Define geometry
Mandrel radius, cylindrical length, dome type, boss opening, and coordinate frame determine the feasible surface paths.
AddWind Learn connects the manufacturing science behind filament winding with the software checks an engineer needs: pattern closure, geodesic behavior, finite tape width, thickness buildup, transitions, and machine kinematics.
A useful winding tool cannot stop at drawing curves. It has to connect structural intent, process constraints, laminate buildup, and the actual machine motion that deposits tape on the mandrel.
Mandrel radius, cylindrical length, dome type, boss opening, and coordinate frame determine the feasible surface paths.
Fiber, resin, tape width, thickness, friction assumption, and consolidation method define the production envelope.
Hoop and helical angles, skip/cycle integers, and coverage must close into a repeatable pattern without unacceptable gaps.
Finite-width rovings accumulate thickness unevenly, especially near polar regions and dense dome turnarounds.
Generated motion needs axis limits, correct tangent tape delivery, visible moving joints, and controller-ready output.
The first AddWind learning series is now live. These guides use AddWind simulator screenshots and conservative literature-grounded explanations for the core decisions behind filament winding.
Filament winding makes hydrogen tanks, COPVs and rocket motor cases by wrapping continuous fibre onto a rotating mandrel. An engineering walk-through with a browser-based 3D simulator.
2-axis, 4-axis, 6-axis robotic — what each axis on a filament winding machine does, how the payout eye decides where each turn lands, and how to read a winding job from a planner.
The three families of filament-winding paths — hoop, helical and polar — what each one carries, when to pick which, and how to combine them into a real pressure-vessel laminate.
Winding angle drives whether your fibre will even reach the dome. A visual explanation of Clairaut's relation, the boss-radius limit, and how 3° can move a part from windable to impossible.
What N, S and cycles actually mean in filament winding software. A visual guide to pattern closure, skip cycles, and how the cylinder gets uniformly covered.
Real fibre is a band, not a line. How band width, thickness, gaps and overlaps decide laminate quality, and how to see them in a winding simulator before they show up on the part.
Why every wound pressure vessel has thicker domes than cylinder, what happens at the boss turnaround, and how dwell strategy controls polar buildup.
How a filament winding simulation becomes a real machine program. The dry-run, the motion analysis, the export, and what to verify before you cut metal.
The broad AddComposites filament winding guide is the right entry point for outreach. AddWind should extend it with visual sub-lessons that use simulator screenshots, simple charts, machine playback clips, and diagrams before asking readers to care about advanced verification.
A beginner-friendly walkthrough of creel, resin bath or prepreg tow, tensioning, carriage motion, rotating mandrel, payout eye, and CNC control.
Show the difference between common winding patterns with one consistent AddWind vessel view and one developed-path chart.
Explain how 15, 30, 55, and 90 degree layers look, why they carry load differently, and why software must inspect the actual path.
Turn skip, cycle, coverage, and closed-pattern selection into a visual explanation that a non-specialist can understand.
These chapters are based on the local key-paper library and the AddWind implementation work already done for winding paths, transitions, thickness texture, and 4-axis playback.
A zero-width fiber path is mathematically convenient, but a real band has finite width and thickness. The effective polar opening shifts, the centerline is not exactly tangent to the opening, and the visible laminate can differ from the ideal trajectory.
Cylinders are straightforward because radius is constant. Domes introduce curvature, changing parallel radius, local surface normal, and the minimum path angle near the boss.
Geodesic paths are stable without lateral friction. Non-geodesic paths can be manufacturable, but only when the required slip resistance stays inside the available friction window.
A winding pattern is not complete until the skip, cycle, and circuit counts produce acceptable coverage. Structural layer requirements must be reconciled with the integer pattern that the machine can actually repeat.
Real winding often cannot cut fiber between layers. The connecting path has to move from one endpoint and angle state to the next without a discontinuity that the material or machine cannot follow.
Thickness is not uniform along a wound vessel. Dome regions can become dense because many circuits converge near the polar opening, while hoop layers mostly build the cylinder.
The visual machine model is not decoration. Operators need to see the rotary axis, carriage motion, vertical payout position, yaw orientation, and tape strip meeting the same contact point used by playback.
The last mile is controller confidence. Generated output needs the right units, coordinate conventions, feed interpretation, axis naming, and syntax for the target machine.
These articles are better for engineers who already understand the process and are evaluating software. They should sit after the 101 series, with links back to simulator projects and pilot feedback.
Why path drawing is no longer enough, and why modern winding software has to verify laminate, transitions, machine playback, and export packages together.
How to decide when friction assumptions are acceptable and how to inspect the angle profile.
Why real roving width and thickness expose gaps, overlaps, polar buildup, and false coverage.
What payout tangent, yaw, carriage, rotation, axis limits, and dry-run export need to prove.
A buyer checklist for closure, thickness, transitions, machine playback, export, and project persistence.
A premium filament-winding product should make engineering risk visible. AddWind should not merely calculate a pretty path; it should explain why a path is windable, how the laminate builds, and whether the machine can execute it.
The extracted 2004 source emphasizes a practical problem that appears throughout AddWind: an "optimal" vessel by ideal structural theory may not be optimal once roving dimensions, pattern integers, transition friction, excessive roving length, and non-uniform thickness distribution are considered. That is exactly why the software needs pattern search, friction diagnostics, thickness preview, and machine playback in one workflow.
The local paper folder groups the literature into design theory, path calculation, robotic winding, thermoplastic processing, low-cost machinery, and process comparisons. These references should drive the deeper lesson pages over time.
Netting theory, isotensoid pressure vessels, design principles, finite roving geometry, and laminate efficiency.
Geodesic and non-geodesic path planning, dry fiber winding opportunities, and manufacturability limits.
Offline programming, robotic filament winding cells, industrial RFWT applications, and machine coordinate systems.
Thermoplastic composite filament winding, in-situ consolidation, roving processing, and AFP comparisons.
Low-cost filament winding technology, laboratory equipment, cylinder manufacturing machines, and practical build constraints.
Pressure vessel challenges, square tubes, filament winding applications, and filament winding versus fiber placement.
Open the standard project library, then inspect the same concepts in the simulator: pattern closure, thickness texture, and production playback.