured | low 0.4 | proprietarynot read | CNC machiningGeneral DFM | not assessed |
| ISO/ASTM 52900:2021 additive manufacturing vocabularySCC (Standards Council of Canada) listing Current AM terminology standard from ISO/TC 261 and ASTM F42; 2015 edition withdrawn. Note: Catalogue listing; full text paid | Standard | snippet only | low 0.4 | proprietarynot read | 3D printingGeneral DFM | not assessed |
| MatWeb Nylon 6 carbon fiber filled groupMatWeb Material property database; group index of nylon 6 grades incl. carbon fiber filled from RTP and Techmer. Note: Index only seen; property values not in snippet | Guide | snippet onlyAuthors' claim: numbers stated, not measured | low 0.4 | proprietarynot read | 3D printingGeneral DFM | not assessed |
| Minimum Wall Thickness for 3D Printing (Formlabs blog)Formlabs Blog on minimum wall thickness across printing processes; only title seen in results. Note: Only title and URL appeared; content not shown. | Guide | snippet only | low 0.4 | proprietarynot read | General DFM3D printing | not assessed |
| Protolabs: 3D Printing CostProtolabs Explains cost drivers of 3D printing; advice includes adding material only where critical. Note: Page text mostly not visible. | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Protolabs: 6 Ways to Cut Costs with Industrial 3D PrintingProtolabs Cost-cutting design tips; orientation changes can shorten builds, less material means lower cost. Note: Qualitative advice only. | Guide | snippet only | low 0.4 | proprietarynot read | General DFM3D printing | not assessed |
| Protolabs: 7 Mistakes to Avoid When Designing 3D Printed PartsProtolabs Design mistakes list; notes warpage risk rises as thickness nears minimum feature size. Note: Cost content not confirmed from snippet. | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Protolabs: Understanding 3D Printing TolerancesProtolabs Protolabs design tip on 3D printing tolerances by process; only title appeared in results. Note: No numbers visible in snippet. | Guide | snippet only | low 0.4 | proprietarynot read | General DFM3D printing | not assessed |
| Race to 1,000 Parts: 3D Printing vs. Injection MoldingFormlabs Worked cost model: one part with a $3,600 tool, breakeven versus printing, with resin purchase sensitivity. Note: Vendor selling printers; single-part case. unconfirmed Breakeven 13,050 parts at $3,600 tooling; bulk resin could push it past 40,000 | Guide | snippet onlyAuthors' claim: numbers stated, not measured | low 0.4 | proprietarynot read | Moulding and castingGeneral DFM | not assessed |
| Scholz: Tolerance stack-up analysis (Boeing review, UW copy)University of Washington (F. Scholz, Boeing) Rigorous review of arithmetic and statistical stack-up methods and their assumptions. Note: Only table of contents seen | Paper or thesis | snippet onlyAuthors' claim: numbers stated, not measured | low 0.4 | proprietarynot read | General DFM | not assessed |
| Shaft and housing fits (NSK Technical Report 3)NSK Bearing fitting practice: required interference, temperature effects incl. aluminum and plastic housings. Note: Only table of contents seen | Guide | snippet onlyAuthors' claim: numbers stated, not measured | low 0.4 | proprietarynot read | General DFM | not assessed |
| Shaft and housing fits (SKF catalog excerpt)SKF Tables for shaft and housing fit selection of metric radial bearings; clearance versus interference depends on which ring rotates. Note: Table values not visible in snippet | Guide | snippet onlyAuthors' claim: numbers stated, not measured | low 0.4 | proprietarynot read | General DFM | not assessed |
| Tol-Ninjaslehmann1 (GitHub) Tolerance stackup analysis software. Note: MIT, 9 stars, last push 2022-06-18; stale | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Tolerance stack-up analysis: worst case vs RSS (Wevolver)Wevolver Worst-case versus RSS stack-up methods and when to use each. Note: Search summary flagged an arithmetic slip in a related example corroborated Four +/-0.05 mm parts: worst case +/-0.20 mm; RSS about +/-0.10 mm (4 parts). | Guide | snippet only | low 0.4 | proprietarynot read | General DFM | not assessed |
| Xometry tolerances of machined and sheet metal partsXometry Knowledge-base article on tolerances of machined and sheet metal parts; no figures seen. Note: Title-level read only. | Guide | snippet only | low 0.4 | proprietarynot read | Sheet metal and laserGeneral DFM | not assessed |
| Xometry: injection molding vs 3D printing break-evenXometry Break-even of printing versus molding by volume; sources quote 250-2000 parts but vary widely. Note: Vendor snippet; break-even range differs by source unconfirmed Break-even qty = mold cost / (alt cost per part - molded cost per part). | Guide | snippet only | low 0.4 | proprietarynot read | General DFMMoulding and casting3D printing | not assessed |
Role is the survey's judgement of a record within its class (leader, reference, lessons, context): not a quality grade. A single source, a search snippet, or untested judgement is what “low” means.