Machining vs Moulding vs Both How the Combination of Injection Moulding and Machining Offers the Best of Both Worlds for High Performance Polymers
High performance polymers are used in the most demanding parts of the manufacturing world. Materials such as PEEK, PPS, polyimide, PEI and PESU are specified because they outperform metals and commodity plastics in punishing environments – but that performance comes at a price, both in the cost of the raw material and the difficulty in shaping it into a finished component.
Generally, two production routes dominate all polymers: machining from stock shapes or injection moulding. Each has a definite set of strengths and limitations, and for an increasing number of applications the most effective answer is to combine the two rather than choose between them.
Before we delve deeper, let’s understand each process and evaluate what each does well and where each fall short.
The case for – and against – machining
CNC machining is the workhorse of precise polymer manufacturing. It requires no tooling investment, it is infinitely flexible, and it can hold the close tolerances that critical applications require. For materials such as PEEK and PPS, tolerances of under 25 microns are becoming more common, as machining and tooling technologies evolve.
For prototypes and low-volume runs, machining parts from stock shapes — rod, sheet or billet — is almost invariably the sensible choice. One may argue that 3D printing may also be a viable option, however this is erroneous. 3D printing is great for approximating a shape but has been proven to be more or less useless in functional tests.
As the volumes increase, machining becomes problematic for a variety of reasons.
- Material consumption: High-performance polymers are expensive. PEEK prices are typically 12-50 times the cost of a commodity polymer like polypropylene or Nylon 6, and polyimide can be 10-15 times the price of PEEK again. When a part is machined from an oversized billet or rod, a lot of that expensive material gets machined off as swarf. In a complex part, you can waste more of the raw material than you use in the finished component
- Machining time: Complex geometries take time to cut, and time is money on a CNC machine. There are parts that may require tens of minutes of machining (and more!), so this becomes prohibitively expensive and practically infeasible when you are trying to make thousands of parts. Further, filled and abrasive PEEK or PPS grades – such as those that are glass or carbon reinforced - result in rapid tool wear, reduced throughput and increased costs.
- Residual stress: This is one of the least visible limitations, but one of the most important. Stock shapes have stresses locked in during their own manufacture. Machining removes material on one side more than the other, relieving that stress, and the part can warp or distort, defeating the very tolerances the machinist set out to achieve. This often includes intermediate annealing and stress-relieving steps, adding both time and cost.
The case for - and against - injection moulding
At the opposite end of the spectrum is injection moulding. It creates parts at or near their final shape, wastes very little material, runs at a fast cycle once the tool is in place, and can form features—undercuts, internal geometries, thin ribs—that would be difficult or impossible to machine. The economics get very attractive at scale.
But there are constraints to moulding high-performance polymers.
- Tooling cost: A hardened steel injection mould is a significant fixed investment. Abrasive fillers and, in the case of polyimide, PPS and PESU, corrosive gases released during processing temperature require robust tooling specification. That fixed cost is reasonable only if you spread it over a large number of parts.
- Shrinkage and tolerance: Semi-crystalline polymers like PEEK and PPS shrink significantly on cooling and crystallization and that shrinkage is seldom perfectly uniform, especially across sections of different wall thickness. Warping, due to differential cooling. It is very difficult to hold the very tightest tolerances right off the tool, whether it be a sealing face, a precision bore, or a mating diameter.
- Surface and finish: Gate marks, flash and flatness on critical faces all require attention after moulding and an as moulded surface may just not meet the specification a demanding application calls for.
The hybrid solution: first shape a blank, then machine it
Here the two processes cease to compete and start to complement each other. The idea is to injection mould a blank that is near-net-shape – a part close to final dimensions, with additional machining stock left only on the features that must be held to tight tolerance, rather than moulding a finished component. That blank is then machined. But only where it counts - the critical bores, faces and diameters.
The benefits pile up. The blank is brought close to final shape so that less material is removed than in machining from a solid billet, a direct saving on expensive polymers, with correspondingly less waste. The time to machine each part is greatly reduced when the machine is performing features rather than roughing out an entire component, which reduces the cost per part and increases throughput. And as a moulded blank with reasonably uniform sections has lower, more symmetric residual stress than a part hogged from a billet, there is less distortion after machining and therefore more stable, repeatable tolerances across a batch. Then the finishing operation delivers the precision and surface quality that moulding alone cannot - on precisely the features that need it.
A word of caution to keep in mind: rod strength vs. moulded strength
There is however an important trade-off that any designer should consider before committing to a moulded or hybrid part and this is mechanical strength.
The stock rod is manufactured under slow, controlled conditions – usually ram-extruded or compression-moulded, then annealed. In the case of semi-crystalline polymers this leads to a high uniform crystallinity and a largely isotropic structure. This means that the material behaves uniformly irrespective of the direction of loading. A part machined out of such a rod inherits these favourable bulk properties and has no weld lines.
Injection moulding has another tale to tell. As the melt flows into the cavity, polymer chains - and any fibre reinforcement - orient themselves along the flow direction, leaving the part anisotropic - stronger along the flow direction than across it. Where two flow fronts meet (around a hole, in a multi-gated tool) they form a weld line where the chains never quite fully entangle, creating a local weak point. The moulding process may also result in lower and less uniform crystallinity than an annealed rod because of the relative speed of moulding and, in filled grades, the moulding action tends to cut the reinforcing fibres shorter, thus reducing their effectiveness.
In practice, a part machined from a good quality rod will often be mechanically stronger, more predictable and more uniform than the same part made by moulding. The hybrid part, starting from a moulded blank, takes on most of the properties of the moulded material rather than those of the rod.
The penalty can be lessened – a hot mould and a good post-moulding anneal will increase crystallinity considerably – but it very rarely disappears completely. Where the overriding requirement is direction-independent strength, machining out of rod may still be the better route. For the great majority of components, however, the difference is small, well within the design margin and more than offset by improvements in tolerance, dimensional stability, material saving and cost.
The economics: where 'both' is the smart choice
Side by side, the logic of the three approaches is clear.
Machining alone has no fixed cost but high cost per part, due to the material waste and machining time. It’s the natural choice for low volumes and prototypes.
Moulding in itself has a high fixed cost in the mould but a low cost per part. The problem is it may not be able to meet the tolerances required of a precision component and may on its own be technically insufficient.
Moulding plus machining saves the cost of the mould but the cost per part is less than machining from solid because both material use and machining time are sharply reduced. Tolerances could never be achieved by moulding alone.
The important thing is that there’s a crossover volume. Below this volume, simple machining is still more economical. If you spread the fixed cost of the mould over a sufficient number of parts, it becomes a small figure per component, and the savings on material and machining per part more than pay for it. The hybrid way is not a compromise at all beyond that: it is cheaper than machining alone and more precise than moulding alone, all at the same time, as long as the modest strength trade-off has been considered in the design.
For parts specified in PEEK, PPS, polyimide, PEI or PESU in recurring quantities, the more useful question is not “machine or mould?” but “at what volume does doing both become worthwhile?” Usually – give the cost of the polymer - the answer can come surprisingly sooner than you expect.
