You will find the same claim in a lot of supplier literature: an autoclave-cured part is the real thing, and anything made by compression molding vs autoclave is the budget version. It is a tidy story. It is also wrong, and the error shows up in drawings, tooling budgets and rejected batches. These are two consolidation routes with different control variables. Neither is a grade of the other. What follows is the misconception, the physics each route manages, and the fields worth settling before a tool is cut.
Two boundaries. This is a process-selection discussion, not a claim about any part we make or any machine we run. Where we interpret how a specifying habit is moving, we label it as our assessment, not a statement of fact. We also keep performance figures qualitative: pressure, temperature and cycle numbers depend on the resin system and the part. They belong in a qualified process sheet, not a purchase drawing.
The misconception: a ladder instead of a fit
The shorthand runs like this. Autoclave curing is what aerospace uses, so it must produce the best laminate; compression molding is what high-volume industry uses, so it must cut corners. One word does a lot of work there. “Best” quietly becomes “lowest void content”, which becomes “highest strength”, and a process choice has turned into a material ranking.
That chain does not hold. A laminate’s behaviour comes from the fiber, the resin, the interface, the ply schedule, the fiber volume fraction and the consolidation history. The machine applying heat and pressure is one input to the last of those. Two routes can produce a sound, low-void panel, and two runs on the same route can produce very different ones, depending on how well the window was controlled. Our assessment is that the ranking survives mostly because it is easy to repeat, not because it predicts a test result. That is a judgement about a common specifying habit, not a statement of fact about any specific component.
We also see the reverse error: a team writes the autoclave into a specification for a flat, thin, high-count part, then inherits a long cure cycle and a hand-lay-up bottleneck nobody budgeted for. The process was chosen by prestige.
What each process actually controls
An autoclave is a pressure vessel. The lay-up sits on a single-sided tool under a vacuum bag, and the vessel applies heat and gas pressure to the whole assembly more or less uniformly in every direction. Because that pressure comes through a flexible bag, it conforms to curved surfaces, tapered sections and co-cured substructures. Temperature is ramped along the resin’s cure kinetics, held, then brought down slowly to limit residual stress. The route buys freedom of geometry and a gentle, uniform pressure field. It costs time: bagging is manual, the vessel is a batch asset, and a thick part occupies it for a full thermal cycle.
Compression molding uses a matched set of rigid tools. A charge, a cut prepreg stack, a preform or a thermoplastic blank, is placed in a heated cavity; the tool closes, holds pressure while the material consolidates, then the part is cooled and ejected. Pressure comes from the tool faces, so it is directional and set by how the tool closes, not by a surrounding gas. Heat gets in from both sides. The route buys short, repeatable cycles, tight dimensional scatter and near-net shape with inserts or local pads formed in one shot. It asks for expensive tooling and a well-controlled blank, and it is limited by press size and tonnage.
The resin system is the distinction most ranking charts miss. Autoclave processing grew up around thermoset prepregs, where the material must be held through a chemical cross-linking reaction: there is a cure to complete, and the cycle is built around it. Compression molding of a thermoplastic composite is a physical route: the matrix melts, flows under pressure, and solidifies again on cooling, with no cure reaction to wait for. That is why a thermoplastic blank can be heated, stamped and demoulded on a short cycle, and why a thermoset prepreg generally cannot be treated the same way. This is our reading of the general processing behaviour of the two resin families, not a statement of fact about a particular grade.
Both can consolidate a laminate to a low void content. Both can also produce porosity, dry fiber, resin-rich pockets, fibre wash or warpage if the window is wrong. The failure modes are similar; the leverage over them is different.
Where the autoclave earns its place
Start with geometry. Large, doubly curved skins, deep sections, integrally stiffened panels, sandwich constructions and co-cured assemblies all benefit from a pressure field that follows the surface rather than a pair of flat tool faces. A bag does not care that the part curves. A matched die does.
Volume and variety matter too. When a programme builds a modest number of complex parts, or when geometry is still changing between builds, a single-sided tool plus a bag is cheaper to revise than a hardened matched die set. Prototypes, qualification articles and low-rate structural parts tend to land here for that reason. Our assessment is that this is a tooling-economics argument, not a statement that the resulting laminate is inherently superior.
Autoclave quality is not automatic. Bag leaks, poor vacuum, bridge regions and thick-section thermal lag all create defects. A part made in a vessel still needs the same evidence as any other: void content, fiber volume fraction, and a record that the actual cycle stayed inside the qualified window. This is our assessment of what a specification should demand, not a claim about any supplier’s output.
Where compression molding earns its place
Now the other side. A flat or moderately shaped part, thin walls, a stable design and a real production count are the natural home of a matched-die route. Once the tool is qualified, each cycle follows the same closure, the same thermal path and the same ejection. Dimensional scatter between parts is small, which matters when the panel has to bolt into an assembly. Inserts, ribs, local build-ups and edge details can be formed in the same shot, cutting downstream machining and bonding steps.
Thermoplastic systems widen that advantage. Because consolidation is a melt-and-cool step rather than a cure reaction, the cycle is bounded by heat transfer, not chemistry, and the blank can be reheated and re-formed. That fits stamped organosheet, chopped-fiber compounds and continuous-fiber thermoplastic prepregs alike. It also changes the scrap story: offcuts and rejected stampings are at least candidates for reprocessing, where a cured thermoset part is not. We treat that as a design consideration rather than a promise, since reprocessing depends on the specific matrix and on how the material was handled. This is our assessment, not a statement of fact about a closed-loop result.
The limits are just as real. A matched die is expensive and slow to change, so a moving design punishes you. Very large parts run into press capacity. Deep draws, severe thickness changes and delicate cores can trap the charge or wash fibres out of position. And because the tool sets the pressure distribution, a poorly designed blank or a mis-loaded charge produces a bad part quickly and repeatably. Our assessment is that most compression-molding problems are blank and tool problems found late, not material problems; that is a judgement about where to spend engineering effort, not a measured finding.
How to specify: the fields that decide the process
The practical fix is to stop naming a machine in the first line of a drawing and to write down the part’s constraints first. Four fields do most of the work in an enquiry, and a fifth, the commercial frame, belongs in a separate conversation rather than folded into a process name.
| Field | Common misread | What it depends on | State in the enquiry |
|---|---|---|---|
| Geometry | The autoclave always gives the better laminate | Curvature, depth, plan size, thickness changes, press capacity | Give curvature and depth, not “simple” or “complex”. |
| Volume and takt | A short cycle means a lower grade part | Annual count, cycle length, tooling cost, dimensional scatter | State the annual count and the takt you need. |
| Resin family | One route fits every prepreg | Thermoset cure reaction vs thermoplastic melt and cool | Say thermoset or thermoplastic explicitly. |
| Consolidation evidence | One headline void figure settles it | Void content, fiber volume fraction, the qualified cycle record | Ask for evidence tied to the qualified cycle. |
A few habits help. Describe geometry as curvature, depth, plan size and thickness changes, not as “simple” or “complex”. State the annual count and the takt you need; those two do more to select a route than any property table. Say which resin family the design assumes, thermoset or thermoplastic, since that decides whether a cure reaction is in play. Ask for consolidation evidence tied to the qualified cycle, not a headline void figure from a brochure.
One more boundary. Process parameters such as pressure, temperature, dwell and ramp rate are specific to a resin system and a part, and are qualified by the converter that runs the tool. They should be confirmed per grade against the material datasheet and the process sheet before they go into a drawing. For orientation, our published fabric range covers 3K-24K tow, 50-600 g/m2 areal weight, and widths up to 1270 mm; anything outside that band is confirmed per grade. We are not presenting literature values or reported cycle figures as our own supply specification, and if a number is not in the grade datasheet, it does not belong in your drawing.
We supply carbon fiber fabrics in plain, twill and satin weaves, continuous glass-fiber products, thermoplastic prepregs across CF/PP, CF/PA6, CF/PC, GF/PC and GF/PA6 systems, and laminated plates in a small set of standard thicknesses that can be machined. The fabric, tape, prepreg and plate product pages on this site give the form and resin options. If you send us the four fields above, geometry, volume and takt, resin family, and the consolidation evidence you need, we will confirm the material form and the lay-up with you field by field. Whether the part then goes into a vessel or a press is a decision the part should make, not the other way round.
