Recycled material in injection moulding: investigating batch-to-batch variability

A part made from recycled resin may meet the same specifications over several production runs, then show substantial changes when a new material batch arrives. Investigating those differences requires separating the incoming material characteristics from the conditions it experienced during processing.

Secondary material can vary in composition, contamination, degradation, thermal history, moisture and rheological properties. At the same time, changes in the mould, temperature, drying or auxiliary conditions may explain part of the observed variation. Traceability records need to preserve both groups of data.

Build a technical record for each batch

The incoming-material record can begin with the supplier, grade or trade name, batch number, declared origin, recycled content, certificate of analysis and the results of tests required by the application. Test selection should reflect the product requirements and the behaviour to be controlled.

The melt flow index can be useful for certain polymers and applications, provided the test method and conditions are recorded. Other applications may benefit more from measurements of moisture, density, mechanical properties, composition or contamination. The recycler’s records complement the physical characterisation results.

Link the batch to its processing window

At the injection moulding machine, associate each batch with the machine, mould, approved recipe, drying conditions, temperatures, pressures and relevant times. A change in cycle time or energy consumption may indicate an operational variation; interpreting it requires examining the material, product, machine and auxiliaries before assigning a cause.

When virgin and recycled materials are blended, record the proportion that actually enters the process. Identifying each blend and the transition periods helps establish which parts belong to a production run and which data can meaningfully be compared.

Use quality as an independent check

The assessment should include dimensions, appearance, mass, mechanical performance or other properties appropriate to the product. Record the measurement method, sample identifier, batch and manufacturing condition so that a deviation can be reconstructed.

A statistical correlation between temperature and rejects, for example, may help develop hypotheses. Confirmation requires suitable measurements and, where feasible, structured trials that separate simultaneous changes.

Share the findings with the supplier

Traceability should also support discussions with the recycler or compounder. If several batches sharing a characteristic require similar adjustments, that evidence can inform a supply specification, an incoming-material test or a revised blending strategy.

The structured record of recycled material can be linked to production traceability and quality information. The architecture is defined according to product requirements, the number of product variants and the company’s recording capacity.

A recommended first pilot

• Select a product that has shown variation between material batches.
• Agree which material characteristics and process conditions will be recorded.
• Maintain identifiers linking delivery, blend, order, machine and product.
• Define quality acceptance criteria and comparison rules.
• Review several production runs before drawing conclusions or establishing automatic rules.

The review by Andraju and colleagues (2022) examines the potential of machine learning to predict recycled-polymer properties and the limitations associated with incomplete data and heterogeneous materials. Models can add value when experimental characterisation is adequate and the dataset is sufficiently representative. Read the scientific review.

Further reading: the full study of digitalisation in plastics-processing SMEs and priority digitalisation capabilities.